Yield, quaIity and throughput increase by optimising the transfer bar and finished strip geometry in hot strip mills

For predetermining the transfer bar width, a width prediction formula was advanced to calculate the width after the vertical and horizontal rolling. The comparison of measured industrial data and calculated values shows a very good conformity (R2 = 0.9997). Using FEM calculations, the spread of the free extremities in the successive vertical and horizontal passes was investigated. Regarding the avoidance of transfer bar camber, a camber measurement system was installed at a roughing mill for calculation of the curvature of the bar. Based on the calculated curvature, a statistical evaluation of process data was performed. Camber influencing parameters were defined and verified and a neural network method was used to model the camber generation at a roughing mill, with a quite good result. The correlation between slab dimension variations and hook development in roughing was investigated. Slab dimension variations showed weak correlation with hooking. Laboratory tests were performed to define and verify the influence parameters on camber. The tests have shown that the side guides have a major effect on both the camber and the wedge forming. Based on FEM calculations, rolling mill tilting is found to be an effective tool to control hooking. But to be used, it is necessary to have online information about hooking and strip thickness distribution. In addition, a camber simulator was developed that consists of a stand stretching model and the bar geometry model. Using the camber simulator, camber-related investigations on stand deformations, the entry guide effects and the asymmetric parameters as initial wedge, stand levelling, etc. were performed. For the avoidance of the tail-end chewing phenomenon, operating practices based on quantitative and qualitative relationships have been developed. The introduction of the countermeasures on the mill determined a significant decrease of the phenomenon.