Transversal cracks and their behaviour in the hot rolling of steel slabs

The behaviour of transversal cracks in reverse rolling of steel slabs is analysed. For this purpose a commercial FE-program, LS-Dyna3D has been utilised. Process parameters from an industrial rolling mill have been used as a reference. One basic assumption is that the cracks do not propagate for the steel considered. This is in agreement with observations in industry. Furthermore the material is treated as elastic–plastic. The aim of the study is to investigate the possibility of controlling the plastic deformation so that the cracks disappear at the same time as their boundary surfaces are prevented from getting into contact with each other. If they should get into contact, a defect in the form of an oxide flake should be created. The optimum situations are assumed to be found when the depth of the crack decreases more rapidly than the height of the slab at the same time as the crack width increases. For such rolling conditions it should be possible to find the crack bottom on the workpiece surface without any trace let by the original crack. During rolling two phenomena are present, counteracting each other. At the entrance to the roll gap, the crack widens, which is regarded as beneficial. The explanation to the phenomenon is that when the front edge of the crack gets into contact with the roll it is dragged towards the gap at the same time as its velocity is rapidly changed from horizontal to become parallel to the tangential velocity of the roll. If this phenomenon did not occur, the risk for the creation of a permanent oxide flake should be big. During the passage of the gap the widened crack angle is decreased. The decrease of the angle is explained by the high hydrostatic pressure that is associated with the roll/workpiece contact surface. At the exit the angle is slightly increased once again which might be explained by the fact that the back edge of the crack is still in contact with the roll and its movement is retarded because of friction. Thus the final result depends on which of the two phenomena, widening or contraction that has been dominating. The analysis shows that light reductions/pass, small roll radius and high friction are preferred. A deep crack of V-shape is easily transformed to Y-shape. This means that a remaining defect should be formed at the bottom of the crack. Reverse rolling is found to be beneficial because this way of rolling implies that the crack does not tilt so easily which should result in the formation of an oxide flake.