Abstract
Precision control of the width of slabs is vital for product quality and production economy in steel rolling mills. However, the formation of so called 'dog-bone' at the edge of the slab would affect the final width during the horizontal rolling that follows. Therefore, it is essential to predict and control the dog-bone shape. In this paper, a model is derived by using a number of finite element simulations for edge rolling and a least square regression analysis. The prediction accuracy of the proposed model is examined by comparing the predictions from finite element simulation with experiment results in the literature.