Dimensional Ranges and Rolling Efficiency in a Tandem Cold Rolling Mill

摘要

In this work, physical models and a neural network theory have combined in orderto predict the properties of a steel strip and to optimize the process parameters in cold rolling. The prediction of the deformation resistance of the material and the friction parameter is based on the physical model presented by Bland, Ford and Ellis and artificial neural network computing (ANN). The accuracy of these models has been tested and proved by using a large amount of the measured data. With the aid of these models it has been shown that (1) the small change to the relative reduction distribution can have a clear effect upon the rolling efficiency. (2) the dimensional ranges of the tandem cold roll mill can be determined and optimized and (3) the possibility to cold roll a new product of new width, strength or thickness can be determined and the parameters of the tandem cold rolling process can be optimized. (Copyright (c) 1997 Helsinki University of Technology.)