Multi-variable relationships in a batch annealing process

Empirical models are developed to predict the mechanical properties of fourgroups of cold-rolled annealed steels, based on upstream processing variables in thehot-rolling process, annealing conditions, the chemical composition of the steel andproduct details. Various linear models are considered but the simplest approach ofusing a multiple linear regression model is considered the most satisfactory. Thisanalytical tool will allow better understanding of the steel making process and theability to vary input parameters to improve the product. In particular the numberof coils which fail mechanical testing may be able to be reduced, with a subsequentfall in production costs.