Finite element–based inverse approach to estimate the friction coefficient in hot bar rolling process

Abstract

This article proposes a finite element analysis–based inverse approach to estimate the friction coefficient in hot bar rolling. The focus is to minimize the difference between the spread of material measured from the pilot hot bar rolling test and that computed from finite element analysis. The recursive response surface method was used with a changed observation range to minimize the difference. The pilot hot bar rolling test was conducted at temperatures ranging from 850 °C to 1150 °C and reduction ratios from 20% to 40%. Finite-element simulation of the pilot hot bar rolling test was carried out. A fast running model that can rapidly determine the friction coefficient at the arbitrary reduction ratios and temperatures in the ranges mentioned above was also presented. The estimated friction coefficient was approximately 10%–17% higher than the friction coefficient typically used in hot strip rolling. The effect of temperature variation on the friction coefficient was greater at higher reduction ratios (30%–40%) than at a lower reduction ratio (20%).