Performance Analysis of Work-Roll Wear Models on Hot Rolling

The purpose of this work is to evaluate the performance several wear models, either with different mathematical formulation or definition unknown coefficients, on prediction work-roll amplitude in Hot Strip Mills (HSM). To achieve goal, a classical model calibration approach based inverse optimization has been developed calibrate these models. A large industrial hot rolling database composed by roll measurements for both later finishing mill stands (F6 and F7) from ArcelorMittal Dofasco HSM was considered least-square cost function applied minimize differences between numerical experimental results during process. averaged gap optimized predictions then used as quantitative indicator compare models identify most suitable one prediction. In addition, an Artificial Neural Network (ANN) model. Thus, obtained using ANN were compared ones Classical approaches.