Dalarna University's logo and link to the university's website

Thermal expansion, elastic deformation and roll wear are some of the main factors influencing work-roll geometry in the hot rolling process and thereby rolling gap appearance and finally the flatness and profile of the hot rolled material. The purpose of this industrial project was to compile the parameters that influence strip crown of Steckel hot rolling mill, where the influence of these parameters has significant impact to optimize the roll adjustment to achieve a greater availability and higher yield. Focusing on roll wear influence, the evaluation of the grinding depth of work rolls toward roll wear was conducted. In addition to that, the roll wear prediction model in the strip width was developed and roll wear resistance of work-rolls used in a Steckel hot rolling mill has been evaluated. The rolled material used in this project was martensitic stainless chromium steel.

Hardness measuring with Leeb-method shows that softening occurs in the contact area between strip and work roll, which is still present after a re-grinding process. The grinding depth was varied to investigate the influence on the measured softening. It was clear that an increased grinding depth reduced the softening and the work roll attained a hardness condition close to the one of an unused roll. Wear is a function of the hardness of the roll and thus the wear of the work roll due to different grinding depths was analyzed during actual rolling. The results showed that the wear coefficient remained the same even though the softening was reduced by increasing the grinding depth.

A model to calculate the wear as a function of the position on the work roll width has been developed in this project. This was done to be able to model the local wear variations during a campaign. This was achieved by simulating the varying force along the contact between strip and work roll and between work roll and back-up roll. The software program used in the evaluation is CROWN426.

 By using Design of Experiments and the software MODDE, a polynomial fit was assigned to the rolling force between the work rolls and the strip (F₁) and the force between work rolls and backup rolls (F₂) by MLR and PLS, respectively, for faster calculation and the ability to use the model as an on-line tool. Finally, the polynomials, F₁ and F₂, were used to calculate the wear as a function of the force and number of rotations of the work-roll in contact with the strip. The results show that the model can predict the total varying wear with less than 7 % maximum average error for the rolled steel grade, martensitic stainless chromium steel.