Modelling Friction in Cold Rolling
Leon Jacobs is a PhD student in the department Surface Technology and Tribology. (Co)Promotors prof.dr.ir. M.B. de Rooij and prof.dr.ir. D.J. Schipper from the faculty Engineering Technology.
The cold rolling process is an essential step in the production chain of high quality steel strip that is used in applications such as automotive, packaging or construction. Cold rolling models are of crucial importance for setup of the cold rolling mill, defect analysis, development of new steel grades or for design of new rolling mills. Unlike conventional cold rolling models, mixed lubrication models explicitly take the influence of the lubricant into account. This results in a physically based description of the friction between work roll and strip which is necessary to predict crucial rolling parameters as rolling force and forward slip with higher accuracy.
This thesis aims to further advance mixed lubrication models for cold rolling, with particular attention to:
- A physically based description of the friction mechanisms that play a role in the roll bite.
- The validity of the main assumptions in cold rolling models.
- An extensive experimental validation of the complete cold rolling model.
The tribological core of the model accounts for three friction mechanisms that play a role during cold rolling. Similarly to existing mixed lubrication models, boundary layer shear stress (or ‘adhesive friction’) is proportional to the true contact ratio. The adhesive friction factor is determined by fitting with experimental rolling results. This work demonstrates that both ploughing friction and viscous shear stress cannot be neglected and must be integrated in the mixed lubrication model:
- Ploughing friction is modelled based on results of Material Point Method simulations of a moving spherical indenter through a substrate that shows work hardening behaviour. This spherical indenter represents a single asperity on the work roll.
- Viscous shear stress is modelled based on an experimental characterisation of the rheological properties of the rolling lubricant. It was shown that at pressures, temperatures and shear rates that are common in cold rolling of low Carbon steel, the investigated lubricant shows non-Newtonian behaviour. Furthermore, at high pressure the viscosity increases less than exponential, indicating that Barus-law overestimates the lubricant viscosity.
Furthermore, in this work it was found that plastic yielding behaviour of cold rolled steel strip is significantly anisotropic, described by a Lankford parameter smaller than 1. Therefore the isotropic von Mises yield criterion that is commonly used in cold rolling models is replaced in the mixed lubrication model by the Hill48 yield criterion which accounts for the anisotropy, resulting in a more accurate description of the contact behaviour in the roll bite.
Not only the most important output of the mixed lubrication model (rolling force and forward slip) is experimentally validated, but relevant intermediate results as well. During film thickness measurements using the so called droplet experiment, it was revealed that the strip can enter the mill asymmetrically resulting in a different lubricant film thickness on top/bottom strip surface. With an additional experimental procedure, a perfectly symmetrical material entrance could be guaranteed. For such rolling processes, the experimentally determined lubricant film thickness corresponds well with the model results. Other crucial intermediate results of the model were also experimentally validated: the contact length between roll and strip, the load carrying capacity of the strip, anisotropic yielding behaviour and the contribution of viscous shear stress to friction in the roll bite.
Experimental validation of the complete mixed lubrication model suggests that viscous shear heating significantly influences viscous shear stress in the work zone in an actual rolling process. When this effect is accounted for (by a simple relation), the model outcome corresponds better to the experimental results. Another conclusion from the experimental validation is that taking anisotropic plastic yielding behaviour into account leads to a significant improvement in model performance. Key recommendations following from this work are to integrate viscous shear heating and two phase lubrication (O/W emulsions) via a more fundamental way into the mixed lubrication cold rolling model. The final conclusion is that the mixed lubrication model, for a wide variety of cold rolling processes, significantly better predicts the rolling force and forward slip than a conventional rolling model.
