One of the fundamental rules in web handling is a web always tends to enter a cylindric roller normally—the middle of web will be perpendicular to the axis of the roller. When a roller is misaligned in a web line, the web will steer laterally on the roller until a normal entry condition is achieved, thus, induce a bending deformation in the span of web prior to the misaligned roller. The shear forces developed between web and the misaligned roller have the potential to cause instability of web in spans or on rollers. We used explicit finite element analysis to investigate the instability of web due to a misaligned roller.
Simulations showed the evolution of instability of web due to a gradually misaligned roller. First troughs form in the span. When misalign angle achieve certain value, a micro wrinkle enters the misaligned roller and moves across it, forming a wrinkle on the roller. Last, a foldover formed on the roller. The defect on the web then transported through the whole virtual web line downstream. This evolution is exactly consistent with the lab observation.
A wrinkle and foldover formed and passed through the web line.
A closer look of at the simulation results reveals that the misaligned roller induces the development of a pocket of negative lateral stress prior to the web’s entry point to the roller, which causes the formation of a micro wrinkle. The wrinkle then climbs on and over the roller, a foldover ultimately forms.
Development of the negative lateral stress during the evolution of troughs – wrinkles – foldovers
The discontinuous lateral displacement of web due to wrinkles
This work was later extended to help the development of a failure criterion that predicts the critical buckling stress and critical misaligned angle for web instability in a web line. The criterion provides a guidance for web line design and assembly and defect tracing.
The key of successful modeling this problem is the connector constraint and boundary condition that depict the movement of the misaligned roller. If they are not set up right, we may see some really odd behaviors like this one, which may be viewed as the problem of a loose roller or a special process?
Misaligned roller is swaying about its center point.
Hi,
This is impressive work! I have couple of questions :
1. Can I know what kind of boundary conditions you used in the simulation as well as material properties (I mean its are non linear material properties?)
2. How you are controlling stiffness of web? Is it dynamic value or constant value?
3. Can we run this for high Temperature of roller ?
4. Are you using shell meshing for web?
Thanks,
Sandip
Hi Sandip, just saw your questions, here are some better late than never answers:
1. Tension and longitudinal (MD) velocity BCs are applied at the upstream and downstream edges respectively. Only elastic material model was used for this specific case.
2. Stiffness of web should be taken care of automatically when the geometry and material of web are modeled accurately.
3. You can easily add a spatial temperature field to approximate a hot roller using UTEMP if simulations are run with Implicit. With Explicit, a coupled thermal stress analysis probably is needed.
4. Yes, shell elements.