Q: How do I know if winding caused my defect?

David Roisum, Ph.D., Consulting Technical Editor 920/725-7671 DRroisum@aol.com www.roisum.com -- Converting Magazine, 4/1/2001

A: While winding causes many troubles, it is also convicted of crimes of which it is innocent. For example, if you see a defect on the wound roll, it would be tempting to conclude that it was due to winding. However, you may first see the defect on the winder simply because it is much more convenient to see there. To be certain the defect was not made elsewhere, it is wise to sample upstream of the winder. Sometimes the winder made the defect, but did so with help from manufacturing. Examples include corrugations, gage bands and ridges, which occur when caliper-varying web is wound. While some winders and process settings might minimize the severity, the basic problem remains a varying caliper. It is important to note that the wound roll is by far the most sensitive measure of caliper. Features can show up on the roll which easily exceed the resolution of the best lab test techniques and online scanners.

The first evidence that the problem is related to winding is if the problem is described in references such as Roll and Web Defect Terminology, The Mechanics of Winding, and The Anthology of Winding (all published by TAPPI Press). In these references, the appearance, trends and symptoms are used for diagnosis. Also, the mechanics may be described, so that you know why the defect occurs, perhaps enabling you to make predictions about other behaviors. Good science yields theories general enough to make predictions beyond those used to generate the theories in the first place.

The second evidence that the problem is related to winding is the distribution of the defect. If the defect is diametrally dependent, then it is almost certainly related to winding. Examples of diametral dependence are troubles that occur near the core. The details of the insult may be due to non-cylindrical cores, flimsy cores, bad starts or higher interlayer pressure. The pressure at the core is almost always higher than elsewhere due to winding physics, as well as from loads passed from the core to the inside of the roll when supported by the core.

Physics of winding

Other examples of diametral dependence are defects that disappear toward the very outside of the roll. Winding physics guarantees that the interlayer pressure drops to zero at the outside. Finally, the defect may prefer some intermediate diameter. This might indicate problems with the winding machine or control. To determine diametral dependence, you must make a frequency-distribution diagram of the defect with respect to radial position in the roll.

If the defect prefers specific cross-direction (CD) locations, the problem is almost certainly related to variations in manufacturing upstream. Most web makers and converting processes have banded fingerprints that vary more with respect to CD position than with machine-direction (MD) position or time. It is possible that the winder takes manufacturing bands and turns them into defect bands. If so, they might retain the diametral preference described above.

The last primary evidence that the defect is winding related is that the problem is sensitive to wound-in-tension. In other words, troubles become better or worse if winding is tightened or loosened. To determine this relationship, you will need to use DOE (Design of Experiments) to construct trials with either very exaggerated tension changes or with very large sample sizes.

These three evidences are a good starting point, but are not all-inclusive. For example, you may get defects related to winding nip (wrinkles or air bubbles behind the nip) or lack of winding nip (air entrainment) that do not conform to the above descriptions. You may also see time dependent behavior that may or may not be related to winding. Perhaps this is why winding is often blamed for difficulties. It is tempting to indict the factor whose complexity we do not understand.

If you found this article helpful, ENTER 201 or Inquire Online.