Every winder is equipped with an adjustable tension. Most winders can taper this tension automatically as the roll diameter builds. Some winders can even program a curve. Whether we have a simple setpoint, linear taper or arbitrary curve, we must choose how to set the tension. The best setting is determined the same way here as with any other adjustment: economics. Specifically, we choose a tension which avoids the most, but not necessarily all, defects.

We can group defects into the following categories: low tension, high tension, taper tension or independent of tension. Examples of low-tension defects are out-of-round rolls, some rough roll edges and loose cores. Examples of high-tension defects would include blocking, gage bands and some crushed cores. Some defects are independent of tension, such as wrong diameter rolls due to an operator setup error. A couple of defects, namely subsets of starring and tension, are very sensitive to a lack of taper.

If you had customer complaints of out-of-round rolls, you would look into roll handling practices as well as tighten up the tension. If you had loose cores, you would also consider increasing the tension. You also must check to make sure you don't have wet cores. Increasing tension may not be strong enough to counter shrinking cores.

If, on the other hand, you had gage bands you would look into manufacturing. Obviously, you would also decrease winding tension as much as possible. However, if the gage bands are severe enough you will not eliminate the problem solely by tension changes.

So how far do you move the tension in response to a defect? The answer is always the same: until you break something. In other words, if you have a high-tension defect you will reduce the tension until you clearly have a low-tension problem of some sort. Once you are clear what and where the limits on both ends are, try running half way between. If you are lucky, you will find a sweet spot where both high- and low-tension defects are absent. If not, you may well have both high- and low-tension defects at the same time. Next steps may be to redesign the process or do economic optimization as described in my Critical Thinking book.

We use the same approach for the taper-sensitive defects—starring and telescoping. We make sure that a particular case is amenable to tension treatment. For example, starring on one side of a wound roll is not treatable by tension because it is probably caused by a gage variation. Gage variations may be unintentional, due to profile variations in manufacturing, or intentional such as printed patterns. If your case is amenable to tension, the appropriate amount of taper is maximum. You would start the roll at maximum, just short of breaking something. You would finish the roll at minimum, again just short of breaking something.

This tension strategy outlined here makes sense for several reasons. First, it is based solidly on economics rather than theory. Second, it adapts to any particular situation rather than conforming to a "one size fits all" guideline that probably fits very few very well. Third, anyone can get the answers if they are willing to spend the effort and ruin a few rolls finding the limits.

Notice that we have not promised complete relief from defects. Expecting any single knob to make the hurt go away completely is often unrealistic. Product/process design always offers alternative solutions, and they should also be considered. Sometimes design is so much more powerful that continuing to think about tension is distracting. Even so, tension is always the right place to start.

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