Your eyes. Seriously. If you look at the 100+ defects in the encyclopedic Roll and Web Defect Terminology, you will see a picture for every single defect. However, not a single measurement is given anywhere in this must-have reference. If you consider the customary practice in the industry, visual culling is the norm, and roll quality measurements are not.

Visual rejection is neither quantitative nor consistent. Since it is not quantitative, statistics and other tools are handicapped by the binary nature of the go/no-go data as opposed to the richness of analog data. Since visual measures are not consistent, our saleable product is not as well.

The only way we can make the product more consistent is to reject anything that is visible. There are advantages to this approach. First, it avoids the foolishness of distinguishing between functional and cosmetic defects when there are none. To convince you of this, try to tell your customer that a particular defect is only cosmetic. If you do, you will find that the defect is not only real; you've just made the situation worse. Second, rejecting anything that can be seen more-or-less takes judgment out of the quality assessment process.

Having said all of that, most everyone should have tried roll-hardness measurements. They work on more products than any other class of instruments. They work on most paper grades with the exceptions of extremes in hardness such as tissue or board. They work on many film grades. They work on most laminates. By work, I mean that they have been statistically proven to correlate to a number of common defects. By work, I mean that they have proven themselves trustworthy enough to reject solely based on an excessive hardness variation across a roll. In the paper industry, it is the supplier that usually rejects. In the film industry, it's the customer that usually puts maximum hardness variations into a purchasing spec.

So why has roll-hardness variation proven so useful? Because many of supposed “wound roll” defects are really caused by winding of an excessively gage-varying web. Common examples here are baggy webs and corrugations. However, even roll defects that are not directly caused by gage variation are made more frequent by variation. The reason is that gage variation is often responsible for a greater variation of stresses in a wound roll than the nominal stresses caused by winding itself. So in effect, roll-hardness variation is really measuring gage uniformity that is either the root cause or a major contributor to most winding defects.

Available roll-hardness instruments include the Backtender's Friend, the Parotester, the Rhometer, the Schmidt Hammer and the TAPIO RQP. So which of these are best for you? What alternatives to roll hardness might you try if hardness is not appropriate? What causes wound-roll defects? Can they be predicted? All of those questions and many more are answered by Winding: Machines, Mechanics and Measurements. This book is a must-have for anyone with winding troubles. It's available from TAPPI PRESS and other book publishers and resellers. To paraphrase a Bell Laboratories saying, “A month in the plant can save you an hour in the library.”

920/725-7671, drroisum@aol.com, www.roisum.com