A sheet which is not dead flat as a sheet of glass is wrinkled, baggy and/or curled. The web can curl in two directions (up or down) and in three different orientations (MD, CD and diagonal). Some webs may be troubled by more than one type of curl at a time, though one is usually dominant. To complicate things further, curl can vary across the width, with time, grade and process settings. MD curl is common with lamination and one-sided treatments, while CD edge curl may trouble other processes. Curl can be measured in several ways. My favorite is to make an X-cut on the web and note which direction the center points move.

A similar method is to lay a square sample flat on the table, measure the lift of the corners, then flip the sheet and repeat the measurement on the other side. Some people hang a square specimen and use a radius template. Others clamp a strip on one end and measure the curvature of the other end with a protractor. In all cases, extreme care in handling of the web is vital for consistent test results. In some cases, such as paper, the sample must be conditioned before measurement.

There are many ways to get curl. One is the strain mismatch of the plies that is the bane of laminators everywhere. This can be due to product design (dissimilar mechanical properties of the plies) or poor tension control (not lowering tension sufficiently on the most flexible member). Similar physics are present with one-sided coatings, moisture or heat application. Indeed, there may be as many ways to introduce residual stresses, such as bagginess or curl, as there are ways to make webs.

Another source of curl is environmental. If moisture or temperature changes on a one-sided web, the most expansive side will warp the product into a curl. An example is a paper/poly product that will curl toward the paper in Tucson and away from the paper in Tallahassee. Even if the paper is sized or sealed, moisture from the air will eventually get in. Even if you hold the web flat during these environmental changes, some curl will still spring back when the sheet is released. Similar physics occur when extrusion-coating a hot web onto a cool one. Product developers must be quite careful about any through-thickness mismatches in mechanical properties such as modulus, Poisson ratios, and thermal and hygroscopic expansion coefficients.

Finally, roll-set curl is common on many wound products. Roll-set curl is easy to recognize because it curls in the direction of the wind. Also, the magnitude of the curl is the inverse of radial distance in the wound roll. This means there will be about 10 times as much curl above the 4-in. OD core as at the 40-in. outside of the same wound roll. The factors that reduce roll-set curl are caliper (decrease), core diameter (increase), winding tightness (decrease), time in roll (decrease) and grade. Web caliper and core diameter are much stronger factors than the others.

Because the two biggest factors in roll-set curl are customer determined, the converter is in a somewhat difficult position. If they sell the product, it will have this "defect." If they don't sell it, they will lose the customer. If they try to sell the product on bigger cores, they will be suspected of self-interest and will be compared with other suppliers who are willing to use small cores.

It would be comical, if it were not so common, to see groups struggling with curl issues. Product developers, process developers, operators, salesman and customers all see the problem as not being of their own making, when in fact, it is. It's simple physics.

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