A laminate can delaminate if the adhesive or adhesion between the plies lets loose. This may result in a defect variously labeled as tunneling, gaping, worming, and so on. The defect is readily recognized as a separation of the plies in a narrow area that is oriented in the CD (cross direction). Typically, the delamination will span much of the width and often be spaced periodically in the MD (machine direction). Finally, the defect will not appear until some tension is released from the laminate, such as on the tail of a wound roll. Even then, however, the defect may take time to develop—seconds to hours, depending on the product.

It is tempting to look at the release of ply bond and blame the adhesive. Certainly, if the bond were better the plies would not separate. While the problem is far more frequent with adhesives that cure slowly, it can happen on others as well. It is not merely a problem of "green tack." We will leave it to others to determine whether the failure is one of adhesion, cohesion or shear. We will leave it to others to determine whether the surface preparation and adhesive chemistries are optimum. Finally, we will leave it to others to determine whether the adhesive application has been optimized. What we will show here is that the failure is also one of product and process design.

If you look closely at the tunnel by sectioning it across, you will note the gap is D-shaped. In other words, one ply is tight and the other is buckled. Stated another way, the straight ply is tighter and/or shorter than the buckled ply. Also, careful observation will show the product always curls toward the straight ply. However, the amount of curl will decrease slightly as the tunnels develop after web tension is reduced.

What these key observations are telling us is that tunneling is just a form of curl. The only difference between more conventional curl and tunneling is the added ingredient of adhesive release. Indeed, you can often see a freshly cut web move from one form of curl (conventional) to the other (tunneled) if the release is slow enough.

This gives us an additional set of tools to combat the tunneling problem. That is, like any other form of curl, we need to more closely balance the MD strains caused by MD tension. Briefly, we would consider the following:

• Educate product/process developers, salesman, customers and operators on the mechanics of curl
• Avoid designing/selling products that have widely dissimilar moduli (and thermal coefficients)
• Decrease the tension of the straight, usually lower modulus, ply
• Increase the tension of the buckled, usually higher modulus, ply
• Heat/moisturize the buckled ply
• Cool/dry the straight ply
• Use differential torque on a dual drive nip to further adjust strains.

Note that all of these treatments must take place just prior to the laminating nip. After you reach the middle of the nip, your fate has literally been sealed. You only need to wait to observe the outcome. One final diagnostic is to note that if tunneling occurs, it should do so uniformly across the entire width. If the tunnel favors a lane or edges, for example, we know the product and/or process are not uniform. In some cases, one of the plies may have baggy lanes. If tunneling is not uniform, you may need to address its cause at least as aggressively as trying to balance the strains in the two plies.

David Roisum, Ph.D.

Consulting Technical Editor

920/725-7671

DRroisum@aol.com

www.roisum.com