How Much Misalignment Is Trouble?
- Published: January 01, 2009, By By Timothy J. Walker Contributing Editor
Most converting equipment suppliers and manufacturers take care to ensure the rollers of their machines are aligned. Designing for good alignment, machining and assembling to close tolerances, measuring to ensure alignment, and periodically rechecking roller parallelism all cost money, so we ought to think about what we are getting for our money.
Some products gain more from precise alignment, while other products quite likely are insensitive to misalignment. How do we know what the alignment specification goal should be for a specific product?
My general rule of thumb on converting equipment alignment calls for targeting a parallelism of 2 mils/ft. This converts to about 0.2 mils/in. or mm/m, 0.01 degrees, or 0.7 seconds of a degree.
It's a small angle, usually imperceptible to the naked eye. You will find equipment aligned to this specification is achievable, measurable, and will eliminate unwanted roller-induced wrinkles, lateral shifting, and edge flutter in most webs.
What problems will occur if roller alignment is only accurate to 10 mils/ft? Three big problems result from roller misalignment:
- Lateral shifting;
- Loose edges;
- Shear wrinkles.
In my previous column of April 2004, “When Rollers Fight, Webs Lose,” I provided more details on the definitions of level and tram and gave some examples of lateral shifts from misalignment. (Find this column on the web at www.pffc-online.com/mag/paper_rollers_fight_webs/index.html).
In short spans, the tracking from roller misalignment will be as imperceptible as the roller misalignment. In long spans, a misalignment of 10 mils/ft will shift by 7-10 mils/ft of web span length. In a 10-ft span, the web will shift 70-100 mils, a noticeable but not earth-shattering amount.
Whether a roller misalignment causes the short side of the web span to go slack is dependent on the strain or stretch in the web. To drive an edge of the web into slackness, the misalignment of two rollers in a web span needs to be two times the stretch in the web.
In this case, the tension in the centerline of the web will not change, but one edge will lose all its tension and the other edge will stretch to double the average tension and strain. Since each span has two rollers that potentially can be misaligned, the critical alignment of any one roller is equal to the strain of the web.
Low modulus materials, such as most nonwoven and softer films, will stretch more that 1% under web tension, so a roller needs greater misalignment to relax this elongation out of the web to a point where looseness would occur. A 1% misalignment in a 2-ft span is about 0.25-in. or 50 mils/ft for a 60-in. web width.
Higher modulus materials (such as polyester films, many papers, and all foils) will have a much lower tolerance for how much misalignment causes edge looseness and associated problems of out-of-plane sagging or flutter. A 2-ft span of 1-mil polyester tensioned at 1 PLI (lbs/in. of width) is stretched about 0.2% or 100 mils. In longer spans, it will take more misalignment to create a slack edge.
This absolute misalignment to induce a slack edge is independent of width, but viewed on a mils per foot basis, wider products are more sensitive. As web width reaches 50-100 in., the 100-mil error converts to 25 or 10 mils/ft.
In aluminum foil, elongation from tension is lower yet, 0.01%. A 2-ft span needs only 5 mils of misalignment to loosen an edge, just 1 mil/ft for a 60-in. wide product. Unless you are running aluminum foil or extremely wide, it looks like 10 mils/ft should be just fine for many processes.
Next month we'll continue to drill down into the complexity of what is an unacceptable roller misalignment. We'll provide more real numbers of how much misalignment will wrinkle a web based on stiffness, width, and span lengths.
Web handling expert Tim Walker, president of TJWalker+Assoc., has 25 years of experience in web processes, education, development, and production problem solving. Contact him at 651-686-5400; This email address is being protected from spambots. You need JavaScript enabled to view it.; www.webhandling.com.