Do You Need a Roller That Is Chillin' and Groovy?

Published: April 01, 2014

How can you help a big chilled or heated roller that air lubricates, losing heat transfer capability at high speed/tension conditions?

Nip it. Ugh. I can’t believe I could even say that. Most people who have worked with me know of my phobia of nipped rollers. However, when you want to get the air out (e.g., high speed film winding), you can’t beat a nipped roller.

You know why I hate nips? Nips make more wrinkles. So can we solve the air lubrication problem without nipping? Sure. Groove the roller. That’s what I’d recommend for any idler roller. Is there any reason to hesitate in grooving a nipped roller? I'm mostly a web handling guy, but I know something about heat transfer, too.

Debate on Grooving a Chilled Roller

The debate on grooving a chilled roller has a couple of competing interests.

No doubt, regarding heat transfer, conduction beats convection. Air between the web and roller can greatly reduce your heat transfer coefficient. I can imagine if you are flying over a roller with large radius and high speed/tension ratios, then you are nearly fully insulated. If the contact time is long enough, the trapped air layer may stay at or near the roller temperature and stay there even after the web departs, helping keep even convection heat transfer somewhat functional.

The best heat transfer is conduction; without intimate contact, I don't think you get it. Grooves would help and hurt conduction. If there was no air layer and you made a roller with 20% area grooves (10 mil wide grooves on a 50 mil pitch), then you would lose 20% of your conduction contact. However, if you have a big air layer, and the grooves pull you down to the roller, then 80% contact is way better than 0%. That's the balance. Will you gain or lose by grooving?

I can imagine someone smarter than I could calculate this by combing air lubrication, groove design, and heat transfer models, but I have a simple experiment in mind that might prove the potential benefit in a real system.

Use an IR pyrometer to measure the temperature of your web before and after your heated/chilled roller. Repeat this measurement over a range of speeds. Plot the temperature change versus contact time. Contact time will be L/V or Radius x (Wrap in radians) / Speed. With no air insulation, the change in temperature should be an inverse function with speed (linear with time of contact, ignoring the cooling effects of the non-wrap roller convection). If the temperature change degrades faster than the inverse of speed (i.e, faster than the decreasing contact time), then you have a secondary factor–air insulation degrading your heat conduction coefficient.

To calculate grooves (for this or any application), consider trying the free air lubrication calculator at Prof. Stephen Abbott’s website. The Air Lubrication Calculator (ALC) will calculate the height of an air lubrication layer, estimate coefficient of friction loss, and recommended groove width, pitch, and depth. You may want to reduce the groove width to pitch ratio to keep your contact area up.

Look over the help page for some good advice on grooves.

Some other comments:

• Too wide of grooves can catch or buckle thin webs (that’s why the ALC recommended groove width < 10X web thickness).
• Grooves can be hard to clean if they get slimed.
• I think the calculation of groove cross-sectional area should equal lubrication height over estimates of the groove depth you need.

Okay. Now our roller is both chilling and grooving. Peace, man.