- July 01, 2008, By Timothy J. Walker Contributing Editor
Most bearings are designed for a hard life, expecting to end up on a drive shaft of a high-speed motor with a heavy load. Some bearings, however, end up with a relatively cushy job in the relatively light-duty life as the bearing in an idler roller on a coating line or slitter/rewinder.
If you run an extremely high-tension web, like steel or an extra-thick paper product, you likely never think about roller drag, so stop here.
People look at me funny when I say you should be able to put in an idler roller and never replace the bearing. To convince yourself of this, get out your favorite engineering handbook and calculate the life expectancy of your bearing based on process rpms and the load from roller weight and web tension.
You likely will find an expected life of 10 or even 100 years. These bearing life expectancies may seem unreasonable, but they are achievable if you consider a few real-world factors. Tend to these five factors and you too may see bearings lasting a decade or more.
What is the right size idler bearing? The easy answer is the bearing needs to have an inner diameter bigger than your shaft and outer diameter smaller than the roller's shell. On the plus side, larger idler shafts will have less deflection, but larger bearings will have longer expected life but also more friction, especially in greased, contact seal bearings.
Some idlers eliminate the need for large inner diameters by eliminating the through shaft. A small diameter stub shaft can reduce bearing diameter greatly, but the tradeoff is life. I think ¾ in. is the place to start on bearing inner diameter unless your idlers are running under extremely light loads.
From a low-drag point of view, no lubricant is best, but a bearing with no lubricant will fail quickly. The answer is to use either of the following options.
- Use 5%-25% fill with a light oil lubricant.
- Avoid petroleum-based lubricant altogether and opt for a PTFE-based, low-viscosity lubricant.
For low-tension web handling, just say no to Zerk fittings on idler rollers. A Zerk fitting doesn't mean a roller will have too much thick grease inside, but I've never met a Zerk fitting that doesn't eventually have a date with the pump-pump-pump of a grease gun.
Contact seals add to idler roller friction, yet with no seals, dust and debris will get in and destroy a bearing. Labyrinth seals, where any particle must travel a tortuous route to get to the balls in the bearing, are the preferred answer. Some low-drag bearings use a lubricated felt seal to block contamination.
The top environmental questions for bearings are heat, humidity (or lack thereof), ozone (near corona treaters), and contamination (such as near slitting). Sufficient lubrication, non-petroleum lubricants, and labyrinth seals are the answer.
Bearings are good at supporting loads perpendicular to their axis of rotation, but the real world may exert loads in other directions. Undesired side loads may result from an over-muscled roller assembly or from a shell's thermal expansion. Excessive radial loads may come from thermal expansion differences between high thermal expansion of aluminum shells and relative-to-low-expanding steel bearings.
Twisting loads are generated from misalignment of the bearing outer and inner race, usually from overly small shafts. A bearing in a spherical bushing mount can minimize this, or you can use basic engineering to model the shaft and shell deflection and keep the bearing angle mismatch to a minimum (usually less than ¼ deg).
I would like to thank Cal Couillard of Componex (www.componex.net) and Pete Eggen of Webex (www.webexinc.com) for their valuable discussion in preparing for this column. I expect, like me, either of them would welcome a call from you to discuss your unique idler roller bearing challenges.
Web handling expert Tim Walker, president of TJWalker+Assoc., has 20+ years of experience in web processes, education, development, and production problem solving. Contact him at 651-686-5400; email@example.com; www.webhandling.com.