Omega style linear belt actuator

[zkling]

Long time reader first, time poster here. I usually just search through and I end up finding a great answer on these boards. Unfortunately I could not locate anything regarding what I believe is to be considered a "Omega drive" linear belt actuator. I am trying to design a system that will incorporate this style of belt driven actuator, however I can not seem to find much information pertaining to the design of the assembly. Basically looking to see if there are any rules or general "keep this in mind" while designing these Omega style drives.

I have talked to Gates (belt manufacture) and they were able to provide a few helpful tips, however I would like to know a little more, especially from those with experience in these drives. In case there is confusion as to what style drive I am referring to or if I have used the improper name please see the attached (crude) drawing.

Thanks in advance

 

[MikeHalloran]

It's almost, but not quite, a Rolamite. Patent expired long ago, but look it up for fun.
Tol-O-Matic could probably provide your assembly as a unit, though most of their stuff is topologically different, so it would be a custom job.

The worst problem I see with it is the reverse wrap on the idlers. You need to pick a timing belt pitch and construction, and go through your design with the fine print in the technical section of the catalog. Wrapping both ways reduces the life, and the thickness of the 'back', non-toothed side of the belt, is not usually controlled very well, so your belt tensioner has to take that into account.

The next most difficult problem is terminating the belt after you've cut it. You will probably try custom machined rack tooth clamps to secure the cut ends, but they typically don't have a good way to grip the fiber/cable tension member, so you are relying a _lot_ on the bond of the tension member and the teeth in a very small area at the anchored ends.

I have considered a similar drive, but it was simpler to just drive one of the end pulleys on an uncut belt loop, like Tol-O-Matic does.





Mike Halloran
Pembroke Pines, FL, USA

 

[tbuelna]

zkling,

I believe your drawing may be missing an important detail regarding the belt. The belt can only provide force in tension, so there would need to some form of compression member in the mechanism.

With toothed belts, it's critical to keep a sufficient number of teeth engaged on the sprocket to carry the PL tangential force. So position your idlers accordingly.

Here's an example of a belt-type rotary-to-linear actuator:

http://www.pt-usa.net/products/speedline.html

Hope that helps.
Terry

 

[JayMaechtlen]

The detail it's "missing" is that each cut end of the belt is probably secured to a stationary part of the machine, and the drive pulley and idler are on the moving element.

Jay Maechtlen

http://www.laserpubs.com/techcomm

 

[zkling]

Yes, I guess I left that detail out about how the belt would be tensioned, I just assumed that someone familiar with these drives would take that as unnecessary information. Attached is a page from PBC Linear's linear actuator catalog, depicting their omega style drives. That is correct, the belt will be tensioned at one end of the drive as well as clamped into place with as I would call a "rack clamp" that matches the tooth profile of the belt.

I can see that the closer the idler pulleys are together the more teeth engaged in the driven pulley, but is there an optimal distance between the two idlers?

Has anyone designed or utilized these drives before? Is it worth my time to design this over a accurate rack and pinion system?

Thanks,
Zkling

 

[MikeHalloran]

I'd be careful about words like 'high'.
Steel reinforced toothed belts, used as intended, are indeed capable of high tension, but in this actuator design, the tension is further limited by the tooth clamp design and the bond between the wire and the rubber.
Steel reinforced belts also don't care for fully reversed bending of the wire.
IOW, "chain pull" data for a normal toothed belt drive should not be used to estimate the belt tension capability of this actuator design.
I'd suggest you get a performance/life/cost quote from the guys who are already building this stuff, and a few exemplars that you can beat to death in a simulation of your application.


Mike Halloran
Pembroke Pines, FL, USA