I'm finally ready to put my Polychain GT 8mm pulley stock to good use, and I want to do things right. I understand that alignment is critical, but there's some flex in any system, so how do I keep the belt from walking off the edge of one of the pulleys?
This typically only works if you're transferring torque in a single direction, aka no regen.
If you want regen, gotta just get the sprocket center to center distance right and pre-load extra belt tension if it's a flexy thing. If it skips or hops teeth, gotta get tension tighter or frame stiffer. They somehow tend to never walk off the sprocket if the sprocket shafts are actually parallel and stay parallel under loading.
You cannot allow any flex in a belt drive train.
You will have to ensure all shafts are suitably rigid with solidly braced supports And with a very rigid tensioner mounting if you use one .
A flange on at least one of the pulleys on both sides of the belt.
If alignment/flex is questionable the flanges will wear the belt & reduce efficiency.
Without flanges the belt will walk off (unless you somehow make at least one pulley slightly barrel shaped.....)
IME of course - LPF above seems to have been luckier
Flanges and anti-jump bars that skim the back of the belt over the pulley are all used to some degree in some commercial/agricultural applications but are arguably bodges that let flawed design work, while the real answer is to have both the pulleys aligned and the belt tensioned correctly. As LFP alludes to, the challenge is ensuring these conditions are met during dynamic load, not just static setup.
Thanks guys,
I imagine that it will be hard to identify flex under load. While I'll try to go overboard regarding rigidity and alignment, would it be prudent to put say a tapered flange on one side of one pulley and the other side of the other pulley?...or am I over-analyzing it and belts stay on well with proper alignment and structure?
Just look at how these kinds of belts are used in cars. No reason to re-invent the wheel. An inch of flex or so, after tightening, depending on the overall length of the belt is all you need. Spring tensioners are not required, but may help. Alignment is critical or the belt will slip off any pulleys regardless of debts of grooves.
This link from Gates has some decent info and guidelines for tracking synchronous belts: http://www.gatesmectrol.com/mectrol/brochure.cfm?brochure=5196
I'm more used to looking at ribbed belts which track themselves quite nicely. About 3x times lower power transmission than a synchronous belt of the same width, though.
This link from Gates has some decent info and guidelines for tracking synchronous belts: http://www.gatesmectrol.com/mectrol/brochure.cfm?brochure=5196
I'm more used to looking at ribbed belts which track themselves quite nicely. About 3x times lower power transmission than a synchronous belt of the same width, though.
Ribbed belts tend to have significantly higher friction losses.
I've broken belts a few times in a few extreme abuse cases, but I've actually never personally had a synchronous tooth belt fall off. A bit of side guide plate on the sprockets with a rigid swingarm and you're good to go.
Oh, yes Just by manufacturer's data we go by ~95% efficiency for a individual-Vee or multi-rib belts versus 98-99% for an equivalent synchronous belt. Industry seems adverse to synchronous belts for non-timing applications, though, presumably because their failure-mode when over-loaded is often to immediately snap/shred rather than slip...
