So we are creating a conveyor/elevator system for the first time (second year team). We have settled on the 2" wide urethane belts (the orange McMasterCarr) with 1 1/2" between them (7 total belts). Due to spacing in the robot it will ride over 3 rollers all about 1" in diameter turning around 3600RPM (to be a bit faster than our 10fps robot speed). This will all be driven by a miniCIM. We have 3D printed crowned pulleys (Diameter+1/8" crown) for one roller. Here are the questions:
Should we crown all the rollers? Just 1? Does it matter which?
Should we reduce speed of the intake as the robot slows down proportionately?
What kind of compression would you suggest with our ~5" ball?
How much undersize is standard for cutting the belts? (I’ve heard 5-10% shorter than calculated length. Does rotational speed or other factors affect this?)
I’d crown all the rollers; it’ll help more than crowning just 1.
No need to reduce intake speed, it’ll only work better going faster than the robot in most cases. Just keep it at full speed.
The exact compression of your belt system is something you have to experimentally determine every year and it depends entirely on your setup. In the future do this during prototyping. For now, 4.75 inches from the belt surface to the back wall is probably a good place to start? Just a guess.
For flat belts you don’t want to cut them THAT much undersize; they stretch much less than the round kind which that standard was invented for. Maybe more like 3%, if that. If you clamp the belts onto your rollers and keep all sides in proper tension, you CAN weld them in place, though it’s a bit of a pain. If you do this, weld them when they are off the crown and then slide them on for a bit of extra tension.
Anyone care to elaborate on the comment about crowning all rollers versus crowning one? I’ve heard both ways but lack direct experience with belted systems in FRC.
I don’t see it as that complicated. Belts track well with crowns but not perfectly / aren’t entirely immune to falling off track. So, it’s much easier for them to stay tracked if they are crowned on every roller versus just one. Why would this not be the case?
Most industrial systems I’ve seen use a flat/crowned setup, but it isn’t that it is technically better, it might just be that one is sufficient (when you don’t have much side load) and it might just be cheaper than 2 crowned ones. In the case of FRC with 3D printed rollers, the cost is no longer an issue and the “standard” might be irrelevant.
In industry I have yet to see belts get pushed sideways like they do in FRC when used in a compressed-ball intake, so the extra crown is not needed. It is super useful to have two crowns in FRC because we can side-load belts quite dramatically and the extra crown prevents the belt from climbing off.
However, using two crowned pulleys it is quite possible to over-constrain the belt and cause binding/drag issues. Usually this isn’t a problem in FRC because it is easy to align two pulleys within their side-to-side positional tolerances.