As we are starting to test all of our systems together, we have run into some problems with our intake system. A picture is attached below, but to explain it briefly we are using 1” wide mcmaster polyurethane orange belt in addition to vectored intake wheels on both sides. We have been using 2.5”x1.25” colsons as pulleys because of their crown, but this has not been enough to withstand the side loads created by vectoring in the balls from the side. We tried to attach some walls made of flexible traffic cone material, but they seemed to just exacerbate the problem by allowing the belts to walk up the sides when they take any side load.
Our next solution is replacing the colsons with some 3D Printed flat pulleys that have walls, essentially smooth HTD pulleys, and though we are implementing this change today, I am worried we will run into the same issue with the belts riding up the walls. Is it reasonable to even continue move forward trying to use the vectored wheels into the belts or should we drop that and make it a straight intake? We are reasonably confident that the belts deal with the straight loads fine since we are also using them for indexing and we have had no issues once they are already in line and constrained. Thanks for any help
The belts will probably climb any pulley with walls. Have you tried increasing the pitch of the crown by adding electrical tape in the middle of the colson tread?
We had some success with a similar setup, but we added an additional shaft behind the shaft with the vectored intakes and that is where the flat belt is attached. We then had a timing belt that attached it to the vectored intake wheel. This allowed the ball to become centered before it touched the flat belt. This reduced our belt falling off significantly, but we still had a few issues.
I think your problem could be that you aren’t using crowned rollers (rollers that have a bump in the middle that the bands follows). We have a fair bit of tension on our bands and are using rollers like the ones in the picture to center to bands.
We actually were using crowned rollers at first, but I would guess that the rollers are not wide enough/do not have a large enough crown relative to the belt width. We might try 3D printing a model like that instead of one with walls.
We fought similar issues in 2016 and weren’t able to completely solve it with crowned rollers or other solutions proposed in this thread. The root cause is that you are bringing game pieces in perpendicular to your belts and those forces are relatively large, compared to the centering ability of the crowned rollers.
An alternate solution that you might try is replacing your front pulleys with pulleys of smaller diameter so that the belts don’t contact the ball there. In addition, add small omni wheels to your front roller so the balls get centered and get pulled onto the belts on the correct vector.
could a setup like this work where the front flat belt just drives the front most roller by grabbing tape wrapped around it but never contacts the ball? that way the vectored wheels will center it before it contacts those back belts like you said.
As everyone else has said - make crowns, not walls. They don’t even have to be particularly substantial crowns - they can be a piece of string taped to your roller or pulley. Belts will find their way to any high point on a circle - crown, flange, or otherwise. Flanges are for timing belts.
Our first setup looked a lot like your sketch and we slipped that center front belt almost all the time until we just took the flat belting off of the front shaft.
FRC 319 has printed our own crown pulleys for our 2" flat belt. We have both passive and active pulleys on our 2020 bot. Check them out in our Public Onshape.
We have done the crowned rollers like the one shown in @JimboFee picture but this year we chose to go with these colsons from WCP: https://www.wcproducts.com/217-4045 They work amazing for keeping the poly belting in place. They have a nice crown on them. The biggest problem that we found with them is that the hex shaft is super tight but we ran a broach through them and they were fine afterwards. Also bonus, they are under $5 so they dont count towards your BOM.
after our meeting today we got to this point shown below, obviously the overall intake width is decreased now but it works much better than what we had before. the greens are 3” stealth wheels and the orange strip is the flat belt that brings the balls into our conveyor system. we chained from the front to back and catching the middle shaft in the middle to drive the two shafts of wheels. by using wheels instead of belts on the front two shafts we basically ensured any side loads were resolved before the ball reached the first flat belt. thanks for all the helpful responses!
