I’ve been contemplating a West Coast Drive for a while now. And just recently a CNC mill, CNC lathe and a plasma cutter came into our team’s possession, so we’re no longer limited to sheet metal. I’ve used the WCD SS Shifter and bearing blocks from WCproducts. Current specs are:
Dimensions - 28"x27.5"
Weight - 28.9lbs
High gear - 16fps
Low gear - 6fps
Belt driven using 24t GT2 profile pulleys
4" AM performance wheels
Bearing Blocks slide 1/2" to adjust tension, and use CAM tensioners.
My concern at the moment is that the pulley walls will bend during use and allow the belts to slide out of alignment. The reason i say this is that the pulleys are made from pulley stock and two 1/16" walls. These walls are not directly attached to the pulley stock and are held in place with a spacer an retaining ring as shown: http://i.imgur.com/cFGnGK1.jpg
I’m not exactly sure but i think thats how we did our pulleys this year and how no problems with the walls bending on the pulleys. They were great and we didn’t have a single complaint about them.
Any teams have experience using typical WCD cams with a belt drive? We’re finalizing a similar off-season WCD, and we’re concerned that the increments on a typical WCD cam are too large to properly tension a belt drive. Thoughts?
You don’t need a tensioning system at all to use a belt drive. If anything, designing to exact center distances with a belt drive is better as it eliminates any chance of the belt run pairs being tensioned unevenly.
There’s no reason that the belt pulley walls would bend, regardless of how they’re attached to the assembly.
Mike, I’d agree the click CAM tensioners are too coarse. Since the distance is so small, I bet a smooth CAM without any clicks could be used just fine, and with a much smaller ratio during rotation. This shallower CAM would be less likely to loosen from load as well.
Exact C-C actually guarantees all the belt runs will be at uneven tension.
When you do exact C-C the belt tension is set by the c-c distance plus the sum of all your manufacturing tolerances (which includes the belt! and that can be more than a thou on drive length belts!).
Gates has a lot of great documentation out there that explains this and many other factors with their belts. It’s a great read for any team who runs belts.
That makes sense. I was mentally comparing my team’s exact-center set up to our previous belt drive, which used plastic tensioners with a lot of adjustability. In this case, we could easily set up our drivetrain where one belt is much more tense than the other, causing ratcheting. I do think that a WCD bearing block setup, relying on a small cam that’s easy to make fine adjustment with, could result in more optimal tension than our drive. I don’t know how necessary that is, but I’m sure it would help with efficiency.
It’s worth noting that for an exact center belt drive made from a single piece of tubing, the spacing tolerance is just determined by how precise your mill is. For many teams this precision is really “good enough” and better than they could get with a tensioning system.
Gates’s documentation on belts is very useful, and something that should be read by anyone designing a belt drive, but at times the guide’s advice and specs are quite conservative. The Gates manual recommends never using an exact-center belt drive, but for FRC drivetrain purposes (at most a few hundred hours run time on a well used practice robot) the factors they cite that suggest a tensioning system don’t really apply.
From what I remember from the documentation isn’t that they say you should never do it, but rather if you do it you should recognize you must apply a strength derating to the belt to compensate for you not having perfect tension. In reality for all FRC applications this derating should exist as even with tensioners teams aren’t able to measure tension to the levels required to determine it’s proper.
I only make the point for teams that are looking at a pulley setup that is right at the ragged edge. Doing exact c-c in that case will be more likely to fail than a tensioner (if the tensioner is used to really dial in tension right).
I’d rather move away from the ragged edge and just skip the tensioner, less stuff to work on.
Separately, you do raise the correct point that most of their documentation is tailored for continuous operation in factories, etc. and as such is a fair bit conservative.
Based on the comments so far, which have been extremely helpful, I’ve made some changes. I got rid of the CAM tensioners and sliding bearing blocks. It’s now a C-C +.005". I’ve also re-done the bellypan, added a battery holder and added a more universal hole pattern on the frame. As shown it weighs 28lbs.
