Whats the big deal with in tube gearing?

[Joey Milia]

Originally Posted by Dunngeon

I'm more concerned about the chance of snapping a drive belt, rather than the wear. It's nearly impossible to snap chain in FRC if assembled correctly, from my (limited) experience it's quite a bit easier with belts.

15mm wide GT2 belt is pretty much indestructible for FRC purposes.

EDIT: 5mm pitch for clarification

[Dunngeon]

Originally Posted by Joey Milia

15mm wide GT2 belt is pretty much indestructible for FRC purposes.

Good to know

[cadandcookies]

Originally Posted by Dunngeon

I'm more concerned about the chance of snapping a drive belt, rather than the wear. It's nearly impossible to snap chain in FRC if assembled correctly, from my (limited) experience it's quite a bit easier with belts.

I see. I'd second Joey-- 15mm GT2 or HTD belt is in my experience on the same level as chain in terms of breaking. In my experience if you can get to 36t pulleys, 9mm HTD or GT2 is also pretty hard to break (as in, I've never had an issue with it).

[Dunngeon]

Originally Posted by cadandcookies

I see. I'd second Joey-- 15mm GT2 or HTD belt is in my experience on the same level as chain in terms of breaking. In my experience if you can get to 36t pulleys, 9mm HTD or GT2 is also pretty hard to break (as in, I've never had an issue with it).

It wouldn't surprise me if the teams I've seen snap belts were using 9mm belts incorrectly, some teams in the PNW have a bit of a history making odd decisions.

[bobcroucher]

Chain definitely stretches through wear, mostly the pins. Park Tools sells a chain stretch gauge for bike shops. The poor man's way is to measure 24 links at 0.5 inches each totals 12 inches when new. Once the chain reaches 12.25", it's time to buy a new chain. The stretched chain will wear out the aluminum chain rings and the rear cassette, and each costs more than a new chain. A new chain lasts 1000 - 2000 miles depending on cleanliness and lubrication.

Bike wheels are obviously much larger, so the chain life will be reduced accordingly. If we scaled 1500 miles down by 4 / 27, since 4 inches is a typical WCD wheel size, and 27 inches is a typical road bike diameter, you would get about 225 miles. That's still around 11,000 round trips down and back on the 52' field. The chain is quite unlikely to wear out based on revolutions alone. Work done to move each vehicle's mass is at least within a factor of 2, and would benefit the robot.

Robots don't get much road grime, and carpet fuzz isn't very abrasive, but they usually don't get much lubricant. The more lube you use, the more stuff sticks to the chain. By putting the chain in the tube, you can probably lubricate it a bit better without fear of contamination. I'm sure chains will typically last an entire season regardless. Putting the chain inside the frame tube probably helps protect it from game pieces and other robots as well.

According to Gates, a belt's stretch is nearly un-measurable over the lifetime of the belt.

[Chris is me]

Originally Posted by Dunngeon

It's possible, but you'd have to use 2x2 tubing to get both of the belts to fit into the interior. From what I've read on Delphi, chain in tube is better than belt since chains (rarely) snap and don't stretch over time, eliminating the need for a tensioning system.

You have it backwards - chain *does* stretch over time, and belts do not. Chains generally do need a tensioning system, and belts generally do not. My team has run 15mm belts in tube with exact center distances for now four seasons in a row without ever having a belt failure or need to tension. Belts just don't stretch in the liftetime of an FRC robot.

If you undersize your belt drive, it can break - generally failures I've seen involve 9mm wide belts with 24t or smaller pulleys. This just isn't up to snuff. 15mm wide is a safe bet if you're using small, west coast drive friendly pulleys. If you are using larger tooth count pulleys such as the bolt on VersaPulleys that are 36 or 48 tooth, 9mm is generally fine for those.

The reason "chain in tube" drivetrains with no tensioning work is because the chain and sprockets are sized so that they just barely fit in the tube at all. Chain in tube does stretch, but since the chain on the sprocket is such a close fit with the tube, the chain can't jump the sprocket since there's just nowhere for the chain to go. It works well, but it is very much an edge case with chain drives.

[Dunngeon]

Originally Posted by Chris is me

You have it backwards - chain *does* stretch over time, and belts do not. Chains generally do need a tensioning system, and belts generally do not. My team has run 15mm belts in tube with exact center distances for now four seasons in a row without ever having a belt failure or need to tension. Belts just don't stretch in the liftetime of an FRC robot.

If you undersize your belt drive, it can break - generally failures I've seen involve 9mm wide belts with 24t or smaller pulleys. This just isn't up to snuff. 15mm wide is a safe bet if you're using small, west coast drive friendly pulleys. If you are using larger tooth count pulleys such as the bolt on VersaPulleys that are 36 or 48 tooth, 9mm is generally fine for those.

The reason "chain in tube" drivetrains with no tensioning work is because the chain and sprockets are sized so that they just barely fit in the tube at all. Chain in tube does stretch, but since the chain on the sprocket is such a close fit with the tube, the chain can't jump the sprocket since there's just nowhere for the chain to go. It works well, but it is very much an edge case with chain drives.

My concerns with belt were more geared towards snapping, not the stretch of the belt. As Joey and others pointed out, the 15mm GT2 belts are pretty indestructible, as are 9mm when utilized correctly. I had heard that belts stretch second hand, but its pretty clear now that they don't stretch enough to be noticeable.

In contrast, we've run chain (bearing in tube) with no tensioners the last 3 years and never had an issue. It does wear the sprockets a bit, but that's only at the beginning.

[Rangel(kf7fdb)]

Agreed with most in terms of experience with belts. One problem we have had in the past that teams should look out for is tape from the field getting stuck between the belts and the pulleys. We went most of champs in 2013 without ever realizing it because there were no obvious indicator just looking at it(Somehow white tape tricked us). It wasn't until the tape started to really affect the driving that we realized something was wrong and found a lot of tape stuck to the pulleys after taking off the belts.

[connor.worley]

Originally Posted by Dunngeon

In contrast, we've run chain (bearing in tube) with no tensioners the last 3 years and never had an issue. It does wear the sprockets a bit, but that's only at the beginning.

#25 or #35? I'd be surprised if you've never thrown 25 in 3 years doing that.

[Joey Milia]

Originally Posted by Rangel(kf7fdb)

Agreed with most in terms of experience with belts. One problem we have had in the past that teams should look out for is tape from the field getting stuck between the belts and the pulleys. We went most of champs in 2013 without ever realizing it because there were no obvious indicator just looking at it(Somehow white tape tricked us). It wasn't until the tape started to really affect the driving that we realized something was wrong and found a lot of tape stuck to the pulleys after taking off the belts.

This is one reason belt in tubes are good. 192's tube is designed to cap all the main access holes so it's almost completely sealed. The largest open holes are the 3/8in access holes for gearbox mounting bolts.

[Gregor]

Originally Posted by connor.worley

#25 or #35? I'd be surprised if you've never thrown 25 in 3 years doing that.

I've heard various teams running 25 chain with no tensioning. It's not as uncommon as people may think, but you do need to do it correctly.

[Andrey K.]

Originally Posted by connor.worley

#25 or #35? I'd be surprised if you've never thrown 25 in 3 years doing that.

#35. We've never bothered with tensioners because the slop #35 chain allows for. Though we're definitely looking to go to #25 for the weight reduction.

[Greg Woelki]

Originally Posted by Joey Milia

This is one reason belt in tubes are good. 192's tube is designed to cap all the main access holes so it's almost completely sealed. The largest open holes are the 3/8in access holes for gearbox mounting bolts.

I've seen pictures of several of your drive transmissions, but is there anywhere I could find a good image of your belt-in-tube drive train?

[Dunngeon]

Originally Posted by connor.worley

#25 or #35? I'd be surprised if you've never thrown 25 in 3 years doing that.

As Andrey said, #35, the only reason we haven't been running #25 is down to our CNC not being accurate enough (for an unknown reason) to place the three bearing holes correctly.

We've run #35 the last 2 years exclusively as the mill has gotten worse, and used #25 no tensioners for pre-season 2013 and 2012. It's possible to do, and have very little issues, just a little beyond our machining capabilities at the moment. Chain in tube should allow us to do 25 even with our mill issues, which will have nice returns all around.

[Joey Milia]

Originally Posted by Greg Woelki

I've seen pictures of several of your drive transmissions, but is there anywhere I could find a good image of your belt-in-tube drive train?

I've talked about 192's system in several places but I don't think there are that many good pictures online.

I made this post with a picture of the bare frame. It's not the best angle to see the sides though.
http://www.chiefdelphi.com/forums/sh...4&postcount=63

I also found this photo of the 2015 robot with a good view of the outside.
http://www.leilehuarobotics.com/uplo...11496_orig.jpg

Neither of those give a good look at whats going on with the belts on the inside so with the increased interest recently I am thinking about making a detailed post on them in the nearish future.