My team is looking at maybe trying to do a chain in tube drivetrain and were wondering what experiences other teams have had. The main questions are basically how do the real world experiences compare to the “on-paper” factors? How hard is it to service? how often do things break? we will run tensioners, so we wont worry too much about C-C distances. We have one designed and it uses COTS VEX 16t #25 sprockets and the Versablock kit. We are interested to hear from teams who have run this system in competition or even just testing outside of competition, and whether or not it was worth it and if you will or wont go back and why.
2 Likes
5987 made a chain-in-tube WCD as our first custom chassis last summer. We used 60x40x2.5mm tubing for the drive rails, and double 12t #35 sprockets, which gave us about 0.4" of clearance (some of which was taken up by the ~1/8" drop center). Assembly went very smoothly; each drive rail took about half an hour to put together. We put the off-season chassis through its paces, and we were satisfied. We haven’t had a single problem with it since, despite being abused in driver practice.
Our 2018 robot was also based on a chain-in-tube drivetrain: the same setup as the off-season chassis but without a drop center. We didn’t have any problems with it on the competition or practice robots. The original design featured an upper rail that bolted to the drivetrain rail, onto which all of the mechanisms were supposed to be attached so the drivetrain rail could be easily removed for servicing. In the end though the robot design required some mechanisms be attached to the drivetrain rail as well. Luckily, the drivetrain worked flawlessly all season so there was no need to remove it. In all likelihood we will probably use a similar chassis for 2019 unless the game precludes it.
To address your questions, serviceability can be an issue because the chain and sprockets are hard to access. Designing the drive rails to be removable, machining pockets into them for better accessibility, and designing the whole thing robustly to avoid the need for repairs are all good steps you can take to make servicing the drivetrain a bit easier.
I don’t recommend using tensioners with chain-in-tube. One of the major benefits of the chain-in-tube style is that the tube constrains the chain so that it can’t skip over the sprocket (there physically isn’t enough room for the chain to rise over the sprocket teeth without hitting the tube inner walls). Use a calculated C-C distance plus a wear-in fudge factor and you shouldn’t have any problems for the duration of a season.
If you have access to a mill, router, tube laser, or a local company or FRC team with any of those machines, I would recommend drilling the holes directly into the tube instead of using VersaBlocks. Even if you have an accurate-enough drill press and a careful machinist, that’s still probably a better option. There is no real reason to use VersaBlocks here, since the bearing holes can be “set it and forget it” once you calculate the proper C-C distance. VersaBlocks require attention to make sure that they don’t shift during competition, and you can’t set them to an accurate C-C distance.
Also, I would suggest using #35 chain rather than #25. While #25 is probably strong enough, you really don’t want to have to deal with fixing a broken chain mid-competition. #35 is also much more tolerant of sprocket mis-alignment, which means one less thing to worry about during assembly and as your robot takes a beating from defense. #35 will give you a much higher factor of safety without adding too much weight.
While I have never been on a team that ran chain in tube, one of the key features of chain-in-tube as I studied it was the lack (or at least reduced) need to provide tensioning. If you’re going to do tensioners anyway, put the chain out where you can see it unless there just isn’t enough width to fit everything otherwise. You may also want to consider using channel rather than tube if you have the space for it and might want to access your chain easier than if it were inside a tube.
We have run chain-in-tube with the Vexpro 25t #25 chain sprokets for both the 2017 and 2018 seasons. The tube we use is 2"x1.5" which helps a lot with assembly as there is a bit more room to fit the chain and sprokets in (takes ~20min to assemble if you do it properly), and also takes out the need of modifying the sprokets or making bearing blocks. Using 2"x1.5" tube also allows for better mounting options since there is more surface area. I also suggest leaving the ends of the drive tubes open so that you can service them if an issue does arise; just make sure to tape them up when you are done so that nothing gets in. I DO NOT recommend pocketing all the way through the tube as that defeats the purpose of keeping the chain completely inclosed and protected.
As for C-C between shafts, I recommend going exact c-c with the bearing holes directly in the tube and skipping the tensioners, just add ~0.018" to all the chain c-c’s to account for chain strech. This is mainly to reduce excessive backlash in your drivetrain as with chain in tube, it is impossible for the chain to jump off the sproket.
We have not run into any issues in any of the 4 robots that we built using this method, all of which have gone through lots of run time. I do caution, however, double loading the chains as the risk of a chain breaking goes up by a lot. If the double loaded chain breaks, you lose power to 2 wheels instead of just 1 wheel. We prefer to do each chain run directly from the gearbox to the shaft to avoid double loading chains. In the case of our 2018 robot, which is 8WD, we put 1 run outside of the chain to keep all the chain runs indepedent (also with no tensioners or issues). Although #35 chain is stronger than #25 chain, it just really isn’t worth have your drive chain 2.5x heavier (#25 chain is 0.087lb/ft and #35 chain is .21lb/ft).
For center drop, you can fit around a 0.1" drop on the wheels, and if you require more, you can raise the rest of the wheels up by 0.1" for a total of 0.2" drop.
In terms of machability, in 2017 the only milling machine we had was a small Littlemachineshop mill and we were able to make the tubes accurately. In the 2018 season we got a CNC router and now we run the tubes on that with a tube jig.
1 Like
We have had great success with WCD - Chain in Tube drivelines. We made the switch after 2016 - Stronghold, where we broke wheels every match and basically left us crippled since it took so long to put on the spares.
On all of our drivelines we have used 1.5in by 3in by 0.12in wall Aluminum Tube and #35 15Tooth sprockets. We do not use tensioners and use this site to calculate our C-C distance https://www.rbracing-rsr.com/chainlength.html
As far as serviceability, we cut a bunch of lightening holes in all sides of the tube which it makes it much easier to change a chain when needed.
We’re been running CIT for ~3 seasons now using an ever evolving chassis design. This past season we switched to using the versa blocks instead of doing C2C, and we likely won’t be going back to C2C. We made the change because of the loss of a machining sponsor to make our rails. You can follow along in our videos on how to make it here.
In all of the years we’ve run CIT we’ve yet to break a chain and the setup you described sounds very similar to ours.
My biased opinion is that’s its great.
118 has used chain in tube since 2013.
Build the chains properly and it should be maintenance free.
Use a 17 tooth sprocket and the chances of a chain jump are basically 0%.
Please consider Al’s corollary to Murphy’s Law…
Anything that can go wrong will go wrong, on EINSTEIN.
Check your design first for ease of diagnosis and second for how long it will take to repair on Einstein between matches.
I for one, hate mechanisms I can neither inspect or repair/replace easily and quickly.
In 2016, our team had a belt-in-tube system. I’m aware that there are very real differences between belt and chain (we have used chain ever since), but as Al said, maintaining and repairing that chain becomes much, much more difficult. During one of our matches, our belt snapped, and we had no way to fix it. We ended up being basically a hunk of metal for the rest of the competition.
Moral of the story, if you are creating a system that’s very difficult to maintain, have a backup plan. I personally wouldn’t recommend it because of the added risk, but you may find yourself with a different experience.
Done correctly chain in tube can be pretty cool, but done wrong and it is going to cause you a lot of headache.
The chain has to be the right strength for the game application, the right length, and typically I have seen better results when there are no half or master links in the chain, so chain that was built using the DS/VP chain tool.
I have only had 2 experiences with chain in tube design (5632 & 4901 in 2016), but both taught me a lot about how to get it to work well, and how to fail with it.
The thing I worry about with your tensioner approach is from what I have seen if the chain get’s loose enough over time, and jumps it will usually throw the chain off the sprockets completely, and when this happens with chain in tube…it can be a bad time.
Pros:
- More belly pan room
- No risk of robot parts getting caught in chain
- No risk of field debris getting caught in chain
- It looks clean
Cons:
- Longer failure recovery time
What do others think I’m missing?
The way I’ve looked at CIT over the years, it’s a trade off between a lower chance of failure for higher failure recovery times.
Does not tensioning chain (using the floor and ceiling of the tube to keep the chain from skipping) impact autonomous accuracy? I would assume that any time the robot is working to pick up chain slack it’s travel will be impacted. Assuming you are direct driving a wheel on the ground, it shouldn’t have a huge impact, but I would think the impact would be noticeable.
I’m leaving that off of the list mainly because that’s really independent of CIT. You can do C2C with chain in or out of the tube. You can also do tensioners in or out of the tube.
1 Like
https://i.imgur.com/84dtj0A.jpg
RIP the 221 Systems Double 17T Sprockets. These were ideal for doing chain in tube with bearings pressed directly into either side of 2x1 tubing.
Any supplier or person out there that would start selling these would gain a big fan right here.
My only request of a change to the existing model would be to extend the hub past the outer faces of the sprocket on each side so that the width of the total part fits snugly between two bearings pressed into a 1" wide tube. That way when dropping the chain through you don’t have to put spacers on either side. However this is a minor annoyance, I’ll take what I can get.
Seconded. When’s someone going to start 17ToothSprockets.com :]
1 Like
Third-ed? We used 221 Systems 17 tooth double sprockets on both of the teams I worked with that did CIT.
Looks like our wishes came true!
However, with Vex’s re-release of #25H chain, will that fit on these new sprockets properly? In CAD I measured the 17t sprocket center to center is .3588 and width of #25H is .354 (vs .307 of regular 25).
1 Like
#25H chain will not fit the Andymark sprocket. The pins will touch.
We were getting ready to re-launch our version of the product when we heard from AM that they were bringing theirs to market…
We could launch a version for heavy chain but I’m not sure it’s worth the effort.
Might be best to lean on AM and have them update their sprocket when they re-order.
All pins? Or just the masterlink?
All pins will touch.
Masterlink is a no-go with the existing design.