Since I have been a member of 1410 we have exclusively used chains for our drive train and mechanisms. From team mentors I have really only heard bad things about them. I have seen many teams have great success with belts on drive trains and mechanisms. What are the advantages and disadvantages of belts and what are proper applications for them in FRC.
We used belts for years, but switched to chain last year using the Vexpro bearing blocks. We did this for ease of assembly more than anything, you can use them with belts too but we wanted to use the lower profile 25 chain sprockets over the larger 24t+ pulleys needed to make belts work well in the drive train.
Either belts or chain can succeed in an FRC drive train as long as the proper implementation is used.
KitBot Chassis (KoP) uses belt drive and we’ve never had problems with it
We use chain because it is more forgiving and we dont want to risk our drivetrain failing. We experimented with #25 chain during the off season since it is substantially lighter than #35 but are still using #35 since we are running 8" wheels with chain in tube. We use belts on lower load applications like intakes, or even our crawler to pull ourselves onto HAB3.
Here’s a great paper written by Team 234 on the topic, offering some insight into the differences and backed up by some testing.
We’ve had experience using both belts and chain on a drive train. Replacing a wrecked chain is generally simpler and faster than a ruined belt, but you won’t encounter either as long as you design and fabricate everything to specification.
Belts are nice on intakes and endeffectors because they’re significantly lighter than chain, reducing the mechanism’s impact on CoG and weight cost. In my experience, belts are also easier to design for since they seem to be more consistent from suppliers and tend to stretch less over time compared to chain (active tensioning is not necessary).
Since 2018, 3538 has been 3D printing most of its pulleys from ABS plastic. This gives us more options for pulley tooth counts at home, makes them more cost effective, and lighter than aluminum sprockets.
My personal design philosophy is use #25 chain in C2C (center to center) on drive trains then belts on most other mechanisms. This allows for simple drive trains that are robust. Then for mechanisms we can place the belt and forget it. We swap the drivetrain chain every 4-5 days of competition. The one time we have used chain on a mechanism recently is for our arm this year. We are using #35 chain in a C2C configuration because it is easier to swap out chain than design a way to tension it IMHO.
How is chain more forgiving?
We have been using belts on our custom drives for three straight years and we have never had a single failure. Never had to adjust tension and never had to replace one. Even when the chassis rail was severely bent from impact they kept working. Can’t get much more forgiving than that.
4388 transitioned to a belt drive this year. We have our entire belt system inside the drive tube, this has been nice for space on the robot. We’ll see you guys in Colorado and I’d love to show you what we have going on. It’s been nice, and we haven’t had any failures with it yet throughout practice and driving.
Just from experience when we used the Rhino Tracks in 2016, the small 5mm belt was required to be replaced once an event it felt like. Chain is also easier to replace in the case of a failure, allowing us to remove a master link, where belts require disassembly. We’ve had issues in the past with improper use of belts in applications so it’s really just personal preference. We were going to experiment with them more this past season but spent the time messing around with chain in tube drivetrains.
Designed right belts are easier to deal with than chain, in my experience since they have no issues with stretch, tight links, poorly assembled master links, etc. They can be less forgiving to design with since dealing with design errors and sizing can be a pain. Chain is easily available in any full link size.
We have used them exclusively for the last 4 years. The down side is that in high load (big motors, small sprocket) situations it is easy to shear teeth and have catastrophic belt failure. We have twisted and snapped 1/2" hex shaft, without breaking #35 chain.
In terms of driveline belts, we have had absolutely no problems using them with the KOP drivetrain, including with 8" pneumatic wheels and a repeatedly airborne robot in Stronghold.
This has been the first season I’ve used belts on a robot and I’ve been very happy with the results. Sometimes it was driven by comfort of “we’ve always used chain” or costs. The last two teams I’ve been on have had a plethora of #25 sprockets and chain available so it was easier. Depending on the mechanism there was foresight that it might change slightly so it was easier to make a new chain than order a new size.
Spending more time this year in CAD for parts made on a laser cutter, waterjet, or router has been a great year to give the GT2 on the robot. I’ve been very impressed by the ease of WCPs C-C calculator, selection, reduced noise, no grease, and they are easy to install. The 775pro 12t pinions are great! Our Cargo intake packaging is a near copy of 319’s Onshape model and we loved the first stage belt reductions right off a 775. Light and clean.
We are iterating our Hatch mechanism before Pine Tree and will find more applications for GT2 belts.
May I suggest you try #25h chain as it is hardened and it is designed for high intensity drive trains. It also stretches less than the regular #25 chain.
Suggest all you want, but never had a problem with regular 25 chain. Might have just switched to it last year for my current team but used it plenty of times on other robots.
Also never seen stretch enough to have to retention…
I may have come off the wrong way. I don’t mean to say #25 is bad. I am just speaking from my teams experience of chain stretching or breaking, and the frustration with it. We are actually currently using normal #25 on our drive train and have used it on our practice bot. I just wanted to help you possibly in case you had experienced any frustration from it. I apologize if I came off as being rude.
15mm Belt in tube FTW. Chain is loud and not very cute…
Advantages of chain:
- More flexibility in spacing, since the it’s length can be adjust
- More forgiving to manufacturing slop
- More forgiving to mechanism redesign
- Better in high load applications (e.g. an arm), especially shock loads
- Narrower for the same amount of load
- Easier to buy ahead of time, since the same #25 chain and sprockets will work for most applications
- Less upfront design work needed and usuable without ahead of time CAD
- Can be engaged on both sides (useful for elevators)
- More efficient
- Better for high speed applications (e.g. shooter), partially because of efficiency.
- Won’t stretch over the competition system
- Can be designed without tensioning, where as chain is more likely to need tensioning or replacement (some recommend tensioning systems for belts too, though)
I’m not an expert at this, so correct any points that are wrong.
One thing that’s helpful in figuring out these nuances is looking at top teams (who have the resources/manufacturing capabilities/design experience to choose between belt and chain for any particular application), and asking yourself why they chose belt or chain for any particular mechanism.
I completely agree that well designed belts are just as good if not better than chain. I personally like #25 over belt because of the strength to width ratio.
chain is easier to reapair and assemble