Belt Driven instead of Chain driven

[Camren]

Im not sure if this is legal for the wheels of the robot. But ive seen arms that use belts instead of chain which frankly seems like a better and cleaner installation than nocking out segments from 100 feet of chain. so can anyone help me out and tell me if its legal to have a belt driven drive train. Also your opinion on which is better

[akoscielski3]

I have never done it personnaly but im sure its legal. Just after shipping this year a team put a picture up of their drivetrain, driven by belts instead of chain.

[Joe G.]

Im not sure if this is legal for the wheels of the robot but ive seen arms that use belts instead of chain which frankly seems like a better and cleaner isntallation than nocking out chains for 100 feet of chain. so can anyone help me out and tell me if its legal to have a belt driven drive train.

Completely legal (per 2011 rules); don't make up rules that don't exist.

Whether or not it's better is a matter for debate. Belts need to be tensioned much more precisely, and can be a bit tricky to work with, since they can't be split like chain can. You need to plan things out ahead of time, and get exactly the size of belt you need.

[lemiant]

Im not sure if this is legal for the wheels of the robot but ive seen arms that use belts instead of chain which frankly seems like a better and cleaner isntallation than nocking out chains for 100 feet of chain. so can anyone help me out and tell me if its legal to have a belt driven drive train.

Short answer: Yes

Long Answer: If the rules don't disallow something then go ahead. In the case of most years that means you can use whatever mechanical component for whatever purpose you want, as long as you are within the weight and cost limitations (note: there are lots more restrictions on electronics and pneumatics).

Remember though, that just because it is legal does not make it a good idea, the reason most teams use chain is because they have decided it is best for them. There have been some threads comparing belts and chain. If you are interested in using belt, do some research.

[kramarczyk]

You may want to consider reading this white paper on chains vs. belts (http://www.chiefdelphi.com/media/papers/2246) as you contiplate about which is better for you.

[Hawiian Cadder]

Belts require a much more precise frame, so I would never use belts with the kit frame. belts take up a lot more space than chain as well.

[Al Skierkiewicz]

Many teams have used belt drives in the past. They need less maintenance and are quieter. They are much more difficult in some designs to replace though. (Most belting doesn't allow you to thread one end through the drive system and then connect the ends together) Teams have been able to use toothed belts and sprockets to great advantage in the past. While the robot rules have not been released for 2012, I would not expect a change making belting illegal. Please check the rules when they are published. Be aware, that belts can inflict personal injury the same way chain can. This might require some protection of the mechanism to prevent injuring anyone who will be exposed to the mechanism, including field volunteers.

[Chris is me]

Belts require a much more precise frame

They do?

so I would never use belts with the kit frame.

And what about the kit frame is imprecise at all? It's not like the kit holes are randomly placed.

belts take up a lot more space than chain as well.

Indeed true, though usually belt pulleys don't need separate hubs like smaller chain sprockets do.

Belts are workable just like chain but they require a bit more thought and design than a standard 35 chain drivetrain to work. You also have to be careful with belt reductions to make sure you're not putting too much stress on the pulley teeth.

[Andrew Lawrence]

We've used belts for our encoders, but so far never for our drive train, or even for our arm apparatuses.

[annie1939]

My team hasn't done a belt drive, but there are some products out there that could simplify doing one. AndyMark sells a v belt pulley http://www.andymark.com/product-p/am-0592.htm. It was developed by team 1810. It is made in 2 pieces of plastic so that when you need to replace the belt, you just unbolt one side of the pulley and slip the belt on. And being plastic they are light, unlike the traditional pulleys. 1810 could tell you more.

There are also Link v -belts that you can shorten or lengthen yourself. You can also remove a link and thread them like you would a chain and reconnect them. Harbor Freight has some at http://www.harborfreight.com/vibration-free-link-belt-43771.html. Grainger and IBT also have them. I haven't used these, but they look intriguing. I have no idea what kind of slippage they would have.

If you want to go with the Gates Belts, they have a lot of videos to show you how at http://first.gatesprograms.com/video. My team was never organized enough to know what length we needed in time to order.

I know every time I pick up the chain, I keep thinking about how wonderful it would be to shed those pounds.

[AdamHeard]

My team hasn't done a belt drive, but there are some products out there that could simplify doing one. AndyMark sells a v belt pulley http://www.andymark.com/product-p/am-0592.htm. It was developed by team 1810. It is made in 2 pieces of plastic so that when you need to replace the belt, you just unbolt one side of the pulley and slip the belt on. And being plastic they are light, unlike the traditional pulleys. 1810 could tell you more.

There are also Link v -belts that you can shorten or lengthen yourself. You can also remove a link and thread them like you would a chain and reconnect them. Harbor Freight has some at http://www.harborfreight.com/vibration-free-link-belt-43771.html. Grainger and IBT also have them. I haven't used these, but they look intriguing. I have no idea what kind of slippage they would have.

If you want to go with the Gates Belts, they have a lot of videos to show you how at http://first.gatesprograms.com/video. My team was never organized enough to know what length we needed in time to order.

I know every time I pick up the chain, I keep thinking about how wonderful it would be to shed those pounds.

973 recently started looking into belts, and used them on a few things recently with success. We read pretty much all the documentation Gates has to offer, the gt2 tooth profile is very nice. We won't be replacing the chains in our drive with them for various reasons, but we'll certainly use belts in various places. We LOVE the selection sdp-si has, and we also love the plastic sprockets.

[AustinSchuh]

... the gt2 tooth profile is very nice.

This is key. The GT2 profile can take more than twice* the load that the standard tooth profiles can take.

The belts are well documented, and after you figure out what is important, the docs are not bad. You can easily check what belt width, and tooth counts will support what loads.

*It could be even bigger than this. I haven't read the gates manual recently.

[Hawiian Cadder]

They do?



And what about the kit frame is imprecise at all? It's not like the kit holes are randomly placed.


Indeed true, though usually belt pulleys don't need separate hubs like smaller chain sprockets do.

Belts are workable just like chain but they require a bit more thought and design than a standard 35 chain drive-train to work. You also have to be careful with belt reductions to make sure you're not putting too much stress on the pulley teeth.

My wording was not as precise as it should have been, what I meant was the kit frame has some weaknesses in that it lacks rigidity. I also have found that the kit frame doesn't tend to stay perfectly square without a substantial amount of additional support (plywood, or larger gussets) so I would choose to use chain with the kit-bot purely because the flexing might cause damage to the belts, and if the frame warps into a parallelogram, the belts will wear out at an accelerated rate.

This white paper does a good job of quantifying the bending problems on a standard 2010 robot. the frame bent enough to throw #35 (i think) chain, so belts would be out of the question.

http://www.chiefdelphi.com/media/papers/2398

[Chris is me]

This white paper does a good job of quantifying the bending problems on a standard 2010 robot. the frame bent enough to throw #35 (i think) chain, so belts would be out of the question.

The 2010 frame was substantially different than the 2011 frame, since the axle mounts were stood off several inches (6? 8? more?) from the actual chassis. The issue that really caused the bending and flex wasn't the kitbot profile as much as the modules that wheels were mounted to. The parts flexed relative to one another and your end result is thrown chain.

Belts require more or less the same amount of precision that 25 chain requires.

[kramarczyk]

This white paper does a good job of quantifying the bending problems on a standard 2010 robot. the frame bent enough to throw #35 (i think) chain, so belts would be out of the question.

http://www.chiefdelphi.com/media/papers/2398

Yes, it was #35 chain thrown off that chassis, but be careful not to take this one data point as a universal problem with the kit chassis.

The 2010 frame was substantially different than the 2011 frame, since the axle mounts were stood off several inches (6? 8? more?) from the actual chassis. The issue that really caused the bending and flex wasn't the kitbot profile as much as the modules that wheels were mounted to. The parts flexed relative to one another and your end result is thrown chain.

As Chris points out this is an issue of structure. In the case of the 2010 kitobt it is a lack of triangles supporting the chain path and the open section of the wheel risers that cause the problem.

The kit frame should only be one portion of your entire robot's structure. Step back and look at the whole robot. Use the other systems to reinforce the chassis and use the chassis to reinforce the other systems.

I used #25 chain on a kit frame this year with no issues.

[AdamHeard]

This is key. The GT2 profile can take more than twice* the load that the standard tooth profiles can take.

The belts are well documented, and after you figure out what is important, the docs are not bad. You can easily check what belt width, and tooth counts will support what loads.

*It could be even bigger than this. I haven't read the gates manual recently.

As you guessed, it is bigger! (http://www.sdp-si.com/D265/HTML/D265T015.html)

The most relevant piece here is comparing XL belt to 5mm gt2, pretty much identical belts in terms of size and weight. It's amazing the tooth profile makes such a difference.

[akoscielski3]

Im Working on a chassis, and want to put belts on it. Where do I find Belts/pulleys for CAD and how do i buy them?

[Aren Siekmeier]

The SDP-SI catalog (https://sdp-si.com/eStore/Direct.asp?GroupID=217) has lots of different pulleys of various tooth profiles, pitches, tooth counts, etc. and if you click the part number you can find a CAD download in the window that pops up. At the very least this gives you the tooth profile which you could rework into other pulley geometries (different hubs, flange/no flange, etc.).

However, usually you don't care about the actual tooth profile so you could just create part files with all the other geometry (hub diameter, length, pitch diameter, etc.) and put those in the model. Solidworks has a belt feature under Insert>>Assembly Features that allows you to route an "imaginary" belt (it just shows a line) that you can use to measure the length. I'm certain similar tools exist in Pro/E/Inventor.

[Billfred]

I can't say we've ever done them on 2815, but my alma mater 1293 used them this year on their kit-frame drive system. Twisting issues from their arm tower caused enough issues that they took them out for the connections from their gearboxes to their rear and middle wheels; to my knowledge, the middle-to-front belt run gave them no real issues. I'll try to get someone from 1293 into this thread.

[akoscielski3]

Is it possible to do belt drive with a dead axle? im trying to do this because i have NO experience with live axles and want to try belts. If it is possible is there an example i can see, picture or a CAD i can see?

[EricH]

Is it possible to do belt drive with a dead axle? im trying to do this because i have NO experience with live axles and want to try belts. If it is possible is there an example i can see, picture or a CAD i can see?
I've seen it done, but in a superstructure. 330's 2006 robot had a series of dead-axle belt connections in its feeder system, and a dead-axle belt-chain connection in the shooter system. At least, I recall it being dead-axle.

[akoscielski3]

I've seen it done, but in a superstructure. 330's 2006 robot had a series of dead-axle belt connections in its feeder system, and a dead-axle belt-chain connection in the shooter system. At least, I recall it being dead-axle.

Thanks, i think i got it. I put holes in it to bolt it to the wheel, than removed the hub part that isn't needed anymore because its dead axle (usually would attach to live axle itself)

[Aren Siekmeier]

Thanks, i think i got it. I put holes in it to bolt it to the wheel, than removed the hub part that isn't needed anymore because its dead axle (usually would attach to live axle itself)

Exactly. Beat me to it. It's no different from mounting a sprocket to a wheel in a dead axle configuration. And as you said, you can remove quite a bit of the hub material, making it lighter (you also kind of need to for bearing clearance, depending on how it's set up).