Timing Belt Drive System

[maclaren]

We are looking to build a timing belt drive system and I was wondering if anyone has built one out there.

What type and size of belts did you use?
How many wheels were in the drive system?
How did you tension the system?
What problems did you have?

Just curious what peoples experiences have been with a drive system like this.

We are planning on making a six wheeled drive system with spring loaded tensioner and serpenting the belt around the drive sprocket and the wheels. I haven't looked at had size or profile of belt we are going to use yet. We plan on using about 8" drive wheels with about 7" wheel pulleys. Yes we plan on either having spokes on the wheel pulleys or making them hub-less. We are planning on using live wheel with fixed axle and attaching the wheel pulleys to the face of the wheels. We have not decided on what wheels we are going to use ideally we will use a fully custom wheel made at our machine shop and then rapped with rubber from a local rubber mill we work with.

Thanks for the feed back.

[Greg Perkins]

Maclaren,
This is purely my opinion and in no way am I forcing my answer towards you, I also appreciate this community and the fact that we are able to hold discussions and have ones opinions posted with utmost respect.

I don't believe a belt drive is effiecient enough for a few reasons

• It would require a custom made belt, while #25 or #35 chain is cheap and is customizable
• Belts tend to slip, and with a drive system in FIRST, you want your wheels to slip, and not your drive system.

My former team in 2002 made a two speed transmission out of belts, and it was a complete failure. If I was in your shoes, I'd use chain for my drive system (it's a proven system) and not design the most crucial part of any FIRST robot with questionable variables.

Again, this is my professional opinion, and I thank you for the chance to respond to you.

Sincerely,
Greg Perkins

[maclaren]

Thank you for the response.

I welcome design critique and plan on always welcoming it.

The reason that we don't want to use chains is that we have stretched several chains before which is a big pain in the butt.

Also chains are heavy compared to belts. Sprockets are also heavier than pulleys.

The amount of tensioning that we have built in can compensate for the difference in standard length belts. So I'm confident that we can find one that fits.

Also we are going to make sure to serpentine the belt around the drive pulleys as much as possible to eliminate any chance of the belt slipping.

I have attached a sketch of what we are planning on doing.

If you are asking yourself about the engagement of the center wheel pulley yes that is a concern but it can be compensated for with additional idler pulleys around the wheel.

As for the actual transmission that will drive the drive pulley it will be two cims that are geared together and attached to the drive pulley. We plan on making the drive pulley either out of aluminum or steel and it will have a large mating cross-section with the output shaft from the gearbox.

[Arefin Bari]

Maclaren, my old team 108 tried to use belts in 2003 and 2004 instead of chain simply because of the weight issue. 2004 was nothing but a huge mess. I have looked at the sketch you attached with your post. In theory it works (we had it done very similarly in 2004), but in reality it didn't. At the same time in 2003, team 108 used belts and pulleys and was very successful. It really depends on how you design the chassis.

The 2003 robot was a 4wd, and the whole front module would go back and forth horizontally in order to tension the belt (using a quarter inch 4" long bolt). On the other hand, the 2004 robot (4wd) had the tensioner tension the chain vertically which failed.

At this point, what I would suggest is design the chassis, post it up here on the forum and let Andy Baker, John V. Neun, Paul Copioli, Ian Mackenzie, Tim Baird have a look at it. There are many other engineers, mentors, and students who will help you perfect your design.

Here is a picture of the 2004 chassis. I couldn't find a 2003 one, when I do, I will post it here.

Also when you get a chance, please read this thread

[sanddrag]

There's a reason you don't see more belts in place of chain of FIRST robots. It's a brutal field out there.

[Not2B]

Well, success or failure - you'll learn something.

The Stock Drive website has alot of good information about timing belts and pulleys. Reading through it will give you a bit more information. Belt size, tooth profile, pulley types, and more will all effect how much power you can transmit and at what speeds. Also have an effect on belt streach and such.

http://www.sdp-si.com/Sdptech_lib.htm

Don't forget, timing belts drive motorcycles.

I don't know if a belt drive system will work - I haven't done the math. But I think it's worth the thinkin' time.

[Madison]

I haven't touched timing belt since 2002 and had limited success at that, but I can see its appeal. The single largest issue I had to overcome in both 2001 and 2002 was preventing the belt from walking off the pulleys.

Out of curiousity, since it appears as if you're striving simply to replace roller chain with timing belt, have you considered taking things one step further and replacing the wheels with timing pulleys? It aggravates the problem mentioned above, but there are self-aligning belts and things that I've never been able to play with that may alleviate some of that concern. If there's a simple way of keeping your belts on, it'd seem a no-brainer to save even more weight by riding directly on your pulleys rather than attaching them to a wheel.

[Kyle Love]

Maclaren, I do see how you want to save weight, but as I look at it, I would like to see most of the weight put into the drive system. Sure, I may be bias because I drive the base but you need a solid drive system to get around a FIRST field now-a-days. Chains do weight quite a bit more then belts, but chains don't stretch typically, unlike belts that stretch more commonly. As Greg said, belts slip, chains dont, in the game of FIRST you would most commonly like to see the wheels slip before the drive system itself. However, I think it would be a cool endeavour but you may find out that it will give you periodic problems throughout the year. It may be cool to do in the fall, as a team building activity to train new team members.

On a bit of a sidenote, a different type of drivetrain done in the past was a ball drive. It used metal pads to turn the balls, and chains to power the omnis. It was my teams "show bot" that was only used in a few matches during the 2003 season.

[MattB703]

Nice design concept. I think it is great that you are looking at different solutions. We all need to do some of that to get progress.

I have a couple of comments/critiques based on your sketch that I hope will be helpful;

1. You mentioned using timing belts. That usually means cogged or toothed belts. Is that what you intended? One potential problem is that you have your drive pulley contacting the outside surface of the belt while your wheel pulleys contact the inside surface of the belt. With this kind of arrangement you would need to use a belt that is cogged on both surfaces.

2. Your drive pulley is pretty small. You will lose a lot of efficiency trying to drive a belt around a radius that small.

3. This one is a little complicated. A cogged belt is like a chain. It has discreet positions that it contacts at (IE. if the belt is held still you can turn the pulley 1 tooth but not 1/2 of a tooth) The way your design is now you have the center wheel pulley contacting the belt in 2 discontinuous positions. Let's say that the teeth on the bottom of the middle pulley are meshed with the belt. Now we will trace along the belt around the drive pulley and the back pulley and get back to the contact point between the belt and the top of the middle pulley. What is the chance that the belt teeth will be "exactly" lined up with the pulley teeth?

Don't give up. I believe that high strength timing belts are quite viable for FIRST robot drive-trains (I haven't convinced my team yet).

Good luck.

Matt B

ps
here is a concept I have been batting around my head. It is very far from finished, but maybe it will start some thoughts.
[IMG]belt drive system.jpg[/IMG]

[Warren Boudreau]

If you are going to use timing belts, why use wheels?

There are many teams, Team S.P.A.M. included, who have used timing belts as treads.

Recently, we went to a dual tread design so if one tread should break during a match, you can still drive around. It works quite well.

[sanddrag]

I've always thought the belts that drive blowers like on dragsters would make good tank treads. But then again, unelss you are climbing stairs, I see no purpose for tank treads.

[Ben Piecuch]

The Aztech's, Team 157, used timing belts and pulleys on their 2005 robot. As far as I know, they didn't seem to have any problems with them. They used some fairly beefy belts, 1-1/4" to 1-1/2" wide, with a 3/8" or 1/2" pitch. They look to be fiber reinforced, so I can't comment on their strength. However, it was not a continuous loop, as the bot had two belts per side in a 4wd setup.

I'm waiting for a couple pics to be uploaded, stay tuned for links.

BEN

[John Gutmann]

Originally Posted by M. Krass

I haven't touched timing belt since 2002 and had limited success at that, but I can see its appeal. The single largest issue I had to overcome in both 2001 and 2002 was preventing the belt from walking off the pulleys.

Out of curiousity, since it appears as if you're striving simply to replace roller chain with timing belt, have you considered taking things one step further and replacing the wheels with timing pulleys? It aggravates the problem mentioned above, but there are self-aligning belts and things that I've never been able to play with that may alleviate some of that concern. If there's a simple way of keeping your belts on, it'd seem a no-brainer to save even more weight by riding directly on your pulleys rather than attaching them to a wheel.

If your going to go that far, wouldn't it be worth it to just go all the way to treads? Although I can See the advantage of just having the wheels.

Also, why not have a spring tensioner on one side and just drive the middle pulley.

[157weldthebot]

I am from team 157 and the belt system worked so well in 05 that we went with it in 06 we get all the belts and the pully stock from the same company and the belts mesh perfectly so there is no chance of sliping and in 2 years we only changed 1 belt and that was because we got another bot inside and it got scratched and we didnt want to chance it. but this was a huge difference form the many chains i changed sophmore year on our bot. once you go belts you wont go back

Email me if you are intrested in detials pics company names or whatever
xtreme19881@aol.com

tag it belt help or something to that effect

[A_Reed]

Here is a resource for belts in general. Brecoflex...they sell belts to custom lengths and pulleys with Self tracking profiles...commonly used for tank treads but you might be able to find something your looking for in timing belts.