How many rivets on a wheel is "too much"?

Our team was preparing 4in. Andymark Performance Wheels to finish up an off-season project before the new season begins. We purchased and successfully cut blue roughtop tread to fit the wheels.

However, we just realized that these wheels have a whopping 12 holes for rivets in them. In the end, we riveted one wheel with all 12 rivets, and needless to say it was a very difficult and long process.

Is it overkill to rivet these wheels with all 12 rivets? I was thinking we could attach 4 rivets (2 each) at the ends of the tread, and possibly 2-4 rivets in the middle of tread.

http://www.andymark.com/product-p/am-0395.htm

I don’t think there is such a thing as “too many” rivets when it comes to wheel tread. What we found out during competition was that the rivets at the end of the tread (where the seam is) tended not to do the job very well under heavy competition wear. A quick solution we came up with was to simply drill a couple extra 1/8" holes through the tread and wheel (right near the end) and put extra rivets in there after the wheel was completely assembled. They withstood competition much better, and it’s a pretty easy way to fix tread that’s fallen off during crunch time.

Edit: As for your riveting suggestion, that would probably work moderately well (but it’s always better to construct things on the safe side). Just keep in mind that the connection between the tread and the wheel is only supported by the rivets and the tiny lips that the tread sits between on the wheel. Meaning, if your robot is being pushed from the side, a significant fraction of that force is going to be shearing the rivets holding the tread on. If you’re lazy you could just drill new holes (through tread and wheel) once you already have the tread in place and rivet them that way.

This doesn’t answer your question, but where did you buy your treads from?

We use a wheel similar to an IFI traction wheel and we use 4 rivets at the seem, and them about 2-3 around the perimeter.

It helps to use an epoxy between the wheel and the tread before the rivets.

There is such a thing as too many rivets: when the inspectors start having to worry about your metallic wheels.1 Though FIRST doesn’t like hard things in contact with the carpet, they’ve been reasonably lenient with the fasteners used to secure tread to wheels. (Consult the 2012 rules for details.)

Edit: If you’re interested in a bit of history about metal on carpet and tread fastening, see this thread.

1 (http://www.chiefdelphi.com/media/photos/13530)

I’ve found that two in each end worked fine, if installed properly with the tread tight around the wheel. Now keep in mind, this was on custom wheels. After previous seasons and once plunging into my hand with a drill (stupid mistake) while changing treads, I became so tired of dealing with treads that I decided to go with Colson wheels. Install them once and never touch them again (even over parades and demos on asphalt). You do take a bit of a weight penalty, but in my mind it was worth it for the zero-maintenance aspect.

McMaster-Carr.

As for the metal on carpet, this is an off-season project so it doesn’t necessarily have to comply with FIRST rules, although you perfectly right to bring that up, since this project is basically practice for the upcoming season.

Interesting. By any chance do you have a link to the specific ones you use?

I believe we used the 5 x 1 1/2 simply because it worked out well for speeds directly driven off an AndyMark Super Shifter, but they’re available in many sizes. http://www.robotmarketplace.com/products/colson_wheels.html

We have always just put 2 in each end. This goes for our 4" IFI in 2006, custom 4" 07-10, and AM 4" performance in 2011.

How often do you have to do that, though? In my experience, we had to repin the tread at least once per competition, no matter how many rivets we put in, or what wheel and tread combo we used.

Whether you rivet just the ends or in all the holes on a Performance wheel, the ends of the tread might still pop out - more rivets throughout the tread does not make it more secure (though more rivets at each end would). However, having the tread riveted down all over would allow you to keep driving on the wheel fairly safely whereas losing one set of end rivets on a wheel that was only riveted at the ends creates a crappy flappy wheel.

Which is exactly why epoxy is also recommended for competition use.

We only had to reattach tread in 06 when we used the first version of the IFI wheels because there where some design issues (which have since been resolved). I usually tread the wheels once with our lead mechanical student and don’t have to worry about it. In 6 years we have only blown a tread once back in 2006 and that was the afformentioned wheel issue.

BTW In case anyone was wondering we use stainless steel 3/32" rivets for the tread.

Ah, that might help explain it; we use 1/8" because that’s what the performance wheels are predrilled for. Did you use exploding rivets or just normal pop rivets.

Correction we use 5/32" rivets. That was a typo on my part.
We use pop rivets normally, but I may try solid rivets this year just to see how well they hold, and I have a bunch of Inconel rivets left over from a project at work so why not.

Also we use the large head diameter so they wont tear out as easily. I occurred to me that I didn’t see anyone mention whether they used standard or large head rivets.

Kind of heavy? What do you use Inconel rivets for, anyway?

Not that much heavier than stainless for the size, and the price of free makes it worth trying them on a practice robot.

I actually design aircraft engines these days, we use a lot of Inconel solid rivets in various areas of the engines to attach spring clips and nut plates.

We use shoe goop as an adhesive, and drill holes for safety wire to “strap” across the tread material. The wire is then twisted tight near the wheel core, and pulls in far enough not to contact the playing surface. Well worth the extra effort. Haven’t lost a tread yet in the 10 years we’ve been doing this.:slight_smile: