Coefficient of Frictions

[JVN]

Originally Posted by Joe Johnson

I am going from memory, but I believe they the soft rubber cross hatch stuff came out best at 1.3 or something like that.

In my experience, 1.2 - 1.5 is a common value for the traction of most FIRST robots.

I measured the wedgetop (we call it "60 tread") at about 1.2-1.3. That stuff is great, except that it has relatively poor wear characteristics compared with other materials of comparable traction. If you use it, be prepared to swap out your wheels once or twice a regional.

Joe,
I've just recently been "enlightened" to the virtue of mountainboard wheels.
(http://www.chiefdelphi.com/forums/sh...ad.php?t=22844)

I'm thinking of using them this year, game permitting. It's good to hear they have your endorsement, especially concerning traction and wear.

We found that the pneumatic skyway wheels have similar characteristics, however those (especially the 6" ones) are a little harder to put a hub on that these appear to be.


Thanks for sharing,
John

[Veselin Kolev]

Okay, prob with all these friction coefficients and stuff is that its all on carpet. You gotta be good on that wire mesh stuff too, and on the HDPE. We'll see what next years game is really soon.. but I can tell you about my team's wheel problems. Over the last few years my team has used the cross hatch "wedge top" stuff from McMaster. It has really good traction on carpet and fairly good durability on carpet. Its ok on the HDPE, but the wire mesh really kills it. If you try to push your way up onto the ramp with that stuff, all you do is burn out your wheels and cut them to shreds on the wire mesh. In fact, me and rest of the drive crew commonly refered to it as the "cheese grater" Yea... in half a day, our wheels would be slicks. Trash. So.. after seeing this at the Sacremento Regional, we invested some research into special wheels designed just for certain materials. For example, we had wheels with notches on them every one inch, giving amazing traction on the wire mesh. We only had 2 of our wheels like this, and the other 2 were I believe polyethelene... I can't remember, but they were really sticky rubber with crazy traction on the HDPE. Both of these tracks are on McMaster, and are usable just like the cross hatch tred. Of course, we didn't have as good traction on the carpet anymore, but our strategy didn't really call for that. We gave our robot traction where it needed it most. You can't have good traction on all 3 materials, its just too hard.
In the end, I feel that the cross-hatch tred is the best to start with, and it has good traction, but its durability might be a good reason to later change to another set of wheels. As JVN said, be prepared to have a few (actually a lot) of spare wheels.

Now for the fun part... I think that the best drive system (traction-wise) would be tank treds with a fully independant suspension on every support roller!! Just like modern day tanks. Hug the terrain. Man... what I would give to see ANY robot with a suspension on its wheels. So far, I've only seen one team that had 6 wheels use a suspension on the middle wheels.

Here's a question, how do teams that use treds like the cross hatch stuff attach them to wheels? My team uses the 8 inch skyways with sprocket hubs and we epoxy the tred on and use screws to keep the ends from tearing off. Anyone do it differently?

[mzitz2k]

Veselin,

The cross hatch material that you used is similar to the material that we used last year, just we used the "rough top" instead of the cross hatch. Our team lathed down some stock wheel-chair wheels and used carpet adhesive to attach the material to the wheel. Then, we screwed the material into the wheels by using a simple pattern. If I were to guess, I would say that we used about 20-30 screws per wheel on a 6" wheel, which may sound like a lot, but that is what it took to keep the material from ripping off. We also oozed epoxy in between where the two ends of the material met and then we simply filed it down so that it was smooth with the material. The nice thing about this method is that yes you must put the material on again and again, but you can tear it off the wheels if you remove the screws because eventually the epoxy wears away and the carpet adhesive is not too terribly strong. We changed our material before each competition and also after our last match on Friday (for the finals the next day).

Like you said, the material is great for the carpet, but wear does become an issue. As for the rough top, it was nearly perfect for the ramp last year - the teeth of the material seemed to mesh just so nicely. It was hard for other robots to push us off the HDPE, but that was due to our wedges (picture above) - not so much the wheel material.

I'm not sure that I completely understand your "wheel notches" that your team used. Do you have any pictures or could you expand on this a little for us? Thanks!

[Joe Ross]

Here is a little bit of data from a previous post: http://www.chiefdelphi.com/forums/sh...8&page=1&pp=20 Look for Paul Copioli and Garry Dillard's posts.

As far as the life of the mcmaster inclined conveyor belt material, when team 60 tested it, they found that it was very close to the coefficient of friction of other materials, but was much longer lasting then other types. However, this past year, the grate on the ramp did just that, grate the tread. Assuming that we have another flat carpet playing field, it holds up quite well.

[Adam Y.]

Does anyone know if the durometer of a material affects it's traction? I want a material that will have traction but not so much traction that it actually pops the tank treads off.

[Veselin Kolev]

Originally Posted by mzitz2k

I'm not sure that I completely understand your "wheel notches" that your team used. Do you have any pictures or could you expand on this a little for us? Thanks!

Yup, as I said they are all on McMaster, so... I found some pics of the material we used. As for your description of how you attached tred to wheels, we did the exact same thing except that we used high durability epoxy instead of carpet glue. Haha great minds think alike!

The first pic is the notched material, it has a 1/8" groove on 1" centers, fitting right into the wire mesh. The second is the ultra sticky material. It was black gum rubber material, and it had excellent grip on the HDPE. Hope that helps!

[patrickrd]

Here are the results of a test we did for RoboCup. This tested rubber on a felt carpet. It is similar to the FIRST carpet... This data is very good for comparison, but take the values themselves lightly, since the shape of the wheel and the carpet material will be different for FIRST use.

Buna-N: 1.29
Neoprene: 1.60
EPDM: 1.44
Viton: 1.11
Silicone: < 1.29 (not sure of exact value, but it is definately less)

- Patrick

[mzitz2k]

Originally Posted by patrickrd

Here are the results of a test we did for RoboCup. This tested rubber on a felt carpet. It is similar to the FIRST carpet... This data is very good for comparison, but take the values themselves lightly, since the shape of the wheel and the carpet material will be different for FIRST use.

Buna-N: 1.29
Neoprene: 1.60
EPDM: 1.44
Viton: 1.11
Silicone: < 1.29 (not sure of exact value, but it is definately less)

- Patrick

Interesting! What did the neoprene look like? There are so many different types... What was the hardness rating and did it have "teeth" or was it just a flat sheet of material??? If you have a picture of part number, even better!

Thanks again!

[patrickrd]

Originally Posted by mzitz2k

Interesting! What did the neoprene look like? There are so many different types... What was the hardness rating and did it have "teeth" or was it just a flat sheet of material??? If you have a picture of part number, even better!

The neoprene had Durometer = Shore A: 70. We used rubber o-rings as the test specimen, 1/2" outer diameter by 3/32" thickness (purchased from McMaster). This is a much smaller scale than used in FIRST. The o-rings were used for contact with the carpet on each roller of our omni-wheels.Testing each material was simply a matter of replacing the o-rings in the wheel. I've attached photos of the wheel and the test setup. In short, the weight of the setup provided a constant normal force, and a voltage was applied to the motor. The voltage was increased until the wheel slips. From this data it is easy to find out what the coefficient of friction is. The uncertainty is less than 5% for all the data.

- Patrick