Which mecanum wheels to buy?

[Max Boord]

Originally Posted by Ether

Why do you suppose this is ?

This is due to the 8 inch wheels having less (12 vs 16) and larger rollers (~ 2x the size) therefore causing the wheel to act in a slightly less round way.

Electronica1, the wheel upgrades where not out in 2012 which was the last time we used them. I would think the problem would still exist although slightly reduced with the newer wheels.

[Ether]

Originally Posted by Max Boord

This is due to the 8 inch wheels having less (12 vs 16) and larger rollers (~ 2x the size) therefore causing the wheel to act in a slightly less round way.

How does fewer rollers cause the wheel to act in a slightly less round way?

[buchanan]

Originally Posted by Ether

How does fewer rollers cause the wheel to act in a slightly less round way?

There isn't any reason they must necessarily, but it does seem in practice that the 8" AMs do act that way. Some possibilities:

• The curvature of the rollers may simply be incorrect for the wheel geometry. Has anyone confirmed that it's exact?
• There may be significant bending deflection of the rollers across their width, changing the effective wheel radius depending on where on the rollers the load is.
• Since the radius, and hence the thickness of the rubber, is different in the middle vs the end of the rollers, the compression of the rubber itself under the same load will be different at different points. This can be compensated for, but only at one load value, right?

I have not investigated any of them to see which is/are of practical importance, but I can confirm the observed phenomenon, and it surely does have an explanation somewhere in theory.

[Ether]

Originally Posted by buchanan

There isn't any reason they must necessarily,

I concur.

but it does seem in practice that the 8" AMs do act that way.

That's been my observation

Some possibilities:
The curvature of the rollers may simply be incorrect for the wheel geometry.

That is probably a contributing factor.

But my guess would be that the main factor is one not on your list: the transition from one roller to the next is not smooth. It's a not clean handoff. You can clearly see this on a shadowgraph. You can do this experiment at home:

In a dark room (a basement is ideal) place a sheet of paper on the floor.

Put a small coffee or tea cup in the middle of the paper.

Balance the mec wheel on the cup so the plane of the wheel is parallel to the floor, and less than 4" from the floor.

Mount a spotlight on the ceiling facing down directly above the wheel, at least 8 feet above the wheel.

Darken the room and turn the spotlight on.

Observe the circumference of the shadow of the wheel on the paper.

You can clearly see the "identations" where one roller ends and the next begins.

Here's a mec wheel roller design calculator if anyone's interested:

http://www.chiefdelphi.com/media/papers/2390

... scroll down the list of attachments and download the one named
"2011-09-30a mecanum roller design calculator"

[Ben Wolsieffer]

Originally Posted by Ether

But my guess would be that the main factor is one not on your list: the transition from one roller to the next is not smooth.

I have noticed the same thing. The outer rollers on the new AndyMark wheels touch the ground during the transition from one roller to the next. When we compared our old wheels to the new ones we could feel the big difference that the extra rollers had simply by rolling them across a table.

[adciv]

Were those these 8" wheels or a different set?

We used the 6" VexPro ones this year and our bot was shaking quite a bit.

[Ben Wolsieffer]

Originally Posted by adciv

Were those these 8" wheels or a different set?

Those were the wheels we used this year. For us they replaced these wheels which we used in past years.