Maximizing the in-line CoF of Omni-wheels

[AustinH]

Has anyone ever tried adding custom tread patterns to omniwheels in order to maximize front-to-back CoF while retaining free lateral motion? Could you theoretically use custom rubber treaded rollers? I would love to hear of and see examples of people trying this, if it's been attempted at all.

[asid61]

I was thinking about using rubber rollers with plastic hubs on the omnis, but you would have to make them yourself.
Following.

[Dunngeon]

I've never heard of this done, but building the rollers wouldn't be to hard, Polyurathane would work in that application (same type of material as colson wheels). However, I'm not sure it would be worth your time to develop this type of Omni-wheel unless it had significant benefits (which I don't see).

[Bryce Paputa]

Why wouldn't this be beneficial? It could give you significantly better traction with almost no additional weight.

[Nate Laverdure]

Originally Posted by Bryce Paputa

significantly better traction

Significantly better than what? AndyMark lists a forward COF of 1.0 for their 6" dualie omni wheel.

[Bryce Paputa]

Originally Posted by Nate Laverdure

Significantly better than what? AndyMark lists a forward COF of 1.0 for their 6" dualie omni wheel.

Significantly better than omnis with standard rollers. Even if you could only get the COF to 1.1 or 1.2, that would be 10 or 20%. Looking at the difference of the two different vex versawheels (http://www.vexrobotics.com/vexpro/wh...sawheels.html), the W tread adds an additional 20% compared to the DT tread. You could probably get a similar increase by adding a pattern to the omni rollers.

[Nick Lawrence]

I might be concerned about the depth of the cuts. If they were too deep, you definitely could run the risk of chunking out the rollers.

-Nick

[Bryce Paputa]

Originally Posted by Nick Lawrence

I might be concerned about the depth of the cuts. If they were too deep, you definitely could run the risk of chunking out the rollers.

-Nick

Yeah, I think that I would get bigger rollers (anyone know how?) and trim them down with the cuts.

[DampRobot]

Most omnis have rollers with rounded edges, that is, they try to follow the arc of the outside of the wheel to minimize bouncing as the robot drives. If they were kept square, it's possible that they could dig into the carpet at the edges, and still drive relatively well (because the carpet would compress to compensate).

[Dunngeon]

Originally Posted by Bryce Paputa

Significantly better than omnis with standard rollers. Even if you could only get the COF to 1.1 or 1.2, that would be 10 or 20%. Looking at the difference of the two different vex versawheels (http://www.vexrobotics.com/vexpro/wh...sawheels.html), the W tread adds an additional 20% compared to the DT tread. You could probably get a similar increase by adding a pattern to the omni rollers.

All things equal, it would, but given how Omni's behave in a pushing match against another robot that isn't dead center. It would likely negate much of the gain that would be seen. (i.e. they break loose/slide in a direction away from the pusher, effectively rendering the friction gain useless)

Given the amount of time that would be put into building these (taking apart the wheels, rollers, testing,ect.), I think a team interested in Omni's could easily design something else that fits the bill better (Such as Butterfly) and be much happier with the results.

[EricH]

Back when mecanum wheels were still new to FRC, at least one team tried getting better grip with patterned rollers/rubber coatings on plastic rollers. As I recall, it wasn't fully tested, but there was an improvement. (Though as I recall, the biggest improvement came from adding the rubber, so that's out with AM wheels.)

[Andrew Schreiber]

Originally Posted by Bryce Paputa

Yeah, I think that I would get bigger rollers (anyone know how?) and trim them down with the cuts.

3D print a master, make a silicone mold, poured urethane rollers.

Screw trimming, just model it in there initially.

[asid61]

Originally Posted by Dunngeon

All things equal, it would, but given how Omni's behave in a pushing match against another robot that isn't dead center. It would likely negate much of the gain that would be seen. (i.e. they break loose/slide in a direction away from the pusher, effectively rendering the friction gain useless)

Given the amount of time that would be put into building these (taking apart the wheels, rollers, testing,ect.), I think a team interested in Omni's could easily design something else that fits the bill better (Such as Butterfly) and be much happier with the results.

I think with the right wheel configuration there's a way to solve some of the pushing problems. I need to think on this one a bit.

The benefits of this is that it would be a direct swap out for a normal omni drive, so you would only need basic testing in the offseason. With butterfly and similar drives, there is a LOT more weight involved and it has to be custom built for each robot. Plus, you don't get a lot of push power in mecanum mode. So IF (big if) a modded omni could be omproved it could offer an advantage over butterfly drives.

[Dunngeon]

Originally Posted by asid61

I think with the right wheel configuration there's a way to solve some of the pushing problems. I need to think on this one a bit.

The benefits of this is that it would be a direct swap out for a normal omni drive, so you would only need basic testing in the offseason. With butterfly and similar drives, there is a LOT more weight involved and it has to be custom built for each robot. Plus, you don't get a lot of push power in mecanum mode. So IF (big if) a modded omni could be omproved it could offer an advantage over butterfly drives.

Butterfly is Omni, Octo-canum is with Mechanum.

I see the gains to be made here, I just don't think post-manufacturing modification is a very efficient route to go in this application. I'm also not convinced it would be very helpful just because of how the Omni's react to lateral loads.

[asid61]

Originally Posted by Dunngeon

Butterfly is Omni, Octo-canum is with Mechanum.

I see the gains to be made here, I just don't think post-manufacturing modification is a very efficient route to go in this application. I'm also not convinced it would be very helpful just because of how the Omni's react to lateral loads.

You're assuming that you are configuring the omnis in a traditional manner. I think ther emight be a way to get them all nice and on the ground (maybe with thingy-which-name-escapes-me with springs)
Post-manufacture, I agree, is somewhat dead because the rollers can only be modified so much. Swapping out the rollers with polyethurane with a delrin hub might work if you were to modify an existing wheel.
I've never assembled an omni before; is it possible to take the rollers off?