pic: Mecanum Wheel Project - Rev. D

[Madison]

loading photo ...

[SirLancelot]

That's absolutely gorgeous--
a few questions, though,
-will the hub be machined aluminum or some sort of plastic?
-what do you think the (approximate) weight will be per wheel?
thanks
Anna

[Tristan Lall]

Originally Posted by SirLancelot

-will the hub be machined aluminum or some sort of plastic?

With geometry like that, it's got to be a casting or a moulded part—note the sharp corners on the internal features, and the thin-walled sections.

[Andrew Blair]

One quick question: why the decision to use Phillips head? Will it be easier for personnel to change rollers?

Looks really good though. How big is it?

[tiffany34990]

Who is making the rollers for you all?

Just wondering...

Looks really great though!
Good luck on this project!

[David55]

Sorry for the lack of knowledge...but what 3D modulation software was used to make these?

[Joe Johnson]

As you go forward, keep in mind the ratio of axle radius to wheel radius.

This ratio is important. It is what makes rolling better than sliding. As the ratio gets closer to 1, the advantage of rolling disappears. Yes the wheel may be turning but the reduction in friction is almost nil.

Think of it this way. If you lock the wheel and slide it across the carpet, the torque about the axle of the force from the carpet would be:

Tsliding = MUcarpet2Wheel X FonCarpet X Rwheel

If you let the wheel rotate the torque on the axle becomes:

Trolling = MUwheel2Axle X FonCarpet X Raxle

There are only two places to let Trolling << Tsliding

#1 you can MUwheel2Axle << MUcarpet2Wheel

This is possible with bearings, etc. but in a Mecanum wheel application there is not much room for bearings and you sort of want MUcarpet2Wheel to be high because it is how your robot drives around. That pretty much means you're stuck with what you get for MUwheel2Axle. The roller material probably slides better on steel than carpet, but it is not going to be like butter on a bald monkey* either ;-)

#2 you can Raxle << Rwheel
This is your best bet in a Mecanum wheel application but even this option is somewhat limited by geometry.

Bottom line: The nice radius tips of your CAD model are beautiful and will provide nice transitions from 1 roller to another, especially on somewhat uneven ground, but it when that part of the roller is in contact with the ground, it is going to act more like a faceted wheel and less like an omniwheel.

Joe J.

*from Dance of the Cucumber
...
Larry: "sus suaves movimientos"
Bob: "Oh, how smooth his motion"
Larry: "tal como mantequilla"
Bob: "like butter"
Larry: "en un chango pelon."
Bob: "on a ... bald monkey."
...

[Madison]

It's exciting to see some people thinking in the summer

-will the hub be machined aluminum or some sort of plastic?
-what do you think the (approximate) weight will be per wheel?

- The hub is first being printed on a ZCorp Spectrum Z510 3D printing system, as it is a prototype. If there are no improvements needed, which certainly won't be the case, we'll likely be making the final hubs in an injection molding process.

- I haven't the slightest idea what they'll weigh, honestly. This isn't for FIRST use, so it's not a large concern.

One quick question: why the decision to use Phillips head? Will it be easier for personnel to change rollers?

Looks really good though. How big is it?

- The use of Phillips head isn't so much an intentional design choice as it is a matter of convenience. In some respect, I did consider that it'd be easier to replace the rollers if they were held in place by a common fastener, but it may turn out that we don't want them to be removable at all. Of all aspects of this design, the method by which the rollers attach is the least developed.

- The effective overall diameter of the wheel is 2.5".

Who is making the rollers for you all?

The prototyping is being done by someone that I met when I first moved to Seattle who was involved in SWAT Robotics - 824, a defunct team. He happened to mention to me that he was starting a prototyping business at about the same time I needed to start getting quotes for this wheel.

Sorry for the lack of knowledge...but what 3D modulation software was used to make these?

The models were made in Solidworks 2005 and rendered using its internal Photoworks renderer.

As you go forward, keep in mind the ratio of axle radius to wheel radius.

Are you referring specifically to the rollers, Joe? I presume you are, since the main hub is driven by a hex shaft.

The roller axle is currently planned as a 1/4" OD nylon tube and the rollers are to be urethane. As I mentioned above, this is the least developed part of the project at the moment and will likely pass through several more iterations before we decide on something that works effectively while remaining cheap to produce and easy to assemble.

Already, on the following revision, I cut back on the roller length and eliminated much of the taper that is shown above. That can be seen a bit more clearly on the image I posted of that revision found here: http://www.chiefdelphi.com/media/photos/25159?

That revision maintains the 1/4" OD nylon tube, but significantly increases the minimum roller radius.

Also, as a matter of full disclosure, I wanted to again reiterate that these wheels are not for FIRST use at the moment and will, instead, be used in some 'commercial' application. In other words, I'm getting paid to design this and, while I very much appreciate the criticism and advice, please be aware that you're doing some of my job for me. I know that some people might not want to do that

[Joe Johnson]

Originally Posted by M. Krass

Also, as a matter of full disclosure, I wanted to again reiterate that these wheels are not for FIRST use at the moment and will, instead, be used in some 'commercial' application. In other words, I'm getting paid to design this and, while I very much appreciate the criticism and advice, please be aware that you're doing some of my job for me. I know that some people might not want to do that

No conflict at all as far as I am concerned.

I think that the mission of FIRST is servered when the ChiefDelphi.com crowd get to learn that people get paid to work on cool things like mechanum wheels!

The only reason I used a FIRST application as a point of reference is that I am too busy working on my own cool robot project to read your message too carefully. Back to work for me...

Joe J.

[negfrequency]

people like you make me really happy. laying my eyes on that omni thats been modeled so well. Its like an art, which is probably why i love it so much. That is one of the better models ive seen on delphi recently. Id like to see some other parts youve done if you have anything else fancy lying around.

i dont know if you saw my previous challenge in the CAD forum, where i challenged people to design a soccer ball or a golf ball, 2 of the hardest objects to model. id like to see you try it. It looks like youve got the CAD thing down pretty well

[Madison]

I don't know if I have anything particularly fancy lying about and, unfortunately, I can't really share a lot of what I do these days.

I received a picture of the prototype parts sitting in the printer last night, so I wanted to attach it here so people who aren't familiar with the rapid prototyping process can take a look at what these machines do.

The picture includes some parts that don't belong to me. You can see two hubs and two sets of rollers for the wheels between the two largest pieces.