Converting to Mechanum....

[ZipTie3182]

Our team finally is deciding to try mecanum over the summer after much pushing on my part. We were rookies last year. Now, our drive right now is very simple. 4 wheels, two motors, two gearboxes. We used the brackets from the KOP so the bottom of our chassis sits above the wheels. So, my question is, without having to completely rebuild our chassis, how can we convert this to mecanum, because all the pictures on andymark for the drivetrains show the wheels being more IN the chassis if that makes sense...I'm wondering if its possible to use what we have. Also, we're looking to do chain driven mecanum if that makes a difference.

This is sorta what ours looks like (except not with c channel): http://ep.yimg.com/ca/I/yhst-3383317...7_2119_2469960

-Anna

[Taylor]

You're going to need two more gearboxes and two more motors. In order for a mecanum drive to do all it can, each wheel needs to be independently driven. The way the robot moves depends on how the wheels work with/against each other. There are many great white papers in the CD-Media section; search "mecanum" and you'll find some nice resources.
Mechanically, your chassis will be the same. Simply place the gearboxes directly above the wheel pods, and run a chain from the gearbox sprocket to the wheel sprocket.
Good luck!

[JesseK]

Theoretically it's a simple movement of the existing gearboxes to make room for a second gearbox on each side. The towers themselves will stand up to the torsion the Mecanum wheels would produce (at least outside of competition), so a direct wheel swap would be fine. It looks like you would have to purchase (based upon what you've given us):

• 1 Mecanum wheel set
• 2 AM Toughbox
• 2 AM Toughbox chassis braces
• 2 CIM motors

For center of gravity on a competition field, you would want to put the wheels more inline with the chassis; yet for offseason prototyping this will not skew your results much.

[Chris is me]

If I remember your frame correctly, I would suggest you first try and build the kit frame instead of using your old frame. It doesn't take very long to assemble it. Unless you want your mecanum drive to compete in an event, I would skip using the wheel brackets and just mount the wheels to axles going through the c-channel frame. This will get you a simple and rigid platform for testing and stuff.

From there, converting to a mecanum drive would require for you guys, two more gearboxes, two more CIM motors, and a bit of modification. Move the gearboxes you have a bit closer to the wheels, and mount two more basically the same way as the other two were. From there you should be able to get rolling and you'll have a base your programmers can use.

I think if you guys plan to compete with the bot at an off season, a better investment would be in two more CIM motors in your existing gearboxes (this increases your acceleration and your ability to push other robots) and some omni wheels. The omni wheels will allow your robot to turn a lot better than it did at CT, and you won't have to reprogram your whole robot. It's also a LOT cheaper than a mecanum base, though way less cool I know

(I would personally not advise using a mecanum drive in a competition robot)

[Jon Stratis]

As others have said, you need independent drive of each wheel - in other words, 4 victors/jaguars, 4 motors, and 4 gear boxes, all connected to 4 wheels.

The KoP wheel standoffs are fine to use with Mecanum wheels. We used them on our practice bot with no problem (we did build our own for the "real" robot, it was a good experience, and gave us better mounting points for some additional things like ball guards). You can clearly see the Mecanum wheels mounted here, and if you look closely here, you can see a second toughbox in the background for the second wheel on that side.

The only reason Anymark has pictures up that show the Mecanums more in line with the frame is because that's been by far the most common configuration over the years - very few games, if any, have required such a large offset between the ground and frame. If you had wanted to use Mecanums in previous years, that's probably how you would have done it, since you wouldn't have been worried about a high bumper height requirement, the bumps in the field, etc.

[Hawiian Cadder]

for testing purposes only, my team added plywood to the frame to make a deck, to make it more rigid. this increased strength and because it was not for competition we didn't care about the weight.

[Ether]

Originally Posted by ZipTie3182

This is sorta what ours looks like (except not with c channel): http://ep.yimg.com/ca/I/yhst-3383317...7_2119_2469960

Mecanum wheels will "work" with the KoP standoffs, but there's some evidence to suggest that the lack of stiffness in the axial direction with mecanum wheels cantilevered at the end of a "stalk" can contribute to wheel chatter. Were you planning on making or buying the wheels ? Either way, make sure the rollers don't have high bearing friction when under load. Make sure the wheels are either protected from impact, or are tough enough to take a hit without deforming and binding up the rollers. Also, consider using closed-loop wheel speed control for better fine control of the robot with mecanum wheels. With mecanum wheels, it is important that all four wheels are sharing the weight of the vehicle because each of the four wheels must generate forces to get the desired vehicle motion. So I suppose it is possible for your frame to be too stiff.

~

[JamesCH95]

You might have to modify the wheel brackets to accept mecanum wheels, on our prototype chassis in 2010 we had to slice open the wheel stalks to fit 8" AM mecanum wheels.

[Ether]

Originally Posted by JamesCH95

You might have to modify the wheel brackets to accept mecanum wheels, on our prototype chassis in 2010 we had to slice open the wheel stalks to fit 8" AM mecanum wheels.

Yes, and that exacerbates the cantilever/stiffness concern.

~

[Gdeaver]

Good, potentially another mecanum wheeled robot to push and smash around. Our team so loves to torment them.

[Ether]

Originally Posted by Gdeaver

Good, potentially another mecanum wheeled robot to push and smash around. Our team so loves to torment them.

They will have the opportunity to learn a lot of engineering and physics that other students on pushing and smashing and tormenting teams miss out on.

If next year's game isn't about pushing and smashing and tormenting they may even beat you.


~

[,4lex S.]

Want a simple test frame, and don't want to take apart your robot too much? Use 3/4" ply and mount your wheels via pieces of channel into cut outs. You can just screw down the toughboxes with the standard mounting hardware. The bonus of plywood is that it will help keep even weight distribution on all your wheels because of its flexibility.

I think everybody else covered the tech requirements, but make sure you have someone willing to program it, or it will be a pain to deal with. I am sure you won't be convinced until you have built it yourself (because I was the same way ), but a 6/8WD done after some research on CD and careful design is the best drive train for FRC (you can always make modifications to make it work better for the game at hand).

[ttldomination]

Originally Posted by Gdeaver

Good, potentially another mecanum wheeled robot to push and smash around. Our team so loves to torment them.

Mecanum wheeled bots are easy to push around, but with the right mind set, mecanum wheeled robots can be huge pains when playing defense, and just out right speed demons when playing offense.

Originally Posted by Ether

If next year's game isn't about pushing and smashing and tormenting they may even beat you.

I for one believe that the 6/8 WD is still the way to go. IMO, the teams that went with 8WD this year (254, 1114, 1771 to name a few) were able to move around much easier than most mecanum wheeled bots. But of course those teams do a thousand things right in order to get their DT in perfect condition.

[Ether]

Originally Posted by ,4lex S.

make sure you have someone willing to program it, or it will be a pain to deal with.

programming mecanum is not difficult.

2-axis double-joystick "tank drive with spin" and 3-axis single-joystick algorithms are available here:

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

~

[,4lex S.]

Ether,

I was aware of your whitepaper, and I intend to make my way through at some point. The real challenge I see here relates to my experience with my own team, without a skilled programming mentor. You must have a passionate and supported programming squad to attempt anything, let alone something that is outside of the basic drive code, and I do not know the situation of the team at hand. In 2008 I (as a mechanical student) asked my team to work with a holonomic platform, and it did not progress as harmoniously as we would have liked due to the limitations on our programming knowledge. The key is that this kind of work is more multidisciplinary than some other ways of attacking the same problem, so the potential programming hiccups have to be weighted against mechanical advantages (if issues are likely).

The only reason I say there are likely to be programming issues, is because I have been there. Doing this in the off season is probably a really good choice, because the wrinkles will be out by the time the season rolls around.