Mecanums: Let's Settle This

[The VEX Raptors]

There's been a lot of arguing around here about whether the rollers on Mecanums spin and cause a loss of energy when they're driving forward. Has anyone tested the mecanums out in great detail to effectively confirm or deny this?

If not, I have an experiment to propose, so see whether the rollers are spinning during forward or backward motion on a normal setup:
1. Build or obtain a mecanum drivetrain.
2. Put a small, white mark (maybe a piece of teflon tape) on one of the rollers.
3. Drive the robot forward, and see if the white mark moves. If it does, the rollers roller is spinning; if it doesn't, then the roller is not.

I want to see this mecanum argument settled once and for all.

[Telemascope]

I hate to say it, but isn't this obvious? The wheel just acts as a normal wheel. The rollers have no reason to roll. There is no way the rollers could roll while the wheel is spinning that would result in only forward movement (I think). Also, the method of the tape probably won't work properly, because as the rollers leave the ground, and make contact with the ground, they are probably spun a little.

EDIT: Actually, to test this, simply put a wheel on a shaft and spin it. Lower it onto the ground. You should be able to see that the wheel moves back and forwards normally, and any movement of the rollers results in sidewards movement.

[magicode]

The arguing has already happened. What The VEX Raptors wants to see is an empirical proof video with a VEX mecanum wheel drive.

[Team80_Giraffes]

Would you want on the wheel or the side of the wheel/cone to be marked? Would a marker work just as well and not affect the wheel's interaction with the ground?

What about using a go-cam mounted off to the side to look at the wheel as it spun?

Experiment 1 - forward drive
Expeirment 2 - 20 degree right drive
Experiment 3 - 45 degree right drive
Experiment 4 - 90 degree right drive
and so on and so on. Maybe put an LCD screen showing the experiment number to help

[The VEX Raptors]

Originally Posted by Team80_Giraffes

Would you want on the wheel or the side of the wheel/cone to be marked? Would a marker work just as well and not affect the wheel's interaction with the ground?

What about using a go-cam mounted off to the side to look at the wheel as it spun?

Experiment 1 - forward drive
Expeirment 2 - 20 degree right drive
Experiment 3 - 45 degree right drive
Experiment 4 - 90 degree right drive
and so on and so on. Maybe put an LCD screen showing the experiment number to help

I'm just thinking a small piece of teflon tape on one of the green rollers, on a part of the roller that's not touching the ground. This will show visual proof of whether or not the roller is spinning.

My team might try this experiment at our meeting tomorrow. If we do, I'll post the results here.

@Telemascope: You could test it with one wheel, but I've tried spinning one wheel alone, and it has a tendency to move diagonally on its own. With a full drive, this diagonal motion is canceled by the other wheels. I don't believe myself that the rollers spin, but theory and practice are completely different. I want empirical proof.

[GregH]

Originally Posted by The VEX Raptors

I'm just thinking a small piece of teflon tape on one of the green rollers, on a part of the roller that's not touching the ground. This will show visual proof of whether or not the roller is spinning.

My team might try this experiment at our meeting tomorrow. If we do, I'll post the results here.

@Telemascope: You could test it with one wheel, but I've tried spinning one wheel alone, and it has a tendency to move diagonally on its own. With a full drive, this diagonal motion is canceled by the other wheels. I don't believe myself that the rollers spin, but theory and practice are completely different. I want empirical proof.

In theory the rollers should spin. When Mecanum wheels roll, there is a force exerted on each roller that has a component not parallel to the axis of the roller so they should spin. It will be interesting to see if they really do.

[chris@aura]

In theory the rollers should spin. When Mecanum wheels roll, there is a force exerted on each roller that has a component not parallel to the axis of the roller so they should spin. It will be interesting to see if they really do.

There is indeed a force on each roller that has a component not parallel to the axis of the roller. However, you forget that on the other side of the robot there is another wheel with rollers at 90 degrees to these rollers. And hence the component of force that is not supported by one roller will be supported by the one on the other side. So the rollers do not spin at all (on an ideal surface with no slippage).

[fretless_kb]

I don't have a video, but I did a small experiment with our prototype mechanum robot. The 'rollers' sometimes have markings or seams. I aligned the seam in a obvious orientation and rolled the robot manually a couple revs, each time the roller I had noted came around the roller was in the same orientation. Not as good as a real drive test, but put me on the side of they don't roll.
Cheers Kb

[Taylor]

While it does not directly address the OP, I believe this video does answer some questions and misconceptions about mecanum wheels.

[The VEX Raptors]

Originally Posted by Taylor

While it does not directly address the OP, I believe this video does answer some questions and misconceptions about mecanum wheels.

I saw that video, and I liked it. It was very useful in convincing a teammate of mine that we should use mecanums.