Mecanum wheels

[JesseK]

Originally Posted by lcoreyl

While as quoted these statements can both be true...

Please don't mis-quote me and then poorly attempt to make a counter-point which actually says something completely tangential to what the entire quote said.

[thefro526]

There's a bit of mis-information in this thread, and while I don't want to join in the debate, I'd like to set the record straight for any of the newer teams that might be using this thread for information.

Mecanum Wheels can be purchased from two different FRC Suppliers.

- AndyMark http://www.andymark.com/Mecanum-s/53.htm

AndyMark's Mecanums were first released for the 2007 Season and have been evolving since then. They're offered in FRC relevant sizes of 6" and 8" with a Standard Duty and Heavy Duty version of each. The Pricing Varies from ~$250 for a set of 6" Standard Duty to in excess of $400 for a set of 8" Standard Duty.

- VexPro http://www.vexrobotics.com/vexpro/wh...um-wheels.html

VexPro's Mecanum Wheels are relatively new compared to AM's Mecanum Wheels, but have proven themselves to hold up reasonably well in the FRC Environment. They're a bit Cheaper than AM's Mecanum Wheels and rely on a primarily plastic construction in place of Metal.

As far as Mecanum's implementation in FRC, they're a decent way to make your drivetrain more maneuverable, as well as adding omni directional functionality within a traditional frame. The increased agility awarded by Mecanum wheels should allow you to move around traditional defense to the point where 'pushing' isn't necessarily required.

All of that being said, when comparing two similar drives, with the only difference being that one Uses Mecanum wheels in place of a traditional wheel, the Mecanum Drive will almost always have less traction than the traditional drive train. This is due in part to a couple of factors, the wheels always being in a constant state of slip (in that they're applying force in a non-fore/aft direction) being one of them, and the other being that each wheel's motor is independent. This means that if one of your wheels leaves the floor, or starts spinning, you're not only putting down the power of (3) motors.

[TedG]

We have used Mecanum wheels the last 3 years with great success.
Great for manouvering in all directions without haveing to turn.
Last year we used the 6" dia heavy ones from AM and could push our way around as well as resist being pushed around. It helps if you have a heavy robot though.
And it's a good idea to buy a spare wheel or two, but they are expensive.

[lcoreyl]

Originally Posted by thefro526

There's a bit of mis-information in this thread

yes... yes there is...

Originally Posted by thefro526

Mecanum Drive will almost always have less traction than the traditional drive train. This is due in part to a couple of factors, the wheels always being in a constant state of slip (in that they're applying force in a non-fore/aft direction) being one of them

I've linked a paper in my signature. In the paper is a link to a video showing the worms eye view of the contact patch of a mecanum wheel while being moved as it would in normal forward motion. The contact point doesn't slip. The rollers don't spin.

[thefro526]

Originally Posted by lcoreyl

yes... yes there is...



I've linked a paper in my signature. In the paper is a link to a video showing the worms eye view of the contact patch of a mecanum wheel while being moved as it would in normal forward motion. The contact point doesn't slip. The rollers don't spin.

I don't want to debate this too much since it's going to derail the thread, but I'm still not convinced based on the two videos (Glass table, green backlighting) that the rollers do not spin under FRC conditions. Maybe I'm nit picking, but there seemed to be some sort of displacement of the roller (Looks like a few degrees at most) even on the table top test. At FRC speeds, something tells me that this has to add up into the rollers spinning while the robot is moving in standard fore and aft directions. It's been a few years since I've played with Mecanum wheels first hand, but if I remember correctly, rollered seemed to slip while under load.... But I might be wrong.

[Ether]

Originally Posted by thefro526

I'm still not convinced based on the two videos (Glass table, green backlighting) that the rollers do not spin under FRC conditions.

They do spin somewhat under FRC conditions. The roller axial free play in the affordable COTS mec wheels typically used for FRC contributes to spinning. A compliant surface (like carpet) contributes to spinning. This is mentioned on page5 of his paper.

[Keegbot]

Originally Posted by jbsmithtx

Try VEXpro. They have very affordable yet durable mecanum wheels. The other posters are correct in assuming that unless you are trying to catch, mecanums might not work well this year... However don't believe they are easy to push. We had to play defense with our mecanums last year in matches when our shooter failed and were fine. We could block lots of teams. I know this is contrary to popular belief, but a properly done mecanum drive base can in fact be very strong.

For anyone that wants to question our pushing abilites, we have video...

Our team used the VEXpro wheels last year, and I was very pleased with them. They lasted 2 competitions and multiple fundraisers until they started to fall apart. They are also quiet and light compared to the Andy Mark versions we have used in the past. We also had to be purely defensive in one competition and managed to place high because our mecanum drive allowed us to out-maneuver our opponents and get in front of them to block their shots.

[Caleb Sykes]

I encourage anyone and everyone who thinks that they understand mecanum physics to read the whitepaper given here: http://www.chiefdelphi.com/media/papers/2759?

[BEN35678]

Originally Posted by Keegbot

Our team used the VEXpro wheels last year, and I was very pleased with them. They lasted 2 competitions and multiple fundraisers until they started to fall apart. They are also quiet and light compared to the Andy Mark versions we have used in the past. We also had to be purely defensive in one competition and managed to place high because our mecanum drive allowed us to out-maneuver our opponents and get in front of them to block their shots.

Last year we used Andymark mecanum and this year we were looking at the vex pro mecanums, and a couple of our main mentors were worried about them cracking because they are plastic. Did you have anything that looked like it was about to crack or anything like that?

[TheJoe]

Our team has used a mecanum drive train for the last three years and we've had a lot of fun with them.

As the robot driver for the past two years, I can assure you they're wonderful to control, and driving becomes a lot easier with mecanum.

Controlling the a mecanum-powered robot is a lot like moving a character in a first person shooter video game, which I'd assume many students are familiar with already.

Couple a mecanum drive train with gyroscope technology and now your robot can always move relevant to the driver (No matter the orientation), which in my personal opinion makes controlling the robot even easier.

While a downfall may be that you lose a lot of pushing power, the versatility gained to out-maneuver the defense matches said downfall quite well.

[Donut]

Our VexPro mecanums have held up well. They survived one regional and some off season demos with no problems, and in our shop we typically run them on concrete. We are planning to use them again this year.

[purduephotog]

Could anyone point me to resources about the maximum speed you can run Mecanum wheels at ?

I'm wondering (and trying) to find out if all the gear boxing is really necessary. Once they're slipping they're slipping, and the friction factor for all four wheels would be ~equal.

[Ether]

Originally Posted by purduephotog

Once they're slipping they're slipping, and the friction factor for all four wheels would be ~equal.

What do you mean by "slipping" in this context?

[JesseK]

Can anyone speak to the wear of VEXPro 4" Mecanum rollers compared to the 6" wheel rollers over the course of multiple competitions?

In an average year we compete in 2-3 official and 2 off-season competitions, perform in several demos & run a lot of driver practice for the next year's team. So the wheels really need to last. Yet it'd be nice to know what spare parts to get ahead of time (i.e. 1 set of spare rollers per 1 event or 2 events?) when comparing wheels. If we have to replace the 4" wheel's rollers every 1 event, we may go the 6" route so we don't have rollers that are wearing thin at the end of an event.

My team's poking at the idea of Mecanum for a few reasons, and it's easy to implement on our current drive frame.

About Mecanum in general:
The only thing I can say without reading a metaphorical earful from Ether (ha, hi Ether) is that the default field-centric (or robot-centric with no deadband) control algorithm robs the drive train of power just by the nature of the code. When translating on the 45-degree relational diagonal (i.e. 0 degrees = full forward, 90 degrees = full right, so at 45 degrees the bot goes forward & right), 2 motors are completely turned off by the algorithm.

If we implemented Mecanum, we'd most likely do a dual-mode driver control. In the first mode, it'd be field-centric holonomic drive. In the 2nd mode, the driver would be limited to robot-centric tank drive with 90-degree strafing. The driver would be in the 2nd mode most of the time.

[Ether]

Hi Jesse :-)

Originally Posted by JesseK

When translating on the 45-degree relational diagonal (i.e. 0 degrees = full forward, 90 degrees = full right, so at 45 degrees the bot goes forward & right), 2 motors are completely turned off by the algorithm.

Yup. And so the available driving torque at the wheels to keep the bot moving against the rolling resistance is half. And so the top speed in that direction will be substantially less than that predicted on the basis of kinematics assuming no load.

In the 2nd mode, the driver would be limited to robot-centric tank drive with 90-degree strafing. The driver would be in the 2nd mode most of the time.

What driver interface do you have in mind for this? Perhaps divide this 2nd mode into 2 modes?

Mode 2a: Standard 2-joystick Y-axis-only tank drive (X-axes disabled)

Mode 2b: Right (or left) X-axis commands strafe (disable Y-axes)