Mecanum

Hey,

What are the advantages and disadvantages of Mecanum drive trains?

Do you think they were successful this year? In other years? When would you use them?

Can they get up the ramp easily, push, etc.? Are they reliable?

Where they better than having a lazy susan?

Thanks, I’m looking forward to everyone’s replies.

Look up team 357.
Look for white papers on drive train designs.
What do you mean by a “lazy susan”?

Consult many other threads in the forums concerning omni-diretional mobility systems such as mecanum, holonomic, and swerve drives.
In short though, Mecanum drives use wheels with angled rollers to divert some of the wheels velocity sideways. Because of this (a 45deg mecanum wheel) they are roughly about 70% as efficient as a scrub steering system with identical geartrains.
If all 4 wheels remain in contact with the ground, they are usually reliable. Many teams (such as 40) have created active suspensions to keep their wheels in contact with the ground.
Mecanum drives were able to climb the ramp on a number of bots in 2006, such as 357 and 40.
Bots with mecanum drives had varying levels of success, but generally did fairly well (as mecanum drives are also usually built by veteran teams with a great deal of expierience anyway).
Check out what teams 40, 357, 488, and 868 did.

The main advantage of a mecanum drive-train is maneuverability, hands down. Mecanums (and swerve/crab drives to an extent) allow you to literally go in an infinite number of directions, a feat that a standard drive-train would not be able to accomplish.

As for if they were a success, I would say certainly yes. I know my team did very well this year winning not only technical awards but also being a finalist at the Boston regional. Our main advantage was that the drive-train allowed us to have many extremely different autonomous modes. We could go over the forward, forward then over, at an angle or even swing in an arc around the target. Our robot is able to get up the ramp unassisted despite the mecanum wheels. We added some grooves to them to add traction and changed the gearing from the transmissions which allows us to climb with relative ease. As far as pushing is concerned, they do reasonably well. Our robot was primarily designed for offense, but in the few occasions where it was in a pushing match (front to front mind you), it did hold its ground if not hold the other team back.

However, the one flaw with mecanums is that they are almost too mobile. Our robot is easier to move from side to side then our cart is, which should tell you something about how easy it is to push if hit from a corner or the side. But, as a mecanum robot, your main job is to move away from those situations instead of to try and hold your ground, because it is nearly impossible.

As for when to use them, it is best to use them when there is a wide-open field, as to put their maneuverability to best use. If mecanums were used in 2004, no one would really be able to get the full potential out of their robot because of the large center structure. But, this year’s and last year’s games were both almost completely (if not totally) clear on the floor, making for a perfect mecanum environment.

I hope this helps, and feel free to ask any more questions!

I’d actually say the best situation for omni-directional drivetrains is the exact opposite, when the field is crowded and full of difficult (or impossible) to bypass obstacles. It allows the robot to move easily and rapidly around these obstacles with the least amount of time spent repositioning and turning in place. This parallels where these drivetrains are most often used in the workplace, crowded areas. Mecanum drive forklifts are often used to maneuver around crowded warehouses and ships.

actually, our veterans had trouble with it… we wanted to give him a single joystick… but he made us give him two (one for forward, rotation; another for side to side… it defeated the purpose of having a 3 axis joystick)

driving a mecanum is kinda weird… because it can go in any direction, you need to think where is it going next. giving demonstrations is fun, especially for inexperienced drivers… running into people, walls, etc.

The friction coefficient (i believe) is the cosine of the angle of the rollers… rollers at a 45 degree angle will have around 71% traction

Use mecanum wheels when the game requires alot of movement, not pushing.
(but more pushing than what the “basic” omnis would require)

lazy susan as in crab? each has its own unique strengths… if you’re seriously considering either, your best bet is to get started designing now (having the wheels available has a much better advantage than starting from scratch during the build season) Team 868 spent 4 / 6 weeks just getting the wheels done… But i’m sure thats not an issue now, as teams will likely just buy AM’s

Last but not least,
As for coding the drive (keeping the hard working programmers in mind) If you have a decent symmetry to your wheels, then the code can be brought down to 4 lines long… otherwise, it’ll be a nightmare. Also, with the motors, make sure each is lined up in the same configuration, if not, then they deadband at different values, making for weird movements going slow

(yeah, 868 spent alot of time discussing this)

so would anyone actually endorce (i know encouragement wouldn’t nesisarily happen) the use of AM mecanums on a rookiebot? Personally I think it wouldn’t take any more time mechanically than puting togeather IFI traction wheels, and my programers could probably get the programing done in the amount of time it would take to assymble the wheels.

Though I agree that Mecanum are great for maneuverability, you seriously lose traction and power on your train if not done right. The main problem is that it often takes tries upon tries to get it done right, and by then the question is…is it really worth it?

Personally, I’d rather go with standard Skyways in a tank configuration, which gives you nice turning --with tank drive-- and truly awesome traction and overall pushing power.

So in the end, it depends on your goal for the game. I remember some robots trying to shoot with Mecanum drives, but would be pushed around like rag-dolls, often rendering their shooter mechanism useless. It’s a matter of combining form, factor, and efficiency to meet the demands of the game as you see them. If you want incredible maneuverability, often be prepared to pay for it with a loss in traction and pushing power (in comparison to other drives).

Advantages - Extreme maneuverability

Disadvantages - if someone tries to move you, they will own you

Not all mecanum wheels can be pushed easily. it depends on the material and way you make the wheel. team 357 used this drive train this year. we got a lot of pushing power ot of our wheels. mecanum wheels are not omni wheels. this year we were able to get up the ramp perfectly with this drive train. they are a very relible drive trian. i would use this in any year that Mobility is needed or can be used. last year was a great year i think to use them because if you weren’t in the spot u need instead of backing up and going back in, you could move in that direction.

Agreed. One feature about the AM Mecanum wheels is interesting: each roller can be adjusted to spin very little or very loosely.

Each roller is on it’s own shaft, the shaft being a #10-32 screw. One end of the #10-32 screw is a nylock nut. If this nut is tightened completely, the roller does not move and the wheel acts like a grippy robot wheel (the rollers won’t spin). If the nut is loosened, then the rollers will spin and the features of the wheel are utilized.

This tensioning could be adjusted between matches at a FIRST Competition. You could go from a fairly good pushing robot in one match (back wheels w/ rollers tight, front wheels with rollers mostly tight) to a strafing robot in the next match (rollers loose for full mobility). 1 student adjusting 12 nuts on 1 wheel would take about 3-4 minutes.

It should be interested to see how these wheels are used in the upcoming season.

Andy B.

Andy, I like your idea of having variable preload on your rollers. Have you done any actual “bolt torque vs. traction” tests with these wheels?

Also, I read through the assembly instructions for your wheels, which is very helpful. However, I’ve found that one student’s idea of “finger tight” varies greatly, so much that finger tight can mean finger loose, or opps… I stripped all the treads. Do you have any more specific guidelines for the bolt torque?

As a comment to rookie teams, the AM wheels and kit gearboxes and chassis can allow you to quickly build up this system mechanically. However, purchasing two extra gearboxes and the set of wheels is a pretty good chunk of change. Also, I believe the programming and feedback required for this system is also a challenge. I certainly encourage all teams to improve upon their previous designs and challenge themselves. But please, try to understand the limitations of your team, and the resources that you have.

And, as a side note, GUS will be entering its 9th year in FIRST this season. With all the experience we’ve had, and iterations we’ve gone through, we’ve still been beaten by our fair share of rookies using the basic kitbot.

BEN

We have not done quantifiable tests on how much traction these wheels get with variable preload. This does need to be done, but I just have not had time yet.

As for the “finger tight” instructions, those same nuts are to be tightened with tools later on in the assembly. At the “finger tight” stage, the intent is to only hold the screw in place until the next assembly step is complete. Also… “finger tight” on a nylock nut is when the nylon engages the screw (this is described in step 7).

At this point, we don’t have more bolt torque requirements. It really depends on the application and the requirements of what the drive chassis is needed to do. Loose rollers will allow maximum mobility and tight rollers will allow the transfer or more pushing force. FIRST teams will be happy with a range of movement between these two extremes.

Andy B.

That’s a good point Ben, the basic kit-bot is nothing to snear at… and so what if you can’t move sideways… if you can turn quickly isn’t that just as bout as good?