Hello! I am the chief designer of a 2nd year robotics team (7886) and I would like to design a mecanum wheel drive for our team during this off-season (which I plan on shrinking down and prototyping on my own). I have studied up on the basic drivetrains and considerations, but how should I actually go about designing something like this?
1 Like
Sorry that this doesn’t really answer the question, but since we already know the game for 2021 I wouldn’t recommend a mecanum drivetrain, since they aren’t good for this game
A big thing is you first need to look at what manufacturing resources your team has access to, whether it be in your own space, or by a local sponsor.
If your machining resources are currently limited, I highly recommend looking at the versa ecosystem from Vexpro. They make things really simple. I believe most designs could be put together with a robot gun, a drill, and a hole saw.
Here is an example chassis from them
If you have more machining resources, I think it’s a great learning experience to work to design within those resources in mind. The KOP chassis is a great example of how to design within sheet metal. If you have access to a good mill and lathe, there are numerous examples on cheif Delphi on different West Coast drives made out of 2x1 tubing. Many of the same principles apply with a Meccanum drive.
As far as “how to design” I recommend using a CAD software. If your team has already decided on a CAD software, that’s probably a good start. If not, I am personally a huge fan of OnShape (as are many folks on this thread). It is free and easy to get started for students. Hope that helps a little. Feel free to ask more specific questions if you get stuck on your design endeavours.
1 Like
We are using a mecanum for the first time. By no means did we start from scratch, and we have not played the game (both our events were cancelled). I could share an onshape of our drivebase if you would like. We used the kit drivebase with four toughbox micros.
Though there are some things about this game that could cause problems for mecanum, our team has done extensive deliberation, and determined that for us, this is the best way forward.
I urge people to listen to the OP. I feel there are a number of threads about whether mecanum is worth it, this one is specifically about how to do it. Please be gratious and professional.
P.S. I will look up our gearing.
1 Like
OP:
What motors are you planning to use?
1 per wheel or 2?
What size wheels?
I will get our drivetrain shortly, but we made a modified version of the Andymark Kit 6" mecanum upgrade. It should be noted that I think the kit assumes you already have the standard FRC Drivebase kit. We did not buy that, just the gearboxes, gears, and mounting hardware (we also bought a plate and some gears to use the kit toughbox mini for our climber).
We used one Neo per wheel and a 10.71:1 gear ratio. Again, we did not get a chance to compete but did test it going over the metal barriers, and it had no problem. We were (still are) considering changing out the 6-inch wheels for 8, and/ or looking at different gearing, but did not get that far. We wanted to be as low as possible because we do have a trench bot (that will be able to spring the wheel of fortune). I will update with some CADD, but it does not deviate much from the stock other than the fact it is a U frame, not a square. If you are considering a U or H frame design, remember that the gearboxes (of which, you will now have four) do stick into the frame more than you probably realize. This is not insurmountable, but it is a potential design challenge and another place where the Neos help.
Edit: Here is a link to our Drivebase. We are just learning OnShape so some of the pieces may just be dropped in place, but it should give you the idea. We called it basic because we have an idea for a butterfly drive too. We have neither drawn it out nor built it yet. There may be time now though 
.
1 Like
4 Likes
I was planning on using 6 wheels, 4 mecanum paired with 2 omni wheels in the middle. Each mecanum wheel would get a motor, so 4 motors. The wheels themselves would be 4 inches.
What motors were you planning on?
(You listed the wheel size, thanks)
I was heavily considering CIMs with CTRE mags as encoders (I am assuming mecanum drives benefit from econders but I could be wrong).
Pretty much any drivetrain can benefit from encoders, mecanums especially. Generally you’d put the encoders near the gearbox output but that can be worked around.
Question: Why would you put omni wheels in the middle? I’m not saying it’s a bad idea or won’t work, but it’s unconventional. Just trying to get a sense of your thinking on that.
From what I learned, the omni wheels need to be oreinted in such a way that the rollers form a square or x-shape in order to properly function. However, since mecanums wouldn’t work in a 6 wheel configuration, I though that placing omni wheels in the center would allow the drivetrain to gain stability without sacrificing the agility of the mecanum wheels.
I meant mecanum in the first sentence.
In reality, it won’t have much effect on stability. You’re just adding contact points within the existing contact box. The same goes for adding more wheels to a 4WD, 6WD, or 8WD: as far as stability goes, if you don’t do any odd tricks (like dropping wheels), it’s a wash. If you add wheels outside that box, stability does change.
For a mecanum drivetrain, the optimal configuration for stability (and for that matter, ease of control) is a square frame at near the maximum frame perimeter, with the FLOOR side of the mecanums showing an X (top will form an O).
My team has been using a simulation model that I have previously posted on this site.
We examine three characteristics, small movement duration, initial velocity capability, and velocity at range. These three characteristics are selected on a game by game basis and are then used to trade between motor and gearbox options.
As far as stock gearboxes - assuming andymark products - something with a 5.94:1 will provide a reasonable compromise for your project. I believe that you can experiment with the gearing and get ratios on either side of this. For example I believe that you can reach a 5.35:1 ratio. I suspect you can identify a range of options with this method, that are obtainable using existing gears.
CIMs with a ~50 Amp current limit should produce:
| 15.001 | ft/sec | at | 2.218 | seconds | with | 25.000 | foot range |
|---|---|---|---|---|---|---|---|
| 10.000 | ft/sec | at | 0.631 | seconds | and | 3.439 | foot range |
| 8.936 | ft/sec | at | 2.500 | feet | and | 0.533 | sec |
If you are willing to consider NEOs with a ~35 Amp current limit, I expect:
| 16.295 | ft/sec | at | 2.087 | seconds | with | 25.000 | foot range |
|---|---|---|---|---|---|---|---|
| 10.000 | ft/sec | at | 0.573 | seconds | and | 2.914 | foot range |
| 9.395 | ft/sec | at | 2.500 | feet | and | 0.530 | sec |
These predictions do not asses any implications of the mecanum wheels themselves; these are straight line results only.
2 Likes
So essentially something along the lines of this would work?
1 Like
Yep.
That said: There have been some odder mecanum drivetrains out there. Someone in the PNW did an 8-wheel a few years back, Stronghold I think: 4 pairs of mecanums. And someone in MI (1322) did a 6-wheel like you were suggesting, BUT they made it an articulating setup to handle obstacles and their omni was driven by a ball differential from both sides. In both cases there was a reason to do something different.
You won’t be able to spin like that; the diamond needs to be on the floor.
I keep getting that setup confused…
It’s counter-intuitive to me but does make sense if I think about it enough.
1986 used 8 mecanum’s in 2017.
2 Likes