pic: Practice Mecanum Chassis

[Madison]

Originally Posted by Jeff 888

what motors are you going to use?

Each of the four wheels will be driven by its own, small Chiaphua motor. I'm presuming the kit will remain unchanged in that respect next season.

What's the purpose of the large plates on either side of the wheel? Why not a much smaller system to mount the wheels and/or bearings?

They're so large so as to protect the wheels from impact some. Last season, our omniwheels were unprotected and took a few good hits from robots with higher chassis rails, causing them to bend. We were always able to bend them back, but they were also a lot cheaper and easier to reproduce so we had plenty of spares. They also look pretty cool.

Wow, great work. Thats an impressive model. I have one question though, do you plan on putting a suspension on you chassis? I can tell you from our experience with mecanum that constant contact is important. Our team, 40, used rubber bumpers on our wheel pods to keep contact. Also at nationals we saw another team (I'm terrible at remembering team numbers) who used small pnumatic pistons for a suspension.

Thanks. Do you have any photos of the arrangement you used? I've currently arranged for the large wheel guards that hold the axles to use vertical slots so that we can adjust their position relative to the frame. This would allow us to raise or lower the wheels individually to ensure contact in case the frame warps or bends, but it wouldn't do this dynamically during a match. How important do you think it is that the wheels sit on an active suspension?

The design I've made previous to this included a more substantial, pivoted subchassis for the two rear wheels that would dynamically stabilize things, but it seemed a bit like overkill. That's shown below:

It's an older model, so there're no chains on this one. They didn't disappear or anything.

[yongkimleng]

wow that looks cool! im also doing a vex assembly in solidworks.. but doubt it can turn out so nice...
where u ge the macanum wheel part frm? wonder anyone has omniwheel part for me to import :S

[artdutra04]

Originally Posted by yongkimleng

wow that looks cool! im also doing a vex assembly in solidworks.. but doubt it can turn out so nice...
where u ge the macanum wheel part frm? wonder anyone has omniwheel part for me to import :S

You can download the STEP model from this page: http://www.andymark.biz/mecanum.htm

[Crazy Ivan]

Thanks. Do you have any photos of the arrangement you used? I've currently arranged for the large wheel guards that hold the axles to use vertical slots so that we can adjust their position relative to the frame. This would allow us to raise or lower the wheels individually to ensure contact in case the frame warps or bends, but it wouldn't do this dynamically during a match. How important do you think it is that the wheels sit on an active suspension?

The slots are a good feature, but I would say that an active suspension is almost without a doubt nessisary. Some sort of ridged adjustment only works if the playing surface is compleatly level and smooth (unlike the floor of the Georgia dome that is full of warps). And even if all the wheels apear to be touching, just a small amount of slip with one wheel can cause you drive train to go funky. Even with our active suspension, it took a good hour to align and even then it only strafed strait in only the right hand direction, with a small arc to the left.

Unfortunatly we don't have any pictures of the wheel pods up close, but I'll give my best try to describing them. With both systems I mentioned before, one end of the wheel assembly must be on some sort of hinge or pivot (we did this with a removable steel pin). The other end of the wheel assembly was held by screws going through the ridge part of our frame, and threaded into the wheel assembly with cylinders of gum rubber inbetween the frame and wheel assembly. A similar system was used, only with 3 inch pistons at 45 degree angles.

The other method of suspention you can use (as I belive you alluded to earlyer) is the system team 190 used in 2005, where the wheels were ridged, but it was there frame that actually had one big pivot in the middle.

Hope this helps!

[Lil' Lavery]

116 found in 2005 that contact was incredibly important with our holonomic drive (I'd assume it would be the same with a mecanum). Even the very small elevation change of the plywood triangles in front of the loading stations caused problems, as one wheel would lose contact with the ground. (If we drove "North" corner first. Once the North corner was raised happened, either the "East" or "West" wheel would lose contact, as the robot had a slight tip to one side to rebalance itself, which was problematic, as the East and West motors are needed for movement on the North/South axis). We had a system where we could adjust the wheels relative to the frame, but it wasn't active, and it couldn't be done on-field. So, even the slightest imperfections can cause problems if you don't have some sort of suspension system.

[sanddrag]

It seems like you could make a flexible enough frame to avoid having to do suspension.

[Astronouth7303]

Also, when team 190 did this in 2005, they ended up spliting their chasis in 2, so that the back two wheels swiveled, somewhat similar to your design.

I didn't have the opportunity to actually see them in action, but they did describe it well.

[Andrew Blair]

868 used a pnuematic suspension on their mechanum drive, for the purpose of lowering their entire base, not just keeping contact- to my knowledge. But I'd imagine their design might help you with your design problem if somebody's got a good picture.

[Crazy Ivan]

Huzzah! I found a picture of our suspension. The black plastic circles were eventually replaced with solid rubber. Its a little hard to make out but the hinge is near the vertical bar on the left side of the image.

[Billfred]

Originally Posted by Crazy Ivan

Huzzah! I found a picture of our suspension. The black plastic circles were eventually replaced with solid rubber. Its a little hard to make out but the hinge is near the vertical bar on the left side of the image.

(picture deleted for layout's sake)

Interesting.

I assume the rubber circles were attached somehow--does the movement of the module affect how the robot drives?

[Crazy Ivan]

Originally Posted by Billfred

Interesting.

I assume the rubber circles were attached somehow--does the movement of the module affect how the robot drives?

The rubber circles/rubber cylinders/shocks were held by a 2 inch bolt that went through the frame and screwed into the top of the wheel pod assembly. The modules movement actually is surprisingly good at keeping the wheels in contact with the ground. Once it was aligned, it worked quite well on many different surfaces.

[Mike Nawrot]

Awesome design. Nice and simple. There are a few improvements I see possible. Using 25 chain instead of what appears to be 35 will save some weight, because it will allow you to achieve a smaller pitch diameter on the drive sprocket, which will allow you to use a small sprocket on the wheels to maintain the same ratio, resulting in less material in sprockets and a shorter length of chain. Also, the rollers on you mecanums are all aligned in the same direction, and last I checked, they should oppose eachother to actually achieve omnidirectional motion. That's just a small detail though. Also, I'm a bit concerned about the chain on the rear wheels, since when the wheel assembly pivots, the chain will be forced to twist in a bit of an unnatural fashion. Other than that, the design is amazing in the lines of simplicity. And I like the green AM mecanums

[edit] I just realized that AM doesn't have the .step for both left and right wheels, so ignore my comment about the direction of the rollers.

[Madison]

Originally Posted by Crazy Ivan

The slots are a good feature, but I would say that an active suspension is almost without a doubt nessisary. Some sort of ridged adjustment only works if the playing surface is compleatly level and smooth (unlike the floor of the Georgia dome that is full of warps). And even if all the wheels apear to be touching, just a small amount of slip with one wheel can cause you drive train to go funky. Even with our active suspension, it took a good hour to align and even then it only strafed strait in only the right hand direction, with a small arc to the left.

Unfortunatly we don't have any pictures of the wheel pods up close, but I'll give my best try to describing them. With both systems I mentioned before, one end of the wheel assembly must be on some sort of hinge or pivot (we did this with a removable steel pin). The other end of the wheel assembly was held by screws going through the ridge part of our frame, and threaded into the wheel assembly with cylinders of gum rubber inbetween the frame and wheel assembly. A similar system was used, only with 3 inch pistons at 45 degree angles.

The other method of suspention you can use (as I belive you alluded to earlyer) is the system team 190 used in 2005, where the wheels were ridged, but it was there frame that actually had one big pivot in the middle.

Hope this helps!

I was afraid that an active suspension would be necessary. While certainly not too hard to achieve, the added complexity and additional weight hurts the practicality of the design.

On your chassis, were the bolts that ran through the rubber cylinders riding in slots? It seems like they'd need to because the module pivots on the opposite end, but it's hard to see from the photo you posted -- which is otherwise enormously helpful. For as much as people are fascinated by these drives, I can't seem to find very many photos.

Could you describe a bit more about what you mean by 'aligning' the drivetrain?

Originally Posted by Mike Nawrot

Awesome design. Nice and simple. There are a few improvements I see possible. Using 25 chain instead of what appears to be 35 will save some weight, because it will allow you to achieve a smaller pitch diameter on the drive sprocket, which will allow you to use a small sprocket on the wheels to maintain the same ratio, resulting in less material in sprockets and a shorter length of chain.

We have a lot of material for #35 chain on hand already, including a pretty good variety of sprockets and a few dozen feet of chain. It's also really easy to track down aluminum sprockets for this pitch from sources like IFI or AndyMark and they both tend to have much shorter lead times than places like Stock Drive Products. I'm okay with taking on some extra weight here if it means fewer headaches.

Also, the rollers on you mecanums are all aligned in the same direction, and last I checked, they should oppose eachother to actually achieve omnidirectional motion. That's just a small detail though.

Yep. I'm just too lazy to create a mirror of the wheel.

Also, I'm a bit concerned about the chain on the rear wheels, since when the wheel assembly pivots, the chain will be forced to twist in a bit of an unnatural fashion.

Me too! I rationalized it by saying, "that's what prototypes are for," but since it seems like an active suspension will be very important to making this work correctly, I'll likely correct this in the next iteration. I could include it among the pivoting subframe, but then that starts to become very large and it's hard to mount additional mechanisms to it. I was hoping the chain would have enough play in it, since I don't imagine the subframe will raise or lower more than about 1/4".

Thanks for the continued discussion everyone.

[yongkimleng]

man this is getting exciting, I've never experienced mecanums before.

any way to flip the step import in solidworks?

As for the suspension, usually how much vertical movement would be enough to keep the wheels on the ground? Looks like the rubber is sufficient, but that depends on the overall weight of ur final bot right?

I was thinking of a pneumatic suspension where the pressure applied on each wheel could be adjusted as required (maybe not on-the-fly). anyone done that before?

[Andrew Blair]

Perhaps an active suspension can be implemented without too much redesign.



When totally locked down with the spacers you've designed in, I believe that this could be strong enough to give you what you want. It appears you're using 1/8" or 3/16" stock- maybe an upgrade to 1/4" would give you all the strength you need there. I would think that if you got hit really hard, the ability of the mecanums to slide would keep you from bending the modules. The chain would still be an issue though perhaps.