In addition to the above interference issue, I also don’t see any way of retaining the axle bearings between the two side plates. These look like flanged bearings, in which case the easy fix would be to flip them around so that the flanges are inside the module instead.
Seeing how compact you’ve gotten this thing already, I am also led to consider the possibility of ditching the outer branch of the mecanum side of the module… Doing so would require screwing into the outer end of the axle to prevent the side plates from pulling apart from each other, but swapping out that bearing for an X-type (4-point contact) or conical bearing aught to do the trick. (The robot frame would support the opposite side panel, so you’d only need to swap out the one bearing.)
Regardless of whether or not you take that suggestion, however, this is looking pretty good! You’ve made a whole lot of progress in a pretty short time here.
It looks like you are planning to move the traction wheel up and down with the mecanum wheel fixed. If you switch that and float the mecanum wheel you can use air pressure to create a suspension system and set the pressure to assure you have all four wheels in contact with the floor. This makes a much better driving mecanum if the floor is uneven. When one wheel goes off the ground driving goes bad.
Don’t forget about the side thrust created by the mecanum wheels. You need some way to react against that force.
Yes, it actually works out that way. Set the pressure so the robot floats and the air cylinders are not fully extended to the end of the cylinders. That works out to be the pressure where 4 wheels support the robot. If the floor is uneven the wheels will follow the floor. The range of motion needs to be enough to include the maximum amount of floor deviation.
Personally, I would keep the mecanum wheels fixed, you probably don’t want a wheel that is constantly applying side load to be on the end of the module. Also, if you are primarily in mecanum mode, then you probably want that to be stable and not be affected by how much pressure you have.
To counter slide load from the mecanum wheel you can use a thrust bearing, but my team has found plastic spacers on the axel do fine (ex: vex spacers)
On another note, I see you decided to double up the colson wheels. Do you feel this is still considered octocanum, or is it dodecacanum? (since you have more than 8 wheels now) My team has been trying to decide since we do the same thing with our newest octocanum.
I’m guessing the idea was to have the mecanum mode drive faster than the traction mode (different size wheels is a quick way to do that). Of course, you could also accomplish this by creating reduction between the belt sprockets.
Why the dual colsons? And why not a smaller mecanum? This could be a lot lighter if you removed a colson and just used 4" mecanums.
Why are there two standoffs almost next to each other in the middle of the module?
Apart from that, this looks very neat. The lightening pattern is perfect and it all looks quite clean. The pivot looks very nice too, and should let you use a smaller piston. However, this could be a good chance to put the mecanum wheels on a suspension by maing those actuated instead of the traction wheels by swapping the two.
This is the setup I originally had in mind, but really almost anything with a 0.5 hex output shaft would work with this.
Thanks, I actually hadn’t caught that yet. There is enough room on the mecanum shaft to scoot things over a bit and put a spacer on either side of the wheel.
I was actually just thinking retaining clips on the ends to keep things in place.
Also, could you post an example of one of the bearings you mentioned?
Well, my team already has a pair of 6" mecanums in stock, as well as a full set of 6" Omnis. I’m just trying to save some money there
That being said, a 4" model is on my list to design next.
I probably wouldn’t want to have the main drive wheel utilizing system pressure most of the match unless there were large field obstacles (like 2010).
My team ran no suspension mecs this past season and almost never had a problem.
Interesting… I think I’m going with a Duodēcim drive!
Maybe Duodec-drive? Not quite as elegant sounding as nonadrive.
This is true, but it also helps make the speeds vastly different.
I re-worked some math from the last set I posted and made the mecanum run at 13.4 Ft/s and the colson run at about 5.5 ft/s
The added Colson is just to increase the amount of ground contact. The extra colson is easily removable from the system, I was mostly just testing to see if I could fit both on there.
The added spacers is just me trying to over-engineer this thing to make sure it’s rock solid. I could probably get away with removing it.
I would actually suggest flipping the bearings, but also using normal FR8ZZ bearings instead of the hex versions. That way, it takes care of retaining the bearing as well as retaining the shaft. When you have hex bearings on both sides, the axle’s free to move without retaining clips. Regular bearings are also cheaper and have fewer concentricity issues.
I think you have gotten some good feedback, just a few more comments for you.
You can use flat head bolts and countersink the outside of your plates to clean up the surface, that should help when you actuate the module up and down so nothing catches. Also, the inside can be counterbored slightly to keep the standoffs from sliding around. Then the standoffs can be tapped on either end and you can save money and weight on shorter bolts by not running them all the way through. The piston can also be attached to one of the standoffs, so the big shoulder bolt doesn’t have to be used to save weight.
How are the axles retained? You may need a bit more space added for a retention method. One option is a retaining clip between the pulley and wheel, that will keep the axle from moving either way. The down side is that creates a stress riser in the middle of your axle, but it should be okay since it’s not cantilevered. Or you could get rid of the live axle on the traction wheels and go with a dead axle setup, that would solve the retention method and get rid of the hex bearings.
There was one module I saw before, I don’t remember which team it was from, but it was packaged nicely. They had a CIM driving a gear on the mecanum wheel and the opposite side ran a belt to the traction wheel. Then the gearbox was packaged right into the module, it eliminates the complexity of a separate gearbox and hex shafts/bearings with dead axles. It just depends on what you want to do with the module.
There are a lot of design decisions to be made. Have fun with it!