If I was to attempt to make a swerve drive, and wanted my modules to look something like this (http://www.chiefdelphi.com/media/img/8f7/8f791d18825f620892ed048ad670c314_l.jpg), how would I go about supporting the module while being strong enough to withstand the side load of being pushed and minimize the friction while rotating the module. Are there a certain type or combination of bearings that would work here? How have your teams dealt with this problem?
It looks like those two plates at the bottom are used to support it. I vaguely remember the discussion of this module here a while back.
My best guess is that the plates would be Delring or HDPE as they are rather low friction.
The best example to look at for this example of crab drive would probably be 111 and 118. I know there are a lot of crab drives out there, but these two stand out to me.
When my team did swerve, we made the gearbox with a steel shaft sticking out the top, and used 2 radial and 2 thrust bearings on each. This was overkill, we realized lated we could have just used plain old bronze bushings.
For our design, the mounts for the swerve were part of the chassis, and the sprockets that turned the modules were mounted to the top of the steel shaft which extended through the bearings. Look at swerves made by 114, 115, or 258 if you want a better idea of what I mean.
You just need 2 support points, The frarther apart the better. 3 inches apart souns good, but then again I have no expirience with swerve modules. Just apply what you know already.
You are very close to good design. We have found that a combination of things works. We use a bronze sleeve bearing with a thrust plate built in for the top. Added to that is a Delrin circle that surrounds a circular plate at the bottom of the module. This bottom plate is mounted on the bottom of the module and is designed to match the bottom of the chassis. This way the top and bottom of the module is secured by bearing points. The bottom bearing goes a long way to keeping the bottom of the module aligned (so that there is no misalignment in the top bearing producing friction) and prevents the module from moving during driving and collisions. By locking the bottom, you prevent the wheels from getting up on edge (less friction) and you allow the steering linkage(at the top of the module) to maintain a reasonable amount of tension for better control.
I assumed I would need to support the module at the bottom and the top. What I was especially curious about was where you get your delrin, how you mount it to the rest of the module and frame, and if you use small bearings or rollers around the arc to lower friction. I can see how the top bearing would work, seeing as it is a smaller radius.
Another decision i have to make is whether to have one CIM + gearing within each module or to use miter gears and transfer power through the pivot(using outside gearboxes and a sprocket.) The price and inconvenience of mounting the CIM inside the module points me in the other direction, but I was wondering whether one miter gear rotating around the other while they are spinning at high RPM causes problems. I guess now that I think about it it wouldnt really matter, you just need to make sure that your pivot is very precise so that your miter gears maintain good mesh no matter what angle the one that rotates with the module is.
The delrin bearing is cut from a sheet using standard machining technique. It is a “z” shape through the cross section but with the vertical at right angles to the horizontal pieces. This allows the circular plate at the bottom of the module to hold the delrin in place once the module is mounted in the robot. I have been under the impression that bevel gears are just too lossy for this application. We have always mounted the motors inside the module. We have used drill motors, Fisher Price Motors and the small Chalupas. As I remember, the gear box uses standard gears cut from gear stock and mounted on shafts with bearings in the side plates of the module. Even with the Chalupa, the module ends up just slightly larger than the length of the motor with a gear on the shaft. The wheel is mounted almost touching the motor case and the gear box is just off to the side of the wheel. As you can imagine, using crab drive does bring the wheels closer to the center of the robot since the turning radius of the module must fit inside the outer dimension of the robot frame.
You don’t need two support points. Our swerve works wonderfully to this day, and it only supported by a shaft out the top. However, it is heavy…
this year i designed a swerve drive based off of 71,111,118 mainly and some others.
it was a 3-speed shift on the fly coaxial swerve drive using dewalt transmissions and bevel gears. the bottom of the module was supported by rollers inset in the frame to create a huge roller bearing but due to annoyances we’ll be eliminating those next year and using delrin sheets around the exterior of the bottom to save weight and complexity. If you do make a swerve drive make it maintainable we arrived at the pits saturday morning at the midwest regional started some maintenance that ended up taking and hour and a half which according to the match schedule would have been fine but they were running early and we missed our very first qualifying match ever just a little warning to it doesn’t happen to anyone else.