I have been getting really into crab drive since just a little while before championships. I was pumped to get to see so many awesome crab drive modules in Atlanta, many using bevel gears and a hollow shaft so the modules can spin as many times as they want. One question that i still have is quite possibly the biggest thing to overcome in making a crab drive chassis. How do you attach the module to the frame such that it is strong (will not have much play especially in a pushing match) and able to rotate.
Looking at the 221 revolution module i see what appears to be a thrust bearing but it is still confusing. Is it expected that you will slide the 1.5" Coupler piece into a hole drilled in the frame and attach it with a collar? That is what it seems to me from reading the description.
If you have a picture of a module attached to the frame that would certainly be very helpful
Team 228 should be posting their SolidWorks model very soon for this year’s robot. Our mounting was to capture the swerve module surfaces between two 0.25" Delrin bearing plates mounted to the robot frame. It took quite a few man hours to get the bearing plates to run flawlessly, but it was worth it. This connection system worked very well for us this season with no failures.
The bearing surfaces had quite a large OD in anticipation of the cantilevered stress from climbing the bumps and playing heavy defense. On a flat playing field we’d probably go with a much smaller bearing surface.
We did not do full concentric swerve. There are measurable efficiency losses in using bevel gears which we didn’t feel were offset by the ability to spin a drive pods continuously. Our swerve pods were fully independent allowing us to rotate while driving in a straight line (there’s a white paper here on CD covering independent twisting swerve).
We used a three axis joystick and gyros for field-centric control. The third axis (joystick twist) allowed for the rotate while driving in any direction.
The swerve system had a pretty steep learning curve for us. I would definitely recommend prototyping during the off-season to avoid some of the stress that we encountered.
Is it expected that you will slide the 1.5" Coupler piece into a hole drilled in the frame and attach it with a collar? That is what it seems to me from reading the description.
The stock mounting solution for the Revolution swerve is exactly as you described. The outer most hub is intended to be clamp mounted to a plate or whatever surface you prefer.
The outer hub could also be removed and a series of bushings or bearings could be used to capture the module instead.
Rigid mounting is definitely a concern when developing swerve technology. Good luck
The 1986 swerve module is supported entirely from above the module by a vertical, hollow steel pivot stem mounted in bronze bushings. It has often been stated as a given that in order to handle the sideloads seen in a traditional game on carpet a swerve module must have lower support. We have proven that is not necessarily true. Our welded steel pivot stem saw brutal punishment throughout 3 events, and is in perfect shape.
Our drawings are online, look at the 4 Wheel Pivot Drive page. There are links showing how the pivots are built and attached to the robot. You can see how the mounting plates hold the bearings and how the entire assembly is kept together.