An Iso view of our final design for our swerve modules. They have already been made, and are being assembled right now.
Each module is about 5.5 lbs. fully assembled. Any and all questions are welcome
after facing several problems we are currently in the process of deciding what to use as an alternative to keyed shafts
Hex shafts work great. you can either find someone with or buy a hex broach. or even easier buy hubs from AndyMark that are prebroached.
What exactly was wrong with the keyed shafts?
Hal was a bit quick to insist that we are looking for an alternative to keyed shaft. We encountered a small problem with holding one of our keys in place, but we have already figured out a way to address the problem. Thank you for the quick suggestion of hex shaft, but we intend to stick with the keyed shaft we have now.
If it is machined correctly the key should have almost a press fit so it doesn’t twist around. You can hold it in place by simply having spacers on either side of the sprocket/gear to limit its movement or by using a setscrew.
IMO keyways are MUCH easier to make than hex parts. You need a lot of special tools and skills to correctly machine hex parts. Keyways are very easy.
Thank you for the advice, I think the approach that our drive team concocted is very similar to the one you proposed 
Is very cool, I wish we could build a swerve drive, but are you really going to use swerve this year?
We asked ourselves the same question during brainstorming. Our answer is yes. We believe the ability to employ “thrust vectoring” (there is your arbitrary aviation connection of the day) will allow us to maneuver better than a differential drive. While we may not spend all match “strafing,” the ability to use transverse friction for braking by orienting the wheels perpendicular to the direction of motion, and put all of our power into turning, rotating or driving makes us believe a swerve is worth the turmoil that usually comes with building one.
cool, i’d like to see what else you’ve done on inventor
This reasoning also drove our team to go with a steering chassis as well. (http://www.chiefdelphi.com/media/photos/32573?) We agreed that we wanted a chassis that “puts all of its energy in the direction we want to go.” “Thrust Vectoring” is a good way of putting it.
Turning the front and back sets of wheels independently will deliver car-like steering (Front wheels turn +X degrees, rear wheels turn -X degrees). Orienting all the wheels the same way will deliver crab-like motion. Although this probably means we’ll be “orbiting our trailer,” we sill would like this functionality so we can escape from robot jams and possibly align our robot with the fueling station and outpost more accurately.
I’m really anxious to see pictures of the skunk module to see how it compares with our design. I think it’s cool that a lot of teams arrived at this common design solution for this brain-scratching problem of a simulated lunar gravity.
OMG!!!:ahh:
we have a concept just like that…but its a little different than that