We’d started work on designing a modular test bed for future ideas – a basic frame that could accomodate a variety of wheel arrangements and varieties. I’ve had trouble designing a suitably small, similarly modular gearbox, so this is on the back burner for the time being while we work on something a bit more conventional.
The notion here is that the frame will contain a series of locations that will accept a universal bearing block – the gray squares inset into the frame. These bearing blocks have an offset hole and can be rotated to vary and combine wheel types and diameters. Here, there are 6" omniwheels and 4" IFI traction wheels mounted on the same chassis using common parts.
I assume you just used the IFI traction wheels in that because you could easily import the file? Beacuse it appears you inputting a live axle into a wheel designed to be driven by a dead axle at the moment. :rolleyes:
Nope – we’re planning on adding AndyMark Universal Hubs to the IFI wheels so they can be driven by a live axle. I’m interested in seeing how well the new IFI wheels hold up under this sort of arrangement.
Interesting, you must let us know the results. Seeing as there is no AndyMark performance wheel with a 4" diameter, IFI has a relative monopoly on the “lightweight traction” wheels available off the shelf to teams. It’ll be interesting to see how well it can perform in a live axle system.
Seems good… But that looks like a lot of work for a prototype project.
Have you considered just making the frame out of 80/20 and mounting some sliding pillow-blocks on the grooves? Something like that would allow for nearly any combination of wheels/treads/meccanums, and would require significantly less work.
80/20 is really the way to go in terms of prototyping.
Well, I’m pretty confident of the team’s ability to build a reliable drivetrain – we’ve been doing it well for a few years now. Instead, what we’re really prototyping are methods and processes for speeding along the production of the drive and making it as easy to assemble, repair and replace as possible. The 2007 drivetrain, ignoring shipping delays, was assembled in less than a week and this drivetrain, again assuming all material on hand, is designed to be fabricated and assembled in about 12 hours. I’m trying to bring that number down even lower.
The 10WD was borne out of poking and prodding of some of the folks on our team. We’re quite happy with our performance of our 2007 robot, obviously, but one of the regrets I have is that it was so slow. The extra pushing oomph helped us out, of course, but it might have been nice to be able to move faster. We’re looking at two-speed gearboxes as the solution and thus needed to do something to put the available power from them to the carpet.
After toying with a 10WD chassis for awhile, we began to realize that it was certainly overkill for most circumstances – and that’s when the idea of modularity came up. I’m having some trouble making a two-speed transmission as small as I’d like while maintaining the ability to mount it to any wheel of any diameter, however.
In the end, we probably will build something a bit less elaborate than this. We’re looking at a simpler 6WD arrangement, since that’s what we’re most likely to use in the future, and instead focusing a lot of our time on integrating feedback loops into the drive and other components so as to build a test platform for the software and electrical folks. That area needs more work than mechanical reliability does – though that’s not to say that they’ve done a bad job in the past, just that I haven’t given them the time and tools they’ve needed.