View Single Post
  #33   Spotlight this post!  
Unread 18-08-2006, 14:55
Madison's Avatar
Madison Madison is online now
Dancing through life...
FRC #0488 (Xbot)
Team Role: Engineer
 
Join Date: Jun 2001
Rookie Year: 1999
Location: Seattle, WA
Posts: 5,245
Madison has a reputation beyond reputeMadison has a reputation beyond reputeMadison has a reputation beyond reputeMadison has a reputation beyond reputeMadison has a reputation beyond reputeMadison has a reputation beyond reputeMadison has a reputation beyond reputeMadison has a reputation beyond reputeMadison has a reputation beyond reputeMadison has a reputation beyond reputeMadison has a reputation beyond repute
Re: A new drivetrain Idea?

Quote:
Originally Posted by Kevin Sevcik
The one arrangement I can sort of envision is when your hinges are at 45 deg and the sides are now perpendicular to each other. If just one side moves, the opposite side will side-slip and it'd end up going that direction. So I'm skeptical it'd do exactly what you think it will in that arrangement, and thus I'm not entirely certain about any other arrangement.
(I'm quoting Kevin as a starting point, not because he's wrong. He's not.)

The idea is that the chassis will rotate about the point where lines, drawn perpendicular to and from the center wheel on each side, intersect. This is true of how traditional tank-steer robots operate as well, except that they typically lack the ability to move that point as this design allows.

Typical drives can spin in place because, by moving each side opposite the other, the translation vector of each -- that is, the force that would move it forward or backward -- is cancelled by the other. When each side is no longer parallel with the other, some of that translation escapes, as it were, and the chassis starts to behave as an inefficient holonomic platform. Thus, running each side in the 'opposite' direction will result in sideways translation. Driving them in the same direction will result in forward or backward movement. The only exception would be when they sides are again parallel, but collinear, as would occur when opened 90*. In this instance, driving them in opposite directions would do nothing at all (except break your robot, perhaps), while driving them together would result in the most efficient sideways translation possible because you've effectively changed which direction represents 'forward'.

Without a fixed point about which to rotate, all this design seemingly allows is a way to vary the efficiency at which a holonomic platform operates. That's not without merit, necessarily, but it doesn't seem to me that it would work as advertised. All of this assumes, of course, that it uses omniwheels. Other wheels would effectively function the same, but it would be even less efficient.
__________________
--Madison--

...down at the Ozdust!

Like a grand and miraculous spaceship, our planet has sailed through the universe of time. And for a brief moment, we have been among its many passengers.

Last edited by Madison : 18-08-2006 at 14:59.
Reply With Quote