Alt-BEAM Archive

Message #05664



To: beam@sgiblab.sgi.com
From: Sean Rigter rigter@cafe.net
Date: Thu, 12 Aug 1999 08:09:29 -0700
Subject: [alt-beam] Re: "Linear Wheels" (was: LEM videos now on-line)


Ah, right about that but these are an emulation of the Killough Linear Wheel
and meant to be mounted in a triaxial arrangement at 120 degree intervals on a
3DOF platform as shown in the July 97 issue of Discover magazine (somewhat like
Bob Shannons's LEM). Remarkably , this results in a omnidirectional linear
motion which is the vector sum of the rotation and speed of the three motors
driving these wheels. This is entirely different from Bob Shannon's LEM motion
which achieves linear displacement afaik serendipitously through slipping and
dragging wheels. Compare the two videos : they are a must see! (I inserted the
Zoz and Mark Noakes original emails for context below)

regards

wilf

Bruce Robinson wrote:

> Wilf Rigter wrote:
>
> >
> > It seems to me that a conceptually simpler "linear wheel" would, as shown
> > on attached example, consist of a 10 (or more, or less) segmented
> > circumference wheel with a "free wheeling" spindle on each segment.
>
> This might have no bearing (pardon the pun) on what you are talking about,
> Sean, but I strongly suggest that you take a look at a two
> degree-of-freedom "linear wheels". They are more complex than everyday
> wheels, but they ARE free to move (with quite low friction) in directions
> other than the driving direction. See:
>
> http://carol.wins.uva.nl/~leo/lego/killough.html
>
> for the basic description and the following mpg movies to see the Killough
> platform in motion:
>
> http://www.net-info.com/~dcarlson/klw1.mpg high resolution/shorter
> length movie
> http://www.net-info.com/~dcarlson/klw2.mpg low resolution/longer
> length movie with several different motions
>
>
>
> Zoz
>
> This is cool to see Steve Killough's stuff mentioned; I work with Steve and
> am very familiar with this approach.
>
> Mark Noakes
> ====================================
> Mark W Noakes
> Robotics & Process Systems Div
> Oak Ridge National Laboratory
> PO Box 2008, Bldg 7606, MS-6426
> Oak Ridge, TN 37831-6426
> Phone:423-574-5695 Fax:423-576-2081
> Email: mqn@ornl.gov
> ====================================
>
> This certainly works, Wilf, but don't put these designs on a vechicle
> with all the wheels parallel to each other -- your machine will drift
> sideways at the first opportunity if you do.
>
> I once saw a design (somewhere on the net) of an actual vehicle using a
> variation of this design. The segment wheels were turned at 45 degrees
> to the plane of the main wheel, instead of at 90 degrees as your are.
> The wheels were mirror-imaged on both sides of the vehicle.
>
> In linear, straight ahead motion, the segment wheels wouldn't be able to
> turn without skidding, since they would be trying to roll at 90 degrees
> to each other. If the main wheels on one side of the vehicle were locked
> or slowed down, the segment wheels would be able to roll to reduce
> friction. Sort of like a skid-steer tractor, without the skid.
>
> Love that mechanical stuff.
>
> Regards,
> Bruce

Home