Re: Re: Hexapod question

> From: "ballendo" <ballendo@...>
>
> Ray,
>
> That is the only one that is going to be accurate without
> compensation. As Mikkelson's machine shows, you can do without ball
> ends. BUT every fullsize realworld machining oriented hexapod I've
> ever seen or heard about uses ballends. (I don't know the details of
> Till F's full size mill.)

Till uses fixed length struts that ride on vertical linear motors. His
geometry is constantly changing as the motor ends of the strut slide.
While the kinematics equasions are slightly different, changing angle,
same length, the principal of passing a world coordinate in x,y,z through
some set of equasions is the same. So long as the equasion used matches
the nature of the physical geometry, the translation will be accurate.
With some Stewart geometries this may be quite a trick.

When I was there, Till was refining the constants in his kinematic
equasions by drilling a set of holes at fixed distances over a surface
and then measuring actual position with a CMM.

Ray