Re: Die STL.die

From: Stephen Rock (rocks@rpi.edu)
Date: Fri Dec 17 1999 - 07:55:46 EET


ATiburon@aol.com wrote:
>
> In a message dated 99-12-16 14:11:02 EST, gfadel@ces.clemson.edu writes:
>
> << One advantage I saw mentioned by Stephen Rock in the exchanges is
> that slicing would be much faster - say 2 to 3 times faster on the average.
> Would that be good enough a reason to migrate? >>
> The major drawback to STL is the FACET. The STL is great for describing
> boxes. But really sucks for curved surfaces, small radii, internal
> passageways etc. It's ludicrous in the extreme for RP manufacterers to be
> claiming .001 in plane accuracies when the file format won't come that close.
> Not with a zillion triangles! There have been formats already that did
> perform better when used on curved models, things like Cars, Airplanes,
> barbie dolls, etc. One is CATSLICE, available for Catia only I believe, I
> don't know if it is still supported. The other was the approach as by Fockle
> and Schwatrz, and perhaps a few others to use HPGL slices, supported by
> Magics etc. The STL is the major impediment for RP's progression/evolvement
> into RM (rapid manufacturing). Accuracy, surface finish and utlity all suffer
> because STL will never be more than just an approximation. That is OK perhaps
> for many models, but totally inadiquate for tooling or production.
> Andy Scott
> Lockheed Martin Aero Sys

Andy,

Are you saying that you favor data exchange to RP on a plane-by-plane
basis (allowing the native CAD solid modeler to perform the slicing and
then export precise planar contour geometry)? Or, should a geometrically
precise 3D format for RP be developed (so machine operators perhaps not
having your CAD system in-house can process the part)?

The later would seem to be a good challenge, but by time it gets
implemented and working, Moore's law (the doubling of compute power
every 18 mos. [if not less]) and any similar improvements in storage and
network bandwidth may mean that those users happy with the status quo can
process your zillion facet model without all that new [necessarily expensive]
software that deals with precise geometry.

Do you really need a zillion facets to have an approximated model
with precision on the order of that deliverable by current RP process
technology?

Regards,
Steve
=======================================
Stephen J. Rock
CII8015
Center for Automation Technologies
Rensselaer Polytechnic Institute
Troy, NY 12180 USA
(518)276-8652 Fax -4897
http://www.rpi.edu/~rocks
=======================================

For more information about the rp-ml, see http://ltk.hut.fi/rp-ml/



This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:53:49 EEST