Re: VRML instead of STL

From: Gary Scholl (gscholl@handel.jlc.net)
Date: Fri Jan 05 1996 - 17:57:13 EET


Greetings all.

This topic seems to have sparked the first significant volley of opinions
for the new year. I've got some too. Please excuse the "paraphrasing."

1) "The VRML proposal provides no improvement or degradation in the
information now provided in by the STL format." -- It ain't broke. Don't
fix it.

2) "The VRML proposal provides a more compact and elegant data structure."
-- Yea, that's nifty, I like that. However, lots of us have already
invested in dealing with the STL format so the cost of conversion is too
high when compared with the added value of the space saving.

3) "Polygonal Representations are faster to slice than Complex Surfaces" --
Well, there are lots of special cases (e.g. Boxes and Spheres) but it seems
to me that, overall, this perception is false. One must also include the
time and degradation that occurred when the true surfaces were first
approximated by the polygons. I don't see the value in saving time slicing
a sausage if I started out with a tenderloin.

The real problem remains: "How can we best transfer a ^suitable^
representation of a CAD model between the system used by the Designer and
the system used by the Manufacturer. The issues are: economy of time, data
integrity and cost. There are lots of solutions (Least we forget, the most
prevalent in the world today are: "file to fit" followed by the Engineering
Sketch/Drawing.) None yet are acceptable in all cases.

For reasons of fluid flow and stress management, my customers want parts
with smooth complex surfaces. Parts with obviously polygonalized surfaces
simply do not cut it. Since R-Pt devices ultimately deal with only 2 1/2D
data, one solution that works for me is to have the customer provide slice
data. (Well, there seemed to be a good reason for yet another format. BTW
the resulting surfaces are much, much, ... better than that possible with an
STL file of similar size.)

*^I would like to see a better standard for transferring surface data^*.
*^I would also like to see designers create better quality CAD models in
thereown systems^* ...

apologies in advance for the soap box.
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
Gary W. Scholl
Metal Casting Technology, Inc.
Milford, New Hampshire 03055

v: 1 (603) 673-9720 x 437
f: 1 (603) 673-7456
e: gscholl@jlc.net



This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:37:05 EEST