In a message dated 99-12-21 06:18:09 EST, Makela@deskartes.fi writes:
<< To utilize the direct slicing fully we should use IGES/STEP etc.
to transfer the data to the RP machine site and slice the
surface model locally, according to the different needs of
different RP machines. I believe this will be available in a
couple of years. >>
Slice on the fly as we would say. Actually, Helisys has been doing it for
years now(not counting the connectivisation). The whole concept of slicing a
file prior to running the model is an archaic paradigm. That was necessary
when a hot rod puter was a 12 Mhz 286. But totally absurd as processor speeds
approach 1GigaHertz. Orienting and supporting, should be done on the Cad
model directly. Then the machine slices, not necessarily in fixed thickness I
might add, as the part is built. Fockle and Schwartz did something similar
using ProE models on their SLA, with help from Magics. Devotion to STL format
is touching, but doesn't serve to progress the technology. I forgot to
mention in the previous mail about fabricating optical devices. What would be
a good chord height for building a functional set of bifocals, 10 micron,
.005 micron? Or would a format that describes curves more effieciently be
appropriate? What are the present RP machines capable of, as far as accuracy
and surface finish? Who can say, thanks to STL you can only guess. The RP
vendor that can produce the better finish and accuracy will have a
competitive advantage. I would like to see an optional S/W module that could
use the inhouse CAD model directly. Sort of like a post processor for CNC.
One would buy the ony the ones needed. Catia, ProE, SDRC whatever. You could
certainly use always STL in a pinch, but quality would suffer as it does now.
Andy Scott
Lockheed Martin Aero Sys
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 - 22:53:51 EEST