Let us choose horses for courses
STL is not a format that satisfies everyones needs and it is not a standard
format. The output file may be in a standardised expression but the
interpretation from the true surface geometry is not.
The sender is not informed of the approximation errors he has produced and
it does not give any indication of its acceptability to the receiver.
(Bad files recently have come from VX Verimetrix, SDRC I-DEAS, ProDesktop
and Autodesk Mechanical Desktop software packages)
3D Systems did not specify a standard meshing algorithm and CAD vendors have
differing meshing patterns. The users choice of chordal deviation is
sometimes not fully explained and a true geometry error from the triangles
centroid to surface is not checked and may be many times more than the edge
chordal deviation that the user has chosen from the CAD menu.
Any bureau worth its salt would educate its clients in respect to the best
route for the quality level that customer is expecting and inform them of
what can be produced. This was my role in the early years establishing
Formation as 'Top Dog' and even now, years later, here with PERA new users
still need information of what can or can't be achieved.
The direction in which I am steering the process within PERA is one of
continual improvement.
The SLA350 is a higher specification machine to the old SLA250, I had used
previously.
The old resin SL5190 for the SLA350 could not match its possible performance
potential. RPC 300 ND now allows me to build more accurately with better
surface finish and at a lower layer thickness without too much of an
increase to build time.
The finish is so good that problems with data now stick out like a sore
thumb and now require a remedy for the direct SLA tooling projects under
consideration.
3D Systems are sorting out the slight glitches with laser
positioning/performance.
CAD software improvements needs to be tackled next.
If I can improve model accuracy and reduce approximation by using another
format then I will either with a better STL specification solution or by
using direct CAD slicing SLC files.
These are the only two that 3D Systems use and are of direct interest to me.
I had dealings with the laser cladding project that PERA spent 2 years
developing and the software that was produced for this development stage was
to import STEP or STL files and output point to point or interpolation
spline data for the Heidenhien machine controller so I know the problems a
new RP systems would have to look into to produce a new file format.
What I would like to see before any talk of a new file format is a complete
breakdown of each CAD package in terms of the type of mathematical surface
geometry and the tolerance that they are computed.
To know what number of decimal places the output of data such as IGES, STEP,
STL etc. are.
To have a 3D demo geometric model and software that would test the ability
of each CAD package to output files for use with various RP systems and act
as a comparison standard.
Happy Christmas, New Year and New Millennium celebrations to all RP-ML
listeners.
Jan Andrzejewski jan.rp@pera.com
PERA
Technology Centre
Nottingham Road
Melton Mowbray
Leicestershire
LE13 0PB
UK
Tel +44 (0) 1664 501501 Fax +44 (0) 1664 501556
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