I-DEAS .stl

From: wlfoss@collins.rockwell.com
Date: Thu May 11 2000 - 17:04:53 EEST

I have not had much opportunity to work with I-DEAS for a while and then it
was not much more than manipulating part files. So, some of the comments
made sounded a little off to me. I found some time and performed a brief
experiment. However, please note that we are at Master Series 7m2 which is
many revisions ahead of the 2.1 someone else mentioned. I too had problems
with STL files in the earlier versions. Now, everything seems to work quite
well and I have not had any problem files for quite a while.

I pulled up a part at random and like most of our designs, was a milled
chassis with straight sides and surfaces. There were fillets and a couple
of smaller curved surfaces. And to qualify my statements, we really don't
care much about those surfaces in parts like these. In the rare instances
when we make appearance parts with complex curved surfaces, we take more
care with the STL file. However, we have had excellent results with only
dropping the Facet Deviation down to .0015 or .001.

1. Standard settings (ours) -- Facet Deviation - .002, isoline density -
0, facets generated = 10,934
2. Change isoline density only -- Facet Deviation - .002, isoline density
- 2, facets generated = 38,140

3. Change Facet Deviation only -- Facet Deviation - .0004, isoline
density - 0, facets generated = 22,632
4. Change Facets and isoline -- Facet Deviation - .0004, isoline density -
2, facets generated = 61,188

NOTE: This is in inches.

A few comments:

The more facets, the harder to see problems, shaded or not. So we use a
viewing software.

The more facets, the harder to dynamically rotate the part, it gets really
jumpy and hard to control.

I think in the earlier versions of MS, you could only change the isoline
density on the whole part. Now you are able to change individual surfaces.
I would go this route first - at least for our types of parts.

The setting for the size of RP modeler really doesn't matter. The results
are the same if you pick an SLA250 or an SLA500 (note that there is no
selection for the SLA5000 or SLA7000). If your part is 18 inches long and
you pick the SLA500, you get a warning that the part is too big for the
modeler. This is the same for the positioning selections. This is helpful
if you happen to have an SLA modeler, but your service bureau's RP modelers
software is going to take care of the parts size and positioning.
Unigraphics has none of these selections.

As for the number of facets, I have been asked by the service bureau to
re-do some files to a finer definition. But I have also been asked to
back-off on the definition to reduce file size. I don't try to second guess
the SB anymore - at least too much. I have been told by SBs and I-DEAS
developers that the facet deviation of .002 is real close to the optimal
setting, but that was some time ago and could have changed by now. I have
found out that the larger the radius, the lower the Facet deviation needs
to be.

Someone made the suggestion of getting a good viewing software. I have a
well known one because we use eight CAD softwares in-house (only I-DEAS and
UG for the vast majority of work). And I have problems with reading in
files and displaying the parts. It is not a specific problem but a
temperamental one.

IGES files have caused me a great deal of heartburn at times. I prefer to
use STEP.

I have a Stratasys modeler, and it's software is real forgiving. I have
very little problem with surface definitions and other STL errors which
will stop other modelers.

Hey! Anyone want to discuss I-DEAS Team Data Manager?



How about some ideas on Unigraphics settings !

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 - 23:03:28 EEST