Re - bad surface on top facing part with holes/protrusions

From: Stahlhut, Todd A. (
Date: Thu Sep 18 1997 - 20:34:00 EEST

I posted a message last week concerning bad top facing surfaces on parts
with protrusions and openings such as text, bosses, and holes. It is
theorized that the bad surface finish is due to the drawing method
where the laser sweeps between features rather than clear across the
part - skipping openings/protrusions. I have a front panel mockup with
a lot of raised text. The resulting top surface was quite bad due to
all of the surface interruptions. After inputs from several RP mail
users, a call to the 3D hotline (yea its a problem on the 250 - hope
they fix it someday - maybe NT.....) and some actual thought on my part
(some would say a rarity) - I came up with a half way solution that is
a pain - but helped. Hopefully others can use this trick also. What I
did was went back to the ProEngineer model and cut the text off at the
top surface and wrote this out as a STL file. Then I cut the part off
even with the bottom of the text and wrote that out as an STL file. I
then located these STL files in the build set so the text lined up
properly (but moved down a layer thickness to ensure bonding) and build
the part. Because they are two separate parts it completes the bottom
part first - sweeping clear across the part - leaving a nice surface,
then it draws part two - the text. The resulting part was significantly
better than the first attempt at building it as a single part.

This was a pain - but was worth the effort to get a better product to my
customer. If you are listening 3D - please fix this bug (yes I realize
this falls in the product enhancement category - but I consider it a

Regards, Todd Stahlhut

P.S. - Another lesson that I learned in building this part was that at
least with ProE text (about .1 tall) was that beam comp needs to be
significantly reduced or turned off completely. With my method above,
you can leave beam comp on for the main part and turn it off for the
text part.

This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:40:25 EEST