Basically this problem stems from the model resolution not being tight enough. These large or "stray" triangle are being generated by faces that do not have completely closed boundaries. If you examine the boundaries closely enough you will find gaps. Catia will consider a boundary to be closed if the gap is less than the "identical curves" parameter found in the model tolerance settings. For example if the parameter is set to .004" (which is the default) and a gap exists of .002" the boundary will be considered closed and the auto-close function will jump over it. While this works for the limiting function the STL translator does not like this and problems like you are describing arise. The solution is to set the model tolerance tighter and remake the problem faces. We typically use a model tolerance 10X the CATIA default, but note that this will not conform with the rules of everyone's standards committees.
Eagle Design & Technology, Inc. Ph: 616-748-1022
2437 84th Ave Fx: 616-748-1032
Zeeland, MI 49464 E-mail: firstname.lastname@example.org
USA Web: www.eagledesign.com
From: DAN DAVIS, RAPID PROTOTYPE [SMTP:DANIEL@PMISD.PROTON.com.my]
Sent: Friday, March 27, 1998 4:44 AM
To: 'RP Mailing List'
Subject: More CATIA Q's
Since there seems to be quite a few CATIA experts out there, here's another puzzle that I am facing.......what is the cause of planes jutting out from the part in an STL file? The STL translator appears to be trying to connect a reference point out in space to a point or points on the part.
My CAD operators say the the "no show" and "no pick" pages are blank and that they have run a "CAD clean" function, but we still get these planes. There seems to be no pattern either as several files from one guy will be good and another bad, so I can' seem to pinpoint operator error.
Division Head, RP&M
Perusahaan Otomobil Nasional Berhad (PROTON)
Hicom Industrial Estate
Batu 3 PO Box 7100
Shah Alam, Selangor D. E.
+60 3 515-3906 direct line
+60 3 515-2380 direct fax
+60 12 277-5770 hand phone
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 - 22:45:11 EEST