How to identify faults in imported .stp files?

owen_sparks

Hi folks.

The last round of updates allowed parts with "faults" to be imported/translated.  

This a big step forward as previously OS decided that a high proportion of the models I wished to import (from Creo) were "faulty" and just bombed out.  (Despite the CAD in question having had roughly 50,000 production parts made from it.)

Now I'm not saying the CAD wasn't faulty, it may well be, but other systems can read it just fine.

I'd like to develop a method whereby our departments can share data without issue.  With this in mind once I have the model imported is it possible to identify which feature OS has a problem with?  I tried splitting one of the parts with a bunch of planes hoping one part would stay red (in the part tree) and the rest would turn black but no luck.

Any suggestions?

Cheers,

Owen S.

owen_sparks

Just out of interest tried exporting the faulty part as a STEP and re-importing and OS reported the errors.

carl_von_ayres

My only thoughts are 1) use section view to look though your model for overlaps, will be highlighted red. 2) create a seperate part that covers your part and do a booleon subract, this should show you the voids in your original part. Could also turn on curvature view to see if there are areas of deformation.

owen_sparks

carl_von_ayres said:

My only thoughts are 1) use section view to look though your model for overlaps, will be highlighted red. 2) create a seperate part that covers your part and do a booleon subract, this should show you the voids in your original part. Could also turn on curvature view to see if there are areas of deformation.

Thanks, I'll give each of those suggestions a try.

Cheers,

Owen S.

owen_sparks

Hmm interesting.

(1) Section view inspection yielded nothing odd looking.

(2) Boolean of faulty part from a bounding solid yielded a good negative of the faulty part.

(3) A 2nd boolean of the negative from another bounding solid yielded a part that seems identical to my original faulty part but that is deemed "good" by OS.

(4) Transform copy of the original faulty part yielded another faulty part.

I feel a featurescript coming on to automate (3) 

Thanks for the pointers,

Owen S.

owen_sparks

Bad on the left, good on the right:-

owen_sparks

Just out of interest tried exporting the faulty part as a STEP and re-importing and OS reported the errors.