David,
Where was the original section created, in the assembly? If you created a section in the assembly, then created curves using that section, then activated your part and created geometry using "use edge" from those curves, from the subsequent cross-section, then you have created a circular reference based not only on the inter-relationshipfrom the Assembly to the Part, your CRC is alsoeffected by the placement and relative model tree position of the asembly datum plane used to create the cross-section. That is why the assembly doesn't complain if you move the curves around to the other "insert location". the CRC is temporarily non-existent in one position as opposed to the other.Whoa...a mouthful!
If the "curves" you require are that important for the success of the assembly, make them the driving feature(s) and not a resultant entity buried within an assembly.
Dean
In Reply to David Schenken:
A minor puzzle - I have a model using datum curves created from cross-sections. I don't want to use all the section entities, and am using 'use edge' in a sketch to capture the ones I want.
It appears the section is being evaluated after the entire assembly, so the section-cut references parts that come after the 'use edge' curves.
If it was possible - I'd fix the regeneration location of the xsection before the references to the curves are created. It would also be OK if I could change the component list intersected by the xsection, as an assembly cut can be handled.Even though the curves are generated from a section, the section doesn't appear in the Program for the model.
If the Insert point is moved after the curve generation feature (suppressing all later parts) and the item listed as part of the circle, the assembly regens without complaint. Were it a real circular reference, it should list the reference as missing.
