Are you familiar with the concept of "Revision Rolling"? This is the scenario in which a change of a revision to a child part, way down at the bottom of the multi-level assy, causes the child's immediate parent's revision to change, and the parent's immediate parents' revs to change, and so on and so forth upward thru the hierarchy until you reach the top level.
In the literature, this is a bad thing to do, and the application of "interchangeability" rules allows you to decide when you need to roll the revision on upper level assys vs. when you can keep them at their existing rev (such as when a change is "interchangeable")
you have a bracket, with no holes in it, released at Rev A. It is a child part of an assy, and this assy is used in several upper level assy's. Drawings for these assy's exist, and there are views that show the bracket as a member of the assy. The drawings of the assys are also released at rev A.
If you then add a hole to the bracket, but this change is interchangeable with the bracket version without the hole, one simply changes the rev of the bracket to B and releases it. No need to roll the revision on upper level assy's, at least according to the rules I've read. BUT - what about the drawings of the intermediate assys? You released them at rev A and they remain at A, and those dwgs have views that include the bracket, now if you send the drawings and models to a supplier, they are going to open up of those same rev A drawings, only this time they will show the bracket with the hole in it? Will the suppliers freak out and say "hey! you sent us a rev A of this assy last week, and in this view the bracket is shown and there is no hole in it, but this week you send us models and files and I open up rev A of the same dwg and the bracket has a hole in it?"
sorry for the long msg ,but we are struggling with this issue and any comments and advice is appreciated.
MIT Lincoln Laboratory