I'm currently looking at how models are assembled where I work.
The current method was devised by an ex-solidworks user, and he has tried to wrangle things to impose a solidworks style solution.
Copying sketches into multiple parts, so you need to manually change it in every part or errors, clashes and model breakage occurs.
Creating assemblies from parts, fine, creating a master assembly of all sub-assemblies, yup, but then creating an assembly containing just the master assembly just for saving different representations??? Why? Why not have the different representations in the master assembly? Creating an assembly containing just of a sub-assembly for representations? What?!
I'm not terribly happy with how it currently stands.
These are large and complex structures and therefore large and complex models
So my question (without divulging how I think it should be done (derived parts and/or skeletal), to avoid bias and in case I'm way off base)
How do you do it?