
One of the reasons Adaptivity so often fails is this: Inventor doesn’t know which constraints are meant to be actual Assembling constraints, and which are meant to drive an Adaptive part’s geometry.
Sometimes Inventor gets it right, but sometimes it doesn’t—especially as assemblies get larger and have more and more constraints. An Adaptive relationship that solves fine when there are only a few parts in the assembly can suddenly become sick when lots of parts are added and constrained, even if those parts are in no way constrained to the Adaptive part.
I’ve been told many times that Adaptivity is meant to be more of a “helper feature” during the initial design, and should be turned off once an Adaptive part is the right size.
I find this to be a completely unacceptable answer. Why have Adaptivity at all if it can’t even adapt??
In my experience with Adaptivity, it’s robust enough if you’re careful to build assemblies with only one solution, and as long as you don’t have too many parts.
If we could designate specific constraints as constraints which are ONLY meant to drive a part’s adaptive features, and are NOT meant to try to change the position of the part on either side of the constraint (or vice versa), then I think Adaptivity could be 100% robust when carefully-designed assemblies are made.
The ability to specify the direction of the adaptivity-driving constraints would be a bonus, because then we could possibly even push geometry between two Adaptive parts and still be 100% robust, as the constraint would explicitly define which part drives which for each feature.