Several years ago (14-JAN-2013) I had an issue where I have two insert mates (both coincident because they are perpendicular) where the part was flipping to one of four different orientations:
I am now in Creo 4.0, and the previous solution, adding to coincident constraints for (2) plane is no longer working. When upgrading to the newer version of Creo, the coincident constraints were changed to "Parallel". There in lies the problem, two planes can be parallel in one of two configurations when they are aligned and when one of the planes if flipped 180 degrees (unless parallel also includes a orientation vector to say what side of the plane is "up").
Any thoughts? Is this a new bug with the constraints changing from coincident to parallel?
Solved! Go to Solution.
I had also made an idea request to correct this:
Yeah, I think the constraints in creo 3 are a step backward, although, "automatic" seems to be a good thing. Still have to play with it more to 100% be convinced. I liked the older "mate".
One thing you might try, is datum planes and surfaces act differently when trying to mate them. Mates on 2 solid surfaces seem to be more stable because creo knows the surface normal always faces out. With datums, it's the brown side now that has the outward normal. This is why my library fasteners always use solid surfaces instead of datums in the pre-defined placement constraints.
I agree it could be more robust. One thing I liked about NX 8.5 was that you could use a "concentric" (I think it was) constraint that would align the center of 2 circles AND make them coplanar. That was nice.
Best of luck!
The weird this is that the generic shows it in the correct configuration, but the instance on one of the four feet is flipped so it is on top of another foot that is already there (and makes it look like one is missing - though it shows up in the model tree).
Are the feet ref patterned from a hole or something? If the 4 radial holes were done as 2 holes thru both sides patterned at 90deg as opposed to 4 separate holes thru one side only patternd at 90deg, that might do some weird things. But I can't imagine an instance flipping if the surface is the same (i.e. from the same feature).
For kicks and giggles, I flipped the parallel mate on this instance to "make it right" and it flipped the generic to "make it wrong". Gotta love PTC.