Axial Runout using normal vector

[[To...]]

Customer is measuring axial run-out of a plane to a short cylinder and is failing. Their customer checks the parts using same methodology and it is passing. The only difference is the final customer has the "Normal Vector" constraint active for the cylinder. Can anyone explain what this constraint is actually doing? Is it valid? This is actually the 2nd time I've heard of someone doing this.

[[Mi...]]

I'd like to see somebody else chime in but based on what I think I know, this wouldn't be valid as the true vector of the cylinder would affect the runout and therefore shouldn't be constrained (locked to perfect with the base alignment)

I often use constrain to normal vector for short cylinders and cones to help but not if a control like that is in place.

That said, if the final customer deems it correct - get them to put the method in writing and everybody do it that way. lol

[[To...]]

Please sign in to view this quote.

I told supplying customer "do whatever the customer tells you to do", but I agree that getting it in writing is smart.

[[Da...]]

Please sign in to view this quote.

As I see it using Normal Vector will lock that cylinder to belonging spatial feature in its alignment.
And there is a great chance that alignments from both parties differ.
I wonder if remeasuring (closer) to the cylinder axis would help and also wondering if it would make more sense to control perpendicularity of the cylinder to that plane instead of run-out.

[[Ri...]]

Ah, my favorite - do whatever makes the part pass. Haha.

As stated, locking the Vector means that the Cylinder axis is perfectly Perpendicular/Parallel to whatever Space Axis from the Alignment that the feature is measured in.

The only time I would lock a vector would be in the case of a short cone that I was trying to pull an angle from. I have seen people constrain threaded hole cylinders before and this can be a hit or miss. You can always test this yourself and if you see the Form error of the feature go up, then it is incorrect.

I wouldn't suggest swapping to Cylinder Perpendicularity to the Datum Plane unless that's actually closer to the functionality of the part.