What exactly is this implying? I've seen a profile composite to ABC and A, ABC and AB, but not A and A. PCDMIS seems to be calculating it but I can't make sense of it.
If I had to guess, .004 to A and .002 form only while still pinned to datum A.
The argument could be made that the secondary frame is establishing error from left side to right side (or vice versa) once the right side existed. I've never seen that on a position, but technically, it is allowed.
If this was holes, they have a position of .004 to datum A (generally akin to perpindicular) and all the holes must be within .002 of each other while still perpendicular to datum A.
So, sort of, if those two sides are 120° apart, one side must be 30° off of datum A within an angularity of .004, while the other side must match the first at 120° within an angularity of .002. Sort of, you can't substitute angularity for profile of a surface, but that's the concept.
I've never seen it done that way with profile (surface or line), and I don't like the look of it. I am used to the primary having more degrees of freedom controlled and then the secondary only to A.
I would personally ask what they (customer's receiving inspection and engineering) are wanting done, explain what I think they are asking for, and explain I've never actually seen it done outside of the standard itself so I want to try and match their inspection and give them what they want/need.
I know this is a napkin sketch, but I'm assuming this is supposed to portray a composite profile.
Composite dimensions are checked to the datums *rotationally* while the upper segment is checked *rotationally and translationally*.
So in this case, the upper segment wants to ensure that if you leveled and origined to Datum A, the entire surface profile is within tolerance zone normal to the surface.
The lower segment want to ensure that if you leveled to Datum A, but origined to the feature itself, the entire surface profile is within the tolerance zone normal to the surface. This is essentially a form only profile tolerance with one extra rotational constraint (Datum A).
*Thought experiment* Imagine you have the same call-out you have on your drawing, but the profile surface is a parallel plane in relation to Datum A. For the upper segment the profile call-out would be checking both the distance of that surface to Datum A, plus parallelism to Datum A and flatness all at the same time.
The lower segment would only be checking parallelism to Datum A and flatness. When I said before that this is form only callout with one extra constraint, form only would be flatness only in this example.
So, if I were to a dimension this in legacy in PC DMIS, (assuming this is a 3D triangle) I would...
Top segment- Level and origin to Datum A, create a best fit full 2D (both rotation and translations) about the axis of the level using the features used to create the profile surface. Then dimension profile.
Lower segment- Level to Datum A, create a best fit alignment that locks rotation about the axis of the level and also best fits all 3 translational axis to the features used to create the profile surface. Then dimension profile.
*EDIT* Composites can be confusing, but the ASME standard has some pretty nice visual representations of multiple segment composites. Your drawing could technically have a 3rd composite segment using no Datums (form only) and it would still be proper.
The customer for this project generally has a lot of problems with prints. I've heard a lot of new designers. Using Geotol it's showing the the .004 OOT and the .002 in tolerance. So I am assuming it's calculating it correctly because the part is a little thick in some spots.
