Question regarding DRF establishment w/ offset plane

[[Th...]]

In advance, I apologize if my artistic skill causes you physical pain.

Earlier today, the designer in charge of this part took issue with how I'm evaluating it on the CMM. The way this is set up, I have a ~25mm cylinder functioning as my B datum feature. The A datum feature is the plane that faces the viewer, which, for the purpose of this question, is ignored. Essentially this evaluation is taking place on a 2D plane. Some distance away, I have a short planar feature tagged as the C datum feature. This feature is given a basic distance of 9mm from B. Then, there's an all-around profile tolerance of 2mm. Every GD&T callout is RFS.

There's some kind of angular deviation with the C datum feature that's causing the profile actual to fail, where the part is rotated clockwise about the B axis. I sent this to the designer, and he countered that the C datum feature should only rotate the coordinate system until it contacts a theoretical plane at 9mm distant from the B axis, where the CMM is (at least, I assume) rotating so that the calculated feature is parallel to a coordinate axis. This would, incidentally, probably bring the profile back in tolerance.

Is my interpretation correct or incorect? I have no formal GD&T training, and while I was able to find some diagrams in Y14.5-2009 regarding offset planes, the plain standard isn't exactly designed for guided learning. Plus, with all the designers WFH, communication is terribly slow.

IMG_20200827_181133.jpg

[[Ja...]]

When you rotate to, or originate from, a plane, Calypso uses that feature's "origin" (which will be in one of the 4 corners, when using a plane). Depending on what that particular corner is doing, it can greatly affect your alignment. Try measuring Datum C using a poly-line, say step-width of 0.3, speed of 3 or 5 with filter and outlier removal on of course, then recall those points into a singular point feature. That will give you an average of all points, smack dab in the middle of your surface (C), or the "centroid" of the surface. Now, use your new centroid for alignment rotation.

Oh, also, you may need to use the "special" section of the alignment window to rotate your alignment back to where it's parallel with C, since rotating to a point will clock that axis to said point.

[[Th...]]

I should clarify, since it's keeping me awake: the question isn't so much related to the calculation of the C feature as it is related to the proper rotation that should be applied to the profile DRF. Should the coordinate system be rotated so that C is parallel to a coordinate axis, regardless of the plane's deviation from that ideal 9mm distance, or should the coordinate system be rotated so that at least one point on C contacts a theoretical plane at that ideal 9mm distance, regardless of the parallelism (or lack thereof) of C to any coordinate axis? Looking at ASME 14.5-2009 figure 4-31 (sections 4.16.5 and 4.16.6), it seems to me that rotating the alignment parallel to C takes priority, since prioritizing the basic distance would require the [BSC] modifier. However, since I can't come up with a justification of my opinion, I can't be certain that I'm right at all. I'm half-convinced I'm misunderstanding some fundamental concept here.

[[An...]]

Please sign in to view this quote.

The designer decides.
It is according Pic. 4-31 (b).

[[Ri...]]

See this thread. You do need to create the offset.

viewtopic.php?f=13&t=3239

[[Cl...]]

Please sign in to view this quote.

[[Ri...]]

Please sign in to view this quote.

If you use a point, you can do it that way.

[[Th...]]

Is the discrete probing point necessary? If I set a theoretical point on that plane and recalled the plane's feature points into it, would that provide a functionally accurate point to clock to and rotate to the basic distance?

[[Me...]]

Please sign in to view this quote.

The latter. Imagine a hard gage for this. It would level to A, a perfect expanding cylinder would locate B, and the part would rotate until it contacted C. If the part is failing then whatever is controlling C relative to A/B isn't doing it's job. Does C fail it's callout?

[[Th...]]

Please sign in to view this quote.

This was ultimately the problem. I wasn't considering the properties of a hard gage. I was so focused on the idea that C needed to be parallel that it didn't occur to me the part would be embedded in the wall of the gage if I did. Unfortunately, other things are occupying my CMM right now, so I can't verify the full part again, but I managed to re-check one feature with the C datum points recalled into a single point, then rotated to the basic distance. The profile results were much closer to nominal.

[[Th...]]

Sorry to bump this thread, but I now have a question regarding the process of establishing the tertiary point. I looked at the thread linked earlier where the user in question was taking a physical probing point on their C surface. My actual C datum feature is long enough that I'm not really comfortable just probing a point in the middle of the feature. Rather, I want to consider multiple points along its surface and find the one that would make contact with a hard gauge wall. If I have a C datum feature that I'm scanning a grid or polyline on, and I slap down a maximum coordinate construction to pull the MMC point, I can use it in a secondary alignment and offset the rotation to my basic distance. Is there anything wrong with this approach?

[[An...]]

Please sign in to view this quote.

See attached.

Contribution_10_09_2020.pdf