I've been away for a bit, taking care of paperwork. Gov't & lawyers ... what fun.
Yes, J-P, we've discussed this a few times, and, if done properly, I don't have a problem with it when it's appropriate. It's the old cart before the horse thing. (!) You Simulate your datums. (2) You probe off those simulator surfaces to verify features on the actual part.
No ambiguity, and no problem with the original drawing. Asking these questions is reasonable, and what this forum is for. I am surprised, however, that A.K.'s reference material is automatically accepted without question here. I have found other things in his materials that, when asked about, turned out to be opinions rather than standard-based. Some is based on practices with his prior employer. This is not unusual, as many subject-matter-experts ingrain their histories and biases into their material (I ...wait for it... am biased towards profile controls...shocker, right?!), so it's not a negative against A.K., just a reminder that all materials should be challenged in due course. When people put opinions into reference materials, they should post it as such, or as "rules of thumb". By calling it a "Profile Datum Rule", it implies that it is from the standard (Y14.5) ... but it is not, it is merely his preference.
The most common question that I get about perceived self-referencing datums is "how do you check the datum feature surface if it is engaged with a datum simulator?". When it is appropriate and necessary to use such datums, the datum feature simulators must be able to disengage from the datum features without changing the setup; this may involve auxilliary clamps, surrogate datum features, etc., and therefore generally more expensive fixturing.
Jim Sykes, P.Eng, GDTP-S
Profile Services
TecEase, Inc.
