Does anyone have an idea why a Ø1.7714 +.0003/-.0002 would show 1.7727 when I use it in a true position, but show 1.7715 when I use it in a standard location box? Both ways I'm using the same measures cylinder (scanned 3 levels). I'm running PCD v2022.1
Correct (per the Y14.5 standard 2009 and forward) is now Constrained Least Squares (least squares solution and then adjust size until the mathematical feature fits in/around all measured data). Constrained Least Squares is not the same as max/min in/circumscribed. It is not the same as least squares.
Chair of the Y14.5.1 (math for solving features used in Y14.5) committee works for Hexagon, and GeoTol is going to strictly adhere to the standard.
Why? It elegantly solves the problem that occurred when a part was not near form and you try to do a mincircum/maxinscribed solution allowing multiple solutions.
Worst case, you have an actual dome that is supposed to be a plane. You can pick multiple 3-point groups that allow for solution and "sit" on a datum simulator, but they all kick the part in different orientations.
If that happened to be Datum A, you're never going to get correlation to anyone else. CL2 eliminates almost all of that problem, though not on a perfect dome that should have been a plane, but c'est la vie.
In PcDmis 2022.1 it is going to be doing this for ALL GeoTol that is ASME whether you want it to or not.
If you disable that, you aren't per standard, and it won't say you are compliant with ASME.
As others said, in a normal dimensional output, it will use whatever algorithm you want.
If you are having a correlation problem with customers/suppliers, you can do legacy GD&T and not have the problem.
FYI - ISO is probably going to be going to CL2 as well in the not-so-distant future.
