Alignment Features and Location vs Position

XactMeasure (and Legacy Dimensions) use the pre-resolved feature so, if you have measured your circles using a particular math type, that is what the calculation is based on. The only time features are re-calculated is when it is an XactMeasure form tolerance (flatness for example) in which case it uses the Chebyshev calculation.

As for alignments, GX (maximum inscribed) is known to be mathematically unstable in certain cases and can lead to poor repeatability. It is especially limited when you have partial arcs rather than full diameters or when the form error of the hole makes it more of an oblong shape than a true circle. Your alignments would definitely be more stable if built from least squares features but you would then have the problem that the X Y & Z locations based on those alignments would not match those in the XactMeasure dimensions reporting GX features - as you have found. That is because the least squares calculation would yield a different centroid and size for the circle than the GX calculation would.

Hey Neil,

thank you for your answer.

One of my Problems is, that even with GX Alignment and GX Features the X,Y,Z is Not matching the Position.

Creating a plane from vector points and use "auto"-settings will result in a best-fit gaussian plane, right ?
for flatness this will be re-calculated to chebyshev (min/max) ?!

Will XactMeasure re-calculate the plane, if used as Datum A, if not.. how should the plane be constructed ?
Is it necessary to construct a "primary datum plane" or "Bestfit - min_sep" and use this for dimensioning ? ( i don´t know, if this options are available in R2017.1 , i just read it in the help file)

As for alignments, GX (maximum inscribed) is known to be mathematically unstable in certain cases and can lead to poor repeatability

Would you recommend to measure GG and GX, then use the GG-Alignment for the probing part for repeatability and GX-Alignment for the dimensioning ?


Thanks.

XactMeasure will report the Chebyshev flatness value, regardless of what math was used to measure the plane. The following screenshot was taken from PC-DMIS 2020 R1 and contains two plane - PLA-1.3 is a Least Squares BFRE plane and PLA-1.4 is a MINMAX (Chebyshev) BFRE plane. If you look at the report, you can see that the XactMeasure flatness results for both planes are identical and match the legacy flatness results for PLA-1.4.