Profile measurements - when is it appropriate to use best fit alignments?

[[Ni...]]

Good afternoon,

At my organization we do a TON of profile measurements. The vast majority of our part surfaces are controlled by profile callouts to specified DATUMs (typically A/B/C.) I don't know much about best fit alignments other than I know when Calypso calculates a best fit alignment that causes the data point set to be manipulated to match the CAD geometry as best as possible. I will play around with creating a best fit alignment for features that are checking 'out of tolerance' just to see what the effect is, but we typically just verify the surface as best we can with other measurement means and move on. I've seen more often than not that changing a feature/characteristic to a best fit alignment vs it's DATUM alignment makes the measurement 'closer to nominal.' What is the best practice or what are typical situations in which utilizing a best fit alignment is appropriate or possibly a better representation of a feature?

 

Below is an example Im looking at now (Image one is NO best fit, Image two USES best fit.) - while the best fit feature is still out of tolerance, it is much closer to not being out.

 

 

[[Ma...]]

When to use or not use bestfit is determined by drawing. If there is no datum, then you can use bestfit for sure ( ideally by using freeform and selecting no alignment ).
With more datums you are unchecking translation/rotation due to taking freedom from datum.

Also if you are using curve, don't forget to click on button to close curve if it's meant to be closed ( like in this case can be ).