Reporting datumless profile with unilateral tolerance

[RandomJerk]

I'm ok with a datumless profile, and I'm OK with unilateral tolerancing for profile, but having both is causing me a slight headache. Would it be as simple as best fitting the shape with no restrictions then reporting form and location +.003/-.000? Doing form only doesn't protect me from understock...

Seem reasonable?

[Mike Ruff]

I have a part right now with the same requirements. I used Xact and dimensioned it exactly per the print. I'm not sure quite how it calculates the results, but it seemed to work. I'd be interested to get some insight from others on here...

[BKulpa]

How about doing a best fit alignment of your profile surface and then dimensioning profile form and location.

[RandomJerk]

"Would it be as simple as best fitting the shape with no restrictions then reporting form and location +.003/-.000?"

So that statement does indeed seem correct?

[SingularitY]

A couple solutions.
Like BKulpa stated. You can use a best fit alignment to all features in the profile, not the scanned set, and dimension form and location. 3D best fit for anything 3D and you could probably get away with 2D translate and rotate if the features or scan is one level (2D.)
You could also dimension form only with max/min activated reporting to none. Extract the max and min values and insert them into the measured X,Y, or Z location of 2 generic points, make sure nominals of the same axis locations are set to zero, and dimension the location of those points in the same axis using unilateral tolerancing. If you like a cleaner look you can use one generic point and have the Max in one axis and Min in the other and dimension both axis as a location with unilateral tolerancing.