Gps or caliper?

[pernilla]

I'd like to ask all you experienced metrologists about this one thing that the stubborn operators at my place of work claim to know everything about, because it's the way they've always done, long before geometrical product specifications (which is a blur to them) were invented and long before CMM's were invented...

Say you have a part, the bottom is ref A, 3 point locations specified on the drawing (my primary alignment).
All dimensions in Z-axis originates from this plane. The dimensions in Z-axis consists of, say, 26 spotfaces all dimensioned, say, 10.3 +/- 0.1 mm.

Would you measure these 26 spotfaces point-to-point with the exact same X and Y coordinates on the bottom and on the spotfaces to get the best measured values (as you would measure with a caliper)?

Or would you use reference A, by just evaluating the Z-axis location on these 26 individual spotfaces?

Or would you measure this in a totally different way?

[JEFMAN]

I would say it depends on the drawing...
But, without seeing it, if the part is well made, as the tol is very large, there should't be a big difference between both methods.
With a tigher tol, I would measure little planes instead of points.

If operators know everything, they should know that using a caliper can give bad results because of Abbe error, and another because of cosine error...

https://en.wikipedia.org/wiki/Cosine_error

Abbe error :

The-Abbe-error-is-the-product-of-the-perpendicular-distance-of-the-scale-from-the.png

[Douglas]

and another yet because they need to clean the fkn thing

[pernilla]

I know all about Abbe error and cosine error when using a caliper, I find that stuff interesting.

The drawing doesn't specify much about this, here's what it looks like:


1.png

But I've dealt with dimensions like this before and I always find myself doing what the operators want in the end, not knowing if it's the right way to do it.

[Matthew D. Hoedeman]

thickness, as in, point-2-point, with that callout. Unless there is something else, somewhere on that print, with a callout to datums or possibly a flatness callout.

[pernilla]

No, there's a callout for flatness at the bottom plane but not on these small planes. And there's a callout for profile of a surface 0.6 to ABC, on the entire part.

So point-2-point then? Is that because of the "28x" part of the callout?

[Matthew D. Hoedeman]

no, because it is calling out a specific thickness at specific locations, that would be opposable points. It is a refinement of the 0.6prof callout, each side can vary +/-0.3 from nominal, but at those locations it must also pass that 'thickness' callout. So if the 'bottom' surface is 'low' 0.2, then the 'top' surface must also be 'low' but can be 0.1 to 0.3 low (0.3 being the max no matter what). IF the 'bottom' is low 0.3, then the top can only be 0.2 to 0.3 low.

[pernilla]

That clarifies things. Thank you!