Profile Confusion

[[Ky...]]

Datum A is bottom of my part. Datum B is X+ side. Datum C is Y+ side. We have a limited dimension drawing. So everything else is by profile or position tolerance. The thickness of the part is .750 per the model. The thickness measuring with Micro-hite is .753. The thickness from CMM is .759. Profile of Top from Datum A B C is .0206. Flatness is not called for but I checked it and is .006.
Is the flatness and being .003 from nominal causing the profile callout of .0206. Also is the flatness plus the .003 causing the .759 thickness. I need to be able to explain to the floor why their part measures good on the granite table but bad on CMM. 😱

Thanks for any help
Kyle

[[Jo...]]

Please sign in to view this quote.

.759-.750 nominal = .009. Profile worstx2 would be .018. You may have some outliers in there that bump to .0206 or it's not parallel

[[Jo...]]

Used CAD evaluation and you can SEE where and how much using gradient color scale.

[[Da...]]

The difference between the floor inspection and the CMM inspection is the number of points and where those points are located.
If they are taking a few points in the center and the part is cupped or their points are around the edge and the part is bowed as a
couple of examples. Flatness will have an effect on the profile but I don't know to what degree. There are many factors to consider
as to what the root cause of the flatness issue is, if it's being milled or ground, is it held in a fixture, is the fixture flat, feed and speed rates. I would have them move the size closer to nominal and check points at various locations to closer resemble the CMM.

[[Ma...]]

If you are scanning, surface roughness can cause 'bouncing', skewing the results.

Might want to try it with touch points if you suspect that could be the problem.

[[Jo...]]

I don't know if this is solved yet or not but I would need to know how they are checking on the shop floor vs how the CMM is evaluating. The surface plate is going to simulate an outer tangential evaluation if they are checking from the surface plate to the top of the part. If they are just pinching the part with calipers then the thickness number will be smaller then a surface plate to top surface measurement, especially if you have flatness error of .006. If you are evaluating LSQ on the surfaces with the CMM then you will get a smaller value than the shop floor's surface plate measurement. What measurements is the floor getting?