verify keyway position

On both sides make lines then make a mid line, or do 2 Planes and a mid plane. Or you can do a constructed width, but I have no experience with them.

Constructed width seems about right using those two planes, or depending on the version if available there is a "Notch Slot" in the auto features. 

But if its position callout, doesnt it require at least one Datum? 

I was confused about that too, and I was going to mention that. But I opened PC-DMIS and tried to add a position without a datum, and it didn't complain. I guess without a Datum is will use the alignment? So I just brushed that aside in my head, but I was going to say it feels like maybe it should be to datum A. Or another datum

strange. I have seen TP on a primary datum but never a feature that's not. Seems like a "form only" result. I wonder how PC-Dmis would calculate it and what the result would be? Corazon Rosinante can you post the measured result when you have it?

I've generated two planes from both lateral surfaces and I've found the midplane. The coordinate of the midplane is -0.118, that means that the toolerance is positively verified.

Now I have two problems:

1) we have to check 50 pcs. How can I place the the part over the CMM table e block it? In order to not redo the alignment...The part has no holes or other references to be used to palce it all the times in the same way.

2) those pcs was isolated from the production lot, due to a suspect non-conforming Total run-out toolerance. I evaluated it and I found out that the total run-out is about  0.239. My doubt is: could the midplane of the keyway be just 0.118 from theroric center (the center of the A Datum) with a 0.239 total run-out? Is the 0.118 a reliable value?

Due to 1) and 2) I checked just one piece...meanwhile I'm looking for something that allows me to place the part over the table and throw the automatic code without staying there to check every second...the lateral planes are really small...0.5÷0.8 mm from the actual alignment can cause a collision...

Thank you very much for all your support...I'll wait for it

I've generated two planes from both lateral surfaces and I've found the midplane. The coordinate of the midplane is -0.118, that means that the toolerance is positively verified.

Now I have two problems:

1) we have to check 50 pcs. How can I place the the part over the CMM table e block it? In order to not redo the alignment...The part has no holes or other references to be used to palce it all the times in the same way.

2) those pcs was isolated from the production lot, due to a suspect non-conforming Total run-out toolerance. I evaluated it and I found out that the total run-out is about  0.239. My doubt is: could the midplane of the keyway be just 0.118 from theroric center (the center of the A Datum) with a 0.239 total run-out? Is the 0.118 a reliable value?

Due to 1) and 2) I checked just one piece...meanwhile I'm looking for something that allows me to place the part over the table and throw the automatic code without staying there to check every second...the lateral planes are really small...0.5÷0.8 mm from the actual alignment can cause a collision...

Thank you very much for all your support...I'll wait for it

vpt.se 

Plane, level and zero
Circle, zero in the other two axises
One point on side plane in keyway
One point on the opposite side plane in the keyway
Midpoint between these
Align axis using circle and point

Done

Unless you can reliably place the part in the correct place everytime you will need to do the above alignment per part.