Z axis range limits for meauring

[[Ri...]]

We ran into an issue by accident the other day.
One of our Contura G2's had a part setup to run. The setup required the part to rest on top of a 5" 3 jaw chuck, the operator used the previous setup which is a 4" three jaw chuck stacked on top of the 5" chuck, adding about 2.20 or so inches to the height in the Z+ axis.
That's where it got weird.
There are two Coaxiality requirements and the results were far from what was expected. (Inches)
Item #32 was .0236"
Item #42 was .0197"
Upper tolerance for both is .0020"

Our programs use a minimum of two Space Points that are measured at the beginning of the program and again at the end of the program with a tolerance of ±.0003", this lets the operator know whether the part had moved during the CMM program execution.
On this particular run both points recorded "Zero" deviation, the part had not moved.

We then removed the " chuck, placed the part on the 5" chuck per the setup instructions and reran the same part/program.
The results were as expected.
Item #32 was .0015"
Item #2 was .0006"

We repeated the same setup with the same program on our Prismo, although the results were less errant than the Contura, they were still far from expected.
Item #32 was .0078"
Item #2 .0053

After using only the 5" chuck.
Item #32 was .0019"
Item #2 was .0004"

See attached images for each set up.

Is there a limitation fro Z axis hieght while scanning?
Or do nearly if not all of our CMM's, after recent calibration, have the same problem? (Which I would find hard to believe)

Stacked Chuck.jpgSingle Chuck.jpg

[[Da...]]

What are you using for your base alignment ? Most importantly what is your spatial alignment based on?

[[Ow...]]

Along with what Dave is suggesting on the alignment features, are you looping the alignment?
Are the sizes of the shaft at the datum features the same size each time, maybe within spec but different?
It could be an optical illusion but, it looks like the part is not as square to the z-axis (viewed beside the ruler you're using to reference height) when it's in both chucks and if so, them small probes could be shanking out? If you loop the alignment at least twice, it should remove that possibility.

[[Da...]]

As a Reality check can you put a .0001 test indicator on the shaft near the top and lightly touch the shaft, If you see any movement at all you'll have to rethink your work holding. Owen is right, If you're shanking that would kill the reading too.

[[Ri...]]

Please sign in to view this quote.

The large surface at the base of the part is Spatial.
And my base alignment means nothing. We have a rule, Base Alignment features are not used in any measurements, we create a "CNC Alignment" that is used for all measured features. Base alignments are just information to the CMM that something occupies space in the volumetric area.

[[Ri...]]

Please sign in to view this quote.

The part is square to the CMM. All Probes are aligned to .0010 per inch of the CMM Z axis movement.

[[Ri...]]

Please sign in to view this quote.

No movement recorded. I used the stylus to probe points on the upper diameter. The chuck clamps on to a 2.376 diameter hub, its held quite well.
For this part to deviate approximately .0090" or more per axis, I would think I could see it move (light reflection).

[[Da...]]

Does it repeat if you check it twice without removing the part ?

[[Er...]]

In theory the machine has less ridgity the higher you raise the pinole since the mass is shifted from the bottom to the top. But that shall not be the problem, unless there something wrong with the machine.
With that said, I have encounter nothing but problems when fixturing similar parts like that. I really dont know why. But now days I always use v-block or if dub holes are available, I use them.

I remember some German from Zeiss I talked to at the control fair a couple of years ago. He said to always try to be in the middle of the measuring volume and as close to the diabase as possible. Im not saying thats your problem, but I always try to do it like that if possible.

[[Te...]]

Please sign in to view this quote.

Well I agree with your statement about the Base Alignment for the most part, be careful when you say it means nothing. The base alignment still needs to be stable even if you are not using it to measure results from, it still is the bases for your entire measuring plan.

I to always use a secondary alignment for all dimension because it gives much better results but I still make sure my base alignment is as stable as possible for good repeatable measurements. I always loop the base alignment to, even though most are simple 3-2-1 alignments (3 point Plane, 2 point Line, and 1 Point).

[[Ri...]]

Please sign in to view this quote.

It repeats, with and without moving the part.

[[Ri...]]

Please sign in to view this quote.

The base alignments are stable. The fixturing is stable. In our programs the Base Alignment has no basis other than determining there is an object in the space of the working volume of the CMM.
We have many programs that utilize the stacked 3 jaw chucks, this part is the only one that reaches the 18 inch height area.

[[Da...]]

last question, Do you get the same diameter size on the same part in both chuck set ups, and do they match up with a micrometer reading on the same diameter ? ( the diameter farthest away from your chucking diameter )

[[Ri...]]

Please sign in to view this quote.

I will have to rerun the part, the report only showed the location (Coaxiality).