Help Please, correlation issues

[[Lo...]]

I am having correlation issues between our Contura G2 and our new O-Inspect. Some results exceeds 10% of our characteristic’s tolerance. Our tightest tolerance, on average is .03(mm). The Contura G2 has a XT scanning head while the O-Inspect has the XXT. We are also unable to use the same probes, diameters and lengths. The Contura and O-Inspect are also in separate environments, both controlled, in different areas of the facility. The Contura is in the lab, away from any machine center. The O-Inspect is in a hut next to the machines. Basically we have a lot stacked against us and I have exhausted my knowledge base. I understand that the tighter tolerance dimensions can use up to 30%. I just think I need it broken down stupid style on how much difference is "normal". I've been at this for years but this is the first company that I have had more than one CMM. 😱

[[Cl...]]

We have a similar situation, but don't see a 10 % difference. How much temperature difference is there between the Contura environment and the O-inspect environment? Have you checked for excessive vibration caused by surrounding machines?
Those two variables could be your biggest problem. Also bear in mind that the XT is much more accurate than the XXT.

[[Lo...]]

Temp is on point, within 2° and I don't have a way to measure the vibration. It's getting the Production Manager to understand what is acceptable deviation between the two CMMs. He is under the assumption that it has to be better than 10% but that is unrealistic. And telling them to create a stable machining process, so the deviation in inspection methods is still in tolerance, just pisses him off.

[[Wo...]]

I don't know what are the official numbers, but I have two machines: Accura/Vast XT and Accura2/Vast XTR and the differences are minimal, usually below 0.0001".

[[Ow...]]

Room temperature is one thing, part temperature is another.
I once experienced something like you are discussing except it was CMM vs manual gauging.
They'd check it coming right off the cutting machining center where coolant temperatures were heated thru out the day by all the cutting going on. When given an adequate amount of time for the part to acclimate to 68-70°, it checked fine. Check it when it was fist brought to the CMM for 1st article inspection and then much later and it would check much different. Using the magnetic temperature-reading compensation wires helped a bit but, because they were fairly large parts, the part temperature varied though different thickness of the part, long story-short, letting the part acclimate worked best.

The other thing to consider is even though the probes may be exactly the same, it's two different types of scanning systems and one (the xxt) is much more susceptible to probe deviation with longer, skinnier shaft probes and the amount of weight allowed on the XXT is much, much, less than the XT.

[[Lo...]]

I've been using the same part. I goes from my desk to the CMM and back to my desk. As far as values, I have been seeing up to .015(mm) differences. Only on a, very small, group of dimensions. the vast majority are better than .005(mm). But because of these few "flyers" I can't use the new machine for this project. Large paper weight right now. This is definitely not checking apples to apples right now, because of the probe's diameters, lengths and probing heads themselves. Frustrated and at a complete loss.
I have made the call to the Zeiss dealer we purchased the machine from. They may be able to point me in the right direction or tell management about capabilities of this process. Hopefully the decades of combined experience, in this forum, can help me out. You guys have helped me solve countless issues, over the years.

[[Ow...]]

What type of material is the parts made of? As you probably already know, aluminum's thermal expansion is highly subject to temperatures.
Is the part checked in a fixture that constrains it or in a free state?
Do you get the same differences when checking a certified ring gauge of about the same size between both CMM's?

[[Lo...]]

It's GMW17942 A2 Gear Steel according to the print. The part is a turned part, so it is set upon 4 posts, on one of the journal diameters and in a free state. The part is also 235mm long and Ø72mm so it isn't going anywhere. I ran the test again today, definition of insanity. Filtering, evaluations, nothing helps. Next step is to have Zeiss come in and check the head, we are under warranty. But being on an O-Inspect, the table moves. Could the weight of the part have such an effect?

[[Da...]]

So I ran some tests a few weeks ago between two machines. On a 4 inch gauge block, I saw sub-micron differences between two machines. (they were in the same room)

First: make sure that your speeds are good for the XXT. The first thing I do with an XXT giving poor data is slow it down. 2-3mm/s is a good place to start. Slower if the surface finish or form is poor. If the finish is really bad, use single points. In my experience, XXT's don't love the vibration caused by rough surface finish, or running too fast.

Second: you really want to compare apples to apples. At least use the same stylus diameters. I would make a stylus system on the contura that mirrors what you have on the O-inspect. Then measure the part with both systems, back to back. Then measure the part on the O-inspect, and compare the results.

If your plant is having down time over Christmas, I would use the opportunity to measure the parts right at the end of the last day, when everything is off and quiet, then again right when you come back, before anything else turns on, and a third time after everything is running. I have the feeling that the smaller O-inspects are a lot more susceptible to vibration than heavier machines.

[[Lo...]]

I used a 3 inch block to measure the length, using the star probes. And the problem is one of the probes on the Contura, so now there are other parts in question. Management expected to see the issue in the new O-Inspect, because how can our workhorse CMM ever be incorrect, LOL. Well now I have even more work in front of me. Thank you, all, for your advice. Merry Christmas.