Calling All CMM Operators

[tharding247365]

Long story short, we only have 1 person in the company that knows how to program and use the CMM. This is frustrating because when we have conflicting results, there is nobody else to bounce ideas off of, and double check results.

There was a recent conflict about CMM vs Bench Checking parts, and who was right. We ended up sending the said part to a vendor, for a 3rd party CMM result. Their CMM result agreed with our bench check, and conflicted with our CMM.

The CMM operator said the following:

3rd party CMM set up is based on the fixture the part is set on for CMM inspection. Our CMM set up is based on the part already placed on the fixture. Late last week we sent an inquiry to contact Zeiss CMM technical support requesting their thoughts on the validity of the two CMM set ups. Intuitively, our set up based on the part seems like it would provide for more consistent and accurate measurements. (If CMM measurements are based off the fixture, any variability in part placement on the fixture could show up in the measurement.) We wanted to reach out to the experts to confirm. I will send further information when available.


Does the bold statement seem accurate? What is the better way to measure parts on the CMM? We (the engineers) have also doubted our CMM results sometimes. I'm going to wait for the Zeiss CMM support, but I wanted to see what everyone else thought of this.

 

[JNieman]

There's not really enough information to say. I am familiar with Zeiss Contura CMMs and Calypso software, though.

When you say "If CMM measurements are based off the fixture" do you mean you are establishing datum features by having physically probed the fixture surfaces that simulate datums of the part? That's likely a fine way to run things. Depends on the way the geometry is defined and constrained in the drawing/model, somewhat. If your part is not loaded in the fixture correctly, then yes, erroneous results could occur, and would be included in the inspection results.

I've seen some use fixturing only to ascertain a spatial alignment, re-running the part alignment for every part. This means it doesn't matter if your part is slightly misaligned in the fixture, so long as it's close enough to be within the part alignment probing routine. The programmer has the responsibility to ensure they are establishing a datum characteristic in the CMM software that is truly representative of the design drawing/model datum, however.

The finer details of the differences between those two methods is a BIT over my head, however, so I won't speak further on it.

 

[3DDave]

I'm uncertain -
1) Is there a fixture the part is placed in? (I'll guess yes at this point)
2) Is the fixture used for the bench check?
3) Did/does your local CMM check use the fixture?
4) Is the part inspected the same way using the CMM as the bench check?
5) Was the fixture used by the 3rd party?
6) Was the fixture used in the same way by the 3rd party as the bench check/local CMM?
7) Has anyone repeated the inspection on the same part using the same fixture to see how repeatable the inspection method is?

If the fixture is used to simulate datums then of course placement of the part could affect part features measured relative to them.

To do more would probably require a site visit and comparing detailed steps in the methods.

 

[sammcc]

The way in which the parts are checked in the CMM should be decided by the guy who programs the machining centre. He doesn't necessarily have to program the CMM but he should direct the CMM guy as to how it should be programmed.
I find that if the CNC programmer heads off in one direction and the CMM programmer the other then the potential for problems at inspection stage is high.

These guys simply must work together.

Manually checking will almost always be correct and I would never recommend accepting what the CMM has to say over a manual check, that is, providing the method used to carry out the manual check is sound.

 

[KENAT]

What standards are you working to? For ASME with typical conditions invoked then coming off a fixture that simulates datums may better reflect the intention of the standard.

However, that may be a bit of a gross generalization.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[geesamand]

Yes, if the part is measured based on the alignment from the fixture, then it's only as accurate as the fixture and the part location within the fixture. But you can't automatically blame the fixture and walk away.

I would bet my one dollar that the key characteristics on the drawing are not controlled by tolerances set on a clear and robust datum structure. Otherwise the bench and CMM measurements would agree in the first place or the difference in measuring methods would become obvious.

David

 

[dgallup]

I've seen a lot of BS CMM measurements. It's the operators/programmers fault, not the machine. Measuring diameters from 3 points or determining a datum plane from 3 points can give totally ridiculous and unrepeatable results. Correctly putting a part in a good fixture should correct some errors. It is all highly dependent on the part design. Very hard to generalize without seeing the part and how it's being measured.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[Nescius]

Of course, measuring relative to a fixture is not necessarily poor practice. Consider a surface plate with a height gauge. When you put a part on the surface plate, the surface plate is surely a "fixture" in the sense we're discussing. The height gauge sits on and references the "fixture", not the bottom of the part.