Variations in Results of three Diff make CMM for same job -- Explainations requested

[Udayan Patel]

Hii

I have registered as a member to-day only with the hope to get enriched hrough interactions with like minded inquisitive & experianced people across the Globe.

Recently I Checked a Granite Referance Cube 500 X 500 X 500 mm ( Hollowed Constructions ) having Accuracy of Flatness , Perpendicularity & parallelism of all the six faces with in 7 microns with First Principle method by keeping on a Grtade OO Granite Surface Plate, Granite Master Column having 2 microns squareness over 600 mm and inspection grade gauge blocks and Digital probe 0.001 mm resolutions. Flatness was checked wth Digital Electronic level using Grid method, Squareness w.r.to Granite referance Square & Gauge Blocks and Parallelism with Digital Probe on Stand. Accuracy parameters for Flatness, Squareness & Parallelism were found to be within 4 microns. Temp in the Lab was maintained with in +/- 1 Degree C and all the refereance artifacts were calibrated as per ISO/IEC 17025-2005 requirements

When this Granite Cube was checked on a Mitutoyo CMM, parameters were found to be within 12 microns. When this Cube was tested again on two Diff CMM-- Zeiss & other Mitutoyo at other locations results were varying between 8 to 16 microns.

All these CMM are claimed to be having Calibrated with Step Gauge ,Temp in the CMM rooms were maintained with in +/- 2 Degree C

I am preplexed with these glaring diffrennces in the results of three Diff CMMs

I understand these types of Granite Cubes are used to set CMM for mechanical alignments of 3 Axis of CMM for orthogonal movement and CMM uses algorithams of Best Fit while reporting the results

My question is why these glaring variations? Can a CMM be used in abosolute Mode ?

Which method is more reliable for type of inspection carried out for Granite Ref Cube ? I believe First Principle method is more reliable, since the results are repeating within 1 micron variations when re inspected three times.

Please pardon my ignorance on the subject

Udayan

[RICHARD ZASTROW]

Welcome Udayan, To be brief, what degree of accuracy are these CMM's supposed to be producing? The artifact may be near perfect but if the measuring device is not, the results will vary by the variability built into the measuring device. Or by the compounding of variations of the three CMM's.

[Bluesman]

Originally Posted by Udayan Patel Hii I have registered as a member to-day only with the hope to get enriched hrough interactions with like minded inquisitive & experianced people across the Globe. Recently I Checked a Granite Referance Cube 500 X 500 X 500 mm ( Hollowed Constructions ) having Accuracy of Flatness , Perpendicularity & parallelism of all the six faces with in 7 microns with First Principle method by keeping on a Grtade OO Granite Surface Plate, Granite Master Column having 2 microns squareness over 600 mm and inspection grade gauge blocks and Digital probe 0.001 mm resolutions. Flatness was checked wth Digital Electronic level using Grid method, Squareness w.r.to Granite referance Square & Gauge Blocks and Parallelism with Digital Probe on Stand. Accuracy parameters for Flatness, Squareness & Parallelism were found to be within 4 microns. Temp in the Lab was maintained with in +/- 1 Degree C and all the refereance artifacts were calibrated as per ISO/IEC 17025-2005 requirements When this Granite Cube was checked on a Mitutoyo CMM, parameters were found to be within 12 microns. When this Cube was tested again on two Diff CMM-- Zeiss & other Mitutoyo at other locations results were varying between 8 to 16 microns. All these CMM are claimed to be having Calibrated with Step Gauge ,Temp in the CMM rooms were maintained with in +/- 2 Degree C I am preplexed with these glaring diffrennces in the results of three Diff CMMs I understand these types of Granite Cubes are used to set CMM for mechanical alignments of 3 Axis of CMM for orthogonal movement and CMM uses algorithams of Best Fit while reporting the results My question is why these glaring variations? Can a CMM be used in abosolute Mode ? Which method is more reliable for type of inspection carried out for Granite Ref Cube ? I believe First Principle method is more reliable, since the results are repeating within 1 micron variations when re inspected three times. Please pardon my ignorance on the subject Udayan

A CMM is not the most accurate method of mesuring just the quickest, In my exspierience no matter what the salesmen tells us they are only as acurate as the programer and the person loading the part. Check the number of calc points you are taking and make sure they are the same on all machines. The more points the more acurate. Also make sure that they are all asking for the same info in the program. I have seen some pretty screwed up stuff from difrent guys that thought they knew how to program.
Also what I have seen quality do is to run 2 seperate programs 1 for the R&R and one that we use for production. The difrence? The R&R program has 3 times as many hits (Calc Points) as the production program. They can not use the R&R program for production because it takes an hour to run.
Check them and make sure they are all the same.

Bluesman

[NC Cams]

The prior flatness "certifications" were done "how"?

If all three CMM's used EXACTLY the same algorithm (sic) for determining flatness, you SHOULD see exactly the same flatness, PROVIDING that ALL the error and electronic factors on ALL the machines were abolutely identical.

That simply is NOT going to happen.

When you start to measure in microns (0.001 mm = 1 micron = 0.000039") you better have EVERYTHING thermally and mechanically stable, no PERFECT, if you want/expect to measure identical things identically.

I bet that if you measure the same block, day after day, with the same machinee, you won't see the flatness measurements repeat. When/if that doesn't happen, it will be tremendous fun to watch the factory rep explain why not....

[dertsap]

are the cmm's being used in a climate controlled room ?
temperature will greatly affect the accuracy of the system , also if the machines are close to the qc room , the vibrations created from the machines will greatly affect the cmm readings as well
we have one machine in particular that is normally hogging and when its into heavy cuts it not only affects the cmm's but the cnc's around it when we are trying to hold tight tolerances