Methods for Inspection of Checking Reference Granite Cube

CMMs are not absolute measuring devices. If you look at the specification for your machine it will typically claim a repeatability of something like +/-3um. In addition it will show a measuring accuracy that varies according to the length of the feature being measured. typically this will be 'L'/250um. For you 500mm measurement this will give you +/-2um. When you add in the repeatability error, your best accuracy is +/-5um. This means that even if you measure the same feature twice with the same CMM, over 500mm you could get a difference of 10um between readings & still be within the specified accuracy of the machine.

These figures will vary from machine to machine with the larger frame machines being the least accurate. you shound check the specification for your particular machine to see if the errors that you have found are within its specification.

Dear Guest

Thanks for your prompt and valuable inputs. Agreed measuring accuracy and repeatability of the machines would contribute in measurement errors. But in case of checking perpendicularity of Granite Cube what we are checking is Squareness acuuracy between two planes. accuracy of Machine as indicated by you L / 250 mm or Some Numeral value +/- L / 250 normmally relates to linearity errors or volumetric errors in linera distance travelled. For squareness checking while probing vertical face of a cube probe would travel max distance to the tune of Physical value of the error of squarenes. So distance traveled by probe in one dirtection would be quite negilgible and that would be fraction of 100 microns, So even if errors of machine are of the order of L/250 mm total machine error for such small & negligible distance would not contribute as much as has been reported ( around 15 to 18 microns ) on differant make machines for the same job.

Under the circumstances would I be wrong if I say that using method of First Principle would be the most appropriate and reliable ? I understand that Referance Granite cubes are used for setting mechanical accuracy for orthogonality of the travers of X , Y & Zee axis. Will it be correct to USE CMM for checking Granite Cube which it self is used for checking CMM ? Shall appreciate your comments Thanks -- Udayan

Here's my two cents. Most CMM's I've worked on, and with, are good to about 5 microns repeatability. Not accuracy. Those are two different things all together. I did a study to accept a machine for production. My method of acceptance was not based on calibration, that can change and be changed. I've even adjusted the calibration to get better readings or to just get one CMM to read the same as the other.

OK, what I did was use a laser to check the machine when it was traveling it's full length. end to end in all three axis' What I learned on every machine tested, is that repeatability was always 3 or 4 times better then accuracy. When I was satisfied that the machine could repeat, I did a calibration, re-tested and then adjusted the calibration to get the results I needed.

I'm not an expert, but I do know one. if you wish more information, contact the expert on CMM's R. Callaghan at Independent Quality Labs in Westerly, RI. USA. He has authored the ASME spec. on CMM's

Dear Mr. Labyquy,

Thanks for joining discussion and suggestions regarding Laser Calibration. You must have dome error mapping for all the three axis of travel and must have given correction to get better accuracy of measuremnts.

Can I have Email ID / Website of Mr. R Callaghan of Independent Quality Labs , Westerly , RI ? Please also give contact address and phone no if you have.

I am still waiting for some one experianced in the field to clarify " Can CMM be used to check Granite Referance cube for Flatness and perpendicularity having allowable tolerance of 7.5 micons " ? Or First Principle Method is a better and appropriate option especially when results of three diff CMM's are showing substential variations .

Thanks -- Udayan