Machine Accuracy???

[Hilton Roberts]

If your certificate of calibration states that your volumetric accuracy is .0006 inch, do any of you assume that if you only use for example one half the machine volume for your work that your machine is accurate to say .0003 inch?

I am interested in learning if you think your machine is more or less accurate when you use a smaller part of the machine volume than it is when you use the entire volume.

In other words if the manufacturer states one thing for the size of your machine and you only use a proportionally smaller volume for your work, do you think your machine is more or less accurate than the statement on the certificate?

Please let me know what you think and the reasoning behind your statement both for and against.

Thanks!

Hilton

[george frick]

i go with the volumetric accuracy stated on the cert. anything less is a guess.
since our discussion last week i am using our glastonbury precision artifact to check our machines. i check 3 bores 1.0000 dia. in the x/y,z/x,z/y & @ 45 degrees. this gives me a volumetric check in an 8.0000 cube.

[Matthew D. Hoedeman]

You have to assume that if the calibrated spec is 0.0006" that you could have that amount of error anywhere in the envelope

[tking]

Absolutely right, Matt. I've seen it many times, 1/2 spec till the last ballbar position but still in spec, job done!

TK

[MJBolster]

Hilton-
Yes, I did assume that the total volumeric accuracy would be the worse case
if the entire volume were used. Does this mean that my level of confidence
should be no greater in a .1" cube as an 18" cube? I'm glad to learn this stuff as a newbie.

[Hilton Roberts]

I think you cannot assume your volumetric accuracy is linear. I think your total volumetric error can be realized between any two points. Those points may be far apart or they may be close together. I do not believe a CMM is more accurate when you use a smaller portion of the volume to make the measurement. There are just too many things going on at the same time.

Hilton

[johnboy]

Based on the description of the calibration of my CMM by the B&S Tech who calibrated it (the bulk of the error is in the -x+y+z corner) I 'ASSUME' that the error is less in the smaller area I typically work in.

However, for a capability 'STATEMENT' I can only say that the accuracy is only as good as the volumetric accuracy of the entire work envelope, at best.

What I 'believe' and what I can 'guarantee' are two different things.

[John Kingston]

I thought about this for a while before answering. I have always felt it was the same throughout the machine. If I am working in a small area, my "error" is what is determined during the calibration.
You made me think Hilton, I actually questioned my own understanding of volumetric accuracy. Thanks for tickling the brain cells.
John

PS. I'm sure what Johnboy "assumes" can be true. But, we'd need to look at the error map to see where the error is greatest.

[george frick]

looks like everyone is on the same page here, Hilton.

[Hilton Roberts]

I do not know how this would fly with an auditor but if you had a part sufficiently well characterized and small enough to measure in several spots in the machine volume, you might be able to justify higher accuracy by applying the uncertainty of your test part in the area where you were making the measurements. Sure seems like it would take a lot of time and effort to get enough data in enough area to make it pay off.......

H

[Schrocknroll]

I might try that for inhouse use only, but I'm sure that the 1st question an auditor will have is how do you know that everyone will only use that area.

It would be nice to know how your machine is in the "favorite" spot.

[Jeffrey Lovely]

tell him thats the "G" spot !!!!

[Paul Sarrach]

What does the "G" stand for anyway?

[Hilton Roberts]

It is kinda like "that feels good........GEE that REALLY feels good!

[John Kingston]

What does the "G" stand for anyway?

I can't believe I know this...... Grafenburg spot. Not sure of the spelling, but it was the Dr./Researcher who "discovered" it. I want his job!

I like Hilton's answer better though!