Calibration with B89 ball bar verses step gauge calibration

[Towndogg]

It's time for my yearly calibration on the Global 777 CMM. All calibrations up to this time have been done with the B89 ball bar and have been within the acceptable limit of Hexagon certification.
I understand that a B89 calibration is checking for repeatability and not actual length of the bar. The Koba bar or step gauge will check for actual length. What kind of calibration do most of you
have performed on your CMM and how would I know if my machine needed this type of calibration?

[Matthew D. Hoedeman]

ball-bar & step bar. If you only do ball-bar, all you are certifying is that the machine is 'square', you have no idea if the linear is even close.

[Towndogg]

This is so true Matthew. I wonder why a certification like this will pass TS or IATF audits?

[Darrollh]

If all they are using is a ball bar then you need to send them away. You’re spending a ton of money for shortcut work.

[RIDER]

They have 4 basic performance tests and 6 enhanced performance tests. The first basic only checks repeatability (sphere) and volumetric performance (ball bar). In the AS9100 world, this "calibration" would be a major finding. I always opt for the 17025 accredited ISO 10360 cal with before and after results.

[Darrollh]

That’s what dmis100 and the Hexagon guy said that installed our machine.

[Douglas]

Last calibration tech almost insisted he must remove my fixture plate to get that one last step in... I thought he was being a bit of a nut but for the right reasons so I pulled it off and gave him the granite

[JEFMAN]

A ball bar check gives only informations about "perpendicularities" between axes.
Whatever the position, the distance between balls should be the same, even if you don't know exactly the value.
A step gauge gives deviations values from a standard calibration.
You can check the cmm with the same locations / orientations than the ball bar, with more steps than the ISO10360 (or ASME equiv), it will give a large knowledge of the cmm geometry at the moment when yoiu do it.
If the thermal gradient changes, then essentially the pitch of the granite will change, so the cmm accuracy could be decrease a little.
For example, if you measure a KOBA step of 1020 mm, there are 26 steps, so 52 lengthes calibrated.
If you measure them along each axis (3 positions), each plane diagonal on each side of a cube (left, right, at the top and the bottom, in front of the cmm and behind = 12 positions), and then each volumetric diagonal (4 positions), repeat 3 times, you will obtain 2964 values of deviations.
The standard deviation should give a right picture of the cmm accuracy (the average should be very close to zero... !)
Adding some positions along axes could help finding pitch and yaw defects, comparison of diagonals high / low and left / right finding roll defect.
The Y axis defect is harder to detect, it can be done with different lengthes of stylus.

If you ask it to the person who makes the calibration, you will have to pay a lot and to "lock" the cmm during more than a week !!!!!!!!! (but you will "know" your cmm capability !)

[William Johnson]

Also with the B89 they do not make any corrections to the error map as it really only lets you know if it falls within the factory specs. ISO 10360 is the only calibration that I have done. If you have the before and after done it is like paying for 2 calibrations. Unless you have someone requiring a before and after I would not do the before. We only do the after.
The before lets you know the condition of the CMM at the time of calibration but really does not offer any value unless the CMM is way off and then you should have already had issues with inspected parts.
Just my opinion.

[JEFMAN]

Also with the B89 they do not make any corrections to the error map as it really only lets you know if it falls within the factory specs. ISO 10360 is the only calibration that I have done. If you have the before and after done it is like paying for 2 calibrations. Unless you have someone requiring a before and after I would not do the before. We only do the after.
The before lets you know the condition of the CMM at the time of calibration but really does not offer any value unless the CMM is way off and then you should have already had issues with inspected parts.
Just my opinion.

A real quality management should force having the before... And let you try to explain how wrong were your measurements if the cmm is off !!!!!
I like this kind of QM !!!!!!!

I think that ball bar allows perp corrections, taking as "real value" the average of lengthes... But it's a kind of luxury DIY !!!!!

[William Johnson]

I have worked in aerospace 20 yrs and medical 21 and never had any employer or company ask for a before results. I run my own test on the CMM's in between calibrations and (knock on wood) never had it bite me.

[louisd]

Medical devices: we do before/after inspections in accordance with Title 21 CFR 820.72, which states https://www.govinfo.gov/content/pkg/...-sec820-72.pdf

(b) Calibration. Calibration procedures shall include specific directions and limits for accuracy and precision. When accuracy and precision limits are not met, there shall be provisions for remedial action to reestablish the limits and to evaluate whether there was any adverse effect on the device's quality. These activities shall be documented.

Basically, we are obligated to check if the calibration is within spec or not -and document outcome- prior to calibration/adjustments.
In addition to this we've added to internal work instruction that if the B89 ball bar cert (for our TP200 machines) or ISO 10360 (for analog probe CMM) fails, we can justify the failure as having no impact to product as long as the calibration didn't fail by more than 10% (1 standard deviation) of our product line's tightest dimensional constraint. That work instruction has paid off several times now.

[JEFMAN]

Ok a Koba is more expensive than a ball bar, but it gives more and more informations.
You can easily transform the Koba into a ball bar, if you want to perform a quick check, just add this on each end !!! : (and this is not a spam, I sale nothing !!!!!)

bb2.JPG

[Tater]

We use a Laser for dimensionals. Depends on machine travel..........