I have been a quality consultant for years developing automotive supplier and then ISO/TS systems.

The 3rd party auditing firm was correct. Each of the squares on the shop floor must be in your gauge log and confirmed and I would suggest yearly. That should not be a big cost factor then.

Here is how I would do it.

Have 1 square confirmed by an outside calibration firm and retain the certificate on that square. It will now be the master square in your plant and not used for anything else except for calibration purposes.

The remaining squares will be confirmed using the master. If one sees light through the squares as the kiss back to back, then the shop floor square must be replaced.

Remember, place both the master and the square on the granite table. Push them together so that the vertical squares are touching each other. Look for light.

3rd party auditing firms cannot tell a company how to confirm the calibration so perform it in the easiest, most effient way possible.

Hope this helps.

Dave D.
www.qmsi.ca

What square are you referring to, machinist, inspection, etc?

I think I got a good Idea as to what I need to do.  These are mostly machinist squares that our operators are using to monitor the bends on their parts as they are running.

We had some trouble with machinist squares getting a setup to check the angles. As I recall it was three points required for calibration. One thing that bugged us was the scale used in square. We had sveral where the scale had been changed out and it didn't match the original equipment scale.
We used a cylindrical square checker on a granite table to check all inspection squares, which were just as good as the standard.  

One interesting point is they were going to include the carpenter squares used by the fab shop until they put a couple on the table to check. They gave up real quick as nearly all were Al.

I have seen a dual system utilized where if a tool is not being used for obtaining quality data or to actively control/adjust a process it gets labeled as "reference only, calibration not required".  

Since you indicate that part decisions are based on measurements taken, such a system is likely inappropriate.  I agree with the previous posters that you should use as simple a calibration process as is effective for product control.

Regards,