Calibration of Mylars (?)

Define calibration?

Reason for question: Customer print says CAD data is master data for part size. Mylar check fixture is made from CAD data. Mylar is the nominal of the part perimeter with the upper and lower tolerance drawn. Lay the part on the Mylar and make sure the perimeter is in between the lines.

Told mylars have to be calibrated annually. I have always understood calibration to mean adjustments can be made. No adjustments can be made to a mylar visual cheek sheet. It's replace when needed based on visually checking line quality.

Verification is more what is needed and that is to make sure the mylar lines are still readable. Semantics, but verification is much quicker and less expensive than calibration.

Verification would be visual check of revision and line quality on the mylar.

Calibration is the measurement of each part characteristic on the mylar with another calibrated method (CMM - Calibrated steel rule, etc.).


Input?

 

I would agree that mylar tends to be very durable and I definitely agree that in the case described it would at most need a visual check for visibility and condition. Technically I would consider this a verification and not a calibration. It would be a good time to take advantage of interval extensions if you utilize them, since the likelihood of a failure is very small. I would have no problem extending a mylar overlay out to three or even five years.

As was already mentioned, adjustment is totally separate from calibration. Calibration is the comparison of one value to another by means of a traceable standard. The "known" value of a standard is technically a range of values, due to uncertainty. Calibration produces a numeric value.

Verification is concerned with attributes (i.e. is the color correct, what types of defects are present.) There may be accept/reject criteria but these are not generally based on a numeric value. This also tends to be somewhat subjective.

Jennifer's points are fair. I feel most yearly calibration frequencies are based on convenience rather than need. I would however not lump gage blocks in with steel rules or visuals. For dimensional metrology, calibrated gage blocks are a huge part of everything we do. These are one of our traceable standards.

 

I get calibration is determining the value as compared to the desired value, but is reacting to that value if required not part of calibration?

I understand calibration is the determination of the value of measurement as compared to the intended result. I am going to assume you would say it would then kick into corrective action for the adjustment, and then back to calibration to confirm. I was once told you can't calibrate something that cannot be adjusted? It's just verification if it cannot be adjusted. Is that not accurate? Is it semantics?

 

Outside of the semantics (I was wondering how long it would be before the Calibration vs Verification dialog would begin again)

Who told you?

If it was the customer...it may be easiest/cheapest to simply comply. Think about all of the silly things you do to please customers and just add this one to the pile.
This yearly check costs way less than the dinner your salesperson takes the customer out for.

If it was a non-customer auditor there are three ways to go:

1. Comply (sometimes it's easier/cheaper to go along with unreasonable demands than to fight them)
2. "Sort-of comply"...by establishing a calibration period of 10 years and stating why you think that time frame is adequate, but including also a "replace if damaged"
3. fight the fight

I've rarely found it the most profitable road to "fight the fight"...occasionally so, but rarely.
10 minutes/year of doing something I know is ridiculous is less annoying than 20hrs of argument with unknown outcome.
In the big picture...10minutes a year costs less than an ongoing debate.

None of us have the time to fight for everything we feel is right and good, or against everything we think is wrong or ineffective....more's the pity.

 

FWIW, the few Mylars we use here are used flat (as opposed to being wrapped around a part).

It's pretty quick and simple to throw them on a CMM and verify...thus my 10minutes comments.
I know others can use mylars for much more demanding operations where they may not be nearly as easy to measure...

 

I have only seen mylars used as overlays for an optical comparator, or for basic dimensional checks as Tom described. What Bev describes is a very different situation. That's an interesting usage I wouldn't have thought of.

I totally agree with that statement; we calibrate to collect data to show the stability of our M&TE, regardless of whether it's a mylar with complicated dimensions or any other sort of instrument.

 

I used to work in Quality in an Aerospace sheet metal facility and we used mylars almost exclusively. Most of what was being made used tooling that was decades old. This shop also did not have digital opportunities like a CMM. So contours and such were fun.

 

Thanks for the input everyone! I am sure that if the topic comes up again I will be in a better position to debate this need. Come to find out this may have been a self imposed action several years ago before my time, to just end the debate with the auditor, so the organization said we would do this just to move on.

Not my style, so I will be changing this - if it does not add value, or is not required to be done (regardless of adding value) then we are not going to do it just to avoid a discussion with the auditor.

 

In order to validate and verify a mylar, I always had a 1" by 1" box place on it as a standard part of the design. Then, measuring that box with an appropriate measurement system would assure the stability of the mylar has not been compromised, and the check was fairly painless. It was particularly handy when printing them out to assure they were to scale. It was also a lot easier than measuring features.