Jim,
Louis is correct.....MIL-STD 45662A is long since dead and buried.....
The internationally accepted definition of traceability, which is a central key to calibration, includes an unbroken chain of comparisons to National or international standards, and stated uncertainty at each step.
The obvious question: Why is this important in your situation?
You mention that the "golden standards" will be outsourced. I strongly recommend an ISO/IEC 17025 accredited calibration laboratory for those items. In the U.S., there are currently four accrediting bodies whose laboratories are recognized and therefore accepted beyond the accrediting body itself. Thos ABs are: IAS, NVLAP, A2LA, and L-A-B.
For 9K, you will not need measurement uncertainty (for internal cal), but if your auditor (once you go for 9K) knows what s/he is doing, then MU (measurement uncertainty) or more precisely the lack of it will certainly raise questions, if not rate a corrective action.
What is the answer? For the items you send out, use only an accredited (to ISO/IEC 17025) laboratory, including having the logo and MU. For internal cal, base your procedures on the USN (NA 17-20 series) or USAF (T.O. 33K series) procedures and specifically reference them. That will avoid validation issues, as the military procedures are validated. Have a procedure (start with very basic one) for MU. Download software such as Uncertainty Calculator 3.2, available at no cost from the Agilent website, and is considered validated.....one of the simpler of these types of packages. Then work on developing your uncertainty.....but not just for MU.....use it as a process-mapping and process improvement tool.
Obviously, I am answering as a Metrologist and accreditation assessor. I realize that your answer may be much easier than I suggest.
Hope this helps.
Hershal
