the standard or degrade its measurement capabilities.
When a visual
inspection indicates that a possible out-of-tolerance condition exists,
intercomparison tests will be performed where there is a capability to make
such intercomparisons.
b. Intercomparison Tests.
(1) Intercomparison tests will be performed in accordance with TB 9-
4931-489-50, which lists the AACL standards calibrated by the USASL that
require intercomparisons.
(2) Intercomparison tests are required before placing the returned
standard into service and as soon as practicable after receipt from the
USASL or supporting calibration activity. Additional intercomparisons will
be performed when any indication of possible problems creates doubt of the
quality of measurements or at any time the accuracy of the measurement is in
doubt.
(3) Where possible and as workload will allow, intercomparisons will
be performed in conjunction with normal calibration functions.
As an
example, the item to be compared can be substituted as an alternate standard
in a normal workload measurement and the intercomparison data extracted from
the measurement results.
c. Control Limits.
(1) Control limits for intercomparisons, will be determined by the
square-root-of-the-sum-of-the-squares method. For example: Intercomparing
2 like items, each having a accuracy of a 3 parts per million (ppm)
Control limits = ( 3.02 + 3.02 ppm) 1/2
= ( 9.00 + 9.00 ppm) 1/2
= (18 ppm) 1/2
= 4.2 ppm
(2) The computed control limits should be rounded to the same number
of significant digits provided in the error statement of the calibration
reports, 4.2 ppm in the above example.
d. Follow-up Actions. When an intercomparison of measurement standards
results in an out-of-control-limit condition, the appropriate action(s)
noted below will be taken.
(1) Determine which standard is defective by intercomparison with a
third standard, by direct measurement of the standards, or by
68
Previous Page
