How Do You Judge the Accuracy of a Leak Detector?
According to Six
Sigma* guidelines, there are four different parameters that determine
the quality of measurement equipment:
- Repeatability
- Reproducibility
- Linearity
- Stability
When applied to
leak detection, what do these four characteristics mean?
Repeatability
Repeatability
indicates the degrees of variation when measuring
the same leak over and over again (by
the same operator). You can determine
this by simply measuring the same calibrated
test leak several times under the same
conditions. Repeatability is determined
by the maximum and minimum detected leak
rates: the smaller the difference between
the two values, the better the repeatability.
Reproducibility
Similar to repeatability, reproducibility considers
the test results from different operators.
To determine the reproducibility of a
leak detector, different operators measure
the same calibrated test leak several
times. Again, the variation of the different
leak rate measurements gives you reproducibility;
the smaller the variations, the better
the reproducibility of the leak detector.
Linearity
Linearity in a leak detector shows the span of
measurement error over a large measurement range. This is especially
important when calibrating a leak detector with a leak rate value different
from the trigger value used. This is determined by calibrating the
leak detector with a small test leak and then testing this against
larger test leaks (factor 100 to 1000), or vice versa.
Stability
Stability
indicates the reliability of test results over time. To determine the
stability of a leak detector, set the leak
detector with a calibrated leak. Now let the leak detector run for
many days, checking the leak rate reading in regular intervals; the
smaller the deviation after time, the better your leak detector.
INFICON
rigorously tests its leak detectors according
to these four characteristics. We will share the results of these tests
with you in our next issue.
* Six Sigma is
a management strategy used to identify and remove the causes of defects
and errors in manufacturing and business processes.
|