Reliability improvement for anisotropic biased compensated α/β contamination meter
Résumé
Nuclear instruments such as alpha/beta contamination meter are frequently used in a compensated mode where the contribution of gamma radiation background is compensated by a guard detector. The signal of interest is then the subtraction of counting from both channels. In practice, the noise signal measured by the guard detector is not strictly equal to the noise contribution into the first detector due to anisotropic biases.
The random error (under Poisson assumption) is taken into account to build a hypothesis test. The system is also designed to minimize the systematic error but in some cases, this bias could not be completely removed. The measurement system then shows different behavior when the surrounding environment changes exhibiting inopportune false alarms.
A method allowing the false alarms to be suppressed is addressed in this study for compensated measurement. An improvement in terms of reliability has been proven.