Tools
Characteristic evaluation tools
A characteristic evaluation establishes whether an instrument behaves as its standard requires: irradiate it under defined conditions, record what it indicates, and compare the computed response against an acceptance limit. The arithmetic is not difficult and it is easy to get wrong, which is what these tools are for.
What the tools compute
Response and relative response
The indicated value against the conventional true value of the field, and that response normalised to the reference point — the quantity nearly every acceptance rule is written in.
Coefficient of variation
The sample standard deviation divided by the mean at each test point, against the limit the standard sets for it.
Combined standard uncertainty
The Type A component from the repeated readings, propagated through response and relative response, including the reference point's contribution to every other point.
Outcome per test point
The computed value compared with an acceptance criterion you can edit, reported with the margin — and an overall result that distinguishes a failed test from one not performed.
The uncertainty is reported, not applied to the outcome. Applying it requires a decision rule with a guard band, and that is a policy of the organisation rather than a property of the measurement. A tool that silently applied one would produce verdicts that look authoritative and do not transfer between laboratories. See the uncertainty guide and the disclaimer.
Available now
- IEC 60846-1:2009 Tool available
Portable survey meters, dose rate meters and area monitors
Describe the instrument, enter reference values and indications per test point, and read the outcome. Sessions, presets and acceptance criteria stay in your browser; the Excel export is generated locally. Nothing is uploaded and there is no sign-up.
Evaluations it performs — each links to its procedure
How to use them
What to prepare before an evaluation, what the tool does with the readings, and how to read what it reports back.
Planned
Each instrument class needs its own tool: the characteristics, the test points and the acceptance limits all differ. The calculation engine is standard-independent, so what each one below is waiting on is its acceptance criteria — which are in the standard, a copyrighted work that is not reproduced here — and, for several, a measurement condition the engine does not model yet. Those conditions are listed so the scale of each is visible rather than implied.
Electronic personal dosimeters worn on the body
Still needs — Irradiation on a phantom rather than free in air, and angular response about two axes. The personal dose equivalent is defined in the body, so the measurement condition itself differs.
Environmental dose rate monitoring stations and mobile survey equipment
Still needs — The instrument's own contribution to the indication, which is significant at environmental dose rates, and integration time as a declared condition of every statistical result.
High-range portable instruments for emergency response
Still needs — Same quantities and largely the same characteristics as the implemented standard, with a different range and different acceptance limits. Structurally the closest addition — declared values rather than new computation.
Passive dosimetry systems — thermoluminescence, optically stimulated luminescence and film
Still needs — Elapsed time between irradiation and readout as a first-class condition, and the detector batch rather than the single instrument as the population under test.
Neutron survey meters
Still needs — Spectrum uncertainty as a declared component. It does not come from the readings and no amount of repetition reduces it.
Fixed area monitors and warning assemblies in facilities
Still needs — The radiation characteristics are computable, but the requirements that distinguish installed equipment — qualification, safety classification, failure behaviour — are demonstrated by test and analysis of the installation, not by arithmetic on indications.
Surface contamination meters and monitors
Still needs — A different measurement model: activity per unit area rather than dose equivalent, a two-part efficiency that is partly a property of the source, and a detection limit derived from counting statistics and counting time.
Order of implementation
Two things decide it. The first is what the calculation engine already supports: response, relative response, coefficient of variation and uncertainty propagation are the same arithmetic whichever instrument is under test, so a standard that needs only its own limits and test points arrives sooner than one that needs a new measurement condition.
The second is which instruments people actually need to evaluate. If the class you test is not covered yet, say so — that is the part we cannot work out on our own.
RadiMeter is a calculation aid. It does not replace accredited calibration or type testing, and no tool here removes the need to hold the standard it implements.