Skip to content
Radiation Lab
Sign in
ISO 4037 reference fields: what Field Pro computes

ISO 4037 reference fields: what Field Pro computes

Dosimeters and survey meters are calibrated in reference radiation fields whose spectrum is known well enough that the conversion from air kerma to the quantity the instrument displays can be looked up rather than measured. The series of standards ISO 4037 defines those fields for photons: filtered X-ray beams for the low and middle energies, and the gamma rays of Cs-137 and Co-60 above them. Field Pro is the workbench on this platform for that setting. It models the beam a calibration laboratory actually produces, compares it with the quality the standard describes, and carries the result through to operational dose and its uncertainty. This note explains what each part computes, how accurate it is, and how its screen is laid out.

Why reference fields need a model

An ISO radiation quality, such as N-60 or H-100, is specified by its tube voltage, its added filtration and the half-value layers the beam should have. The conversion coefficients that turn air kerma into personal dose equivalent Hp(10) or ambient dose equivalent H*(10) depend on the spectrum, and the standard tabulates them for the nominal beam. A real tube is never quite the nominal one. Its inherent filtration differs, its anode has aged, the air path to the reference point is not exactly the one assumed, and laboratories build qualities that the standard does not list. In each case the question is the same: how far is this beam from the one the coefficients were computed for, and does the difference matter at the accuracy the calibration claims? A model gives a first answer in seconds and shows which input moves the result most.

The X-ray spectrum

Field Pro computes the spectrum of a tungsten anode at a 20° angle for tube voltages from 10 to 400 kV, then passes it through a stack of filters. The filter materials are beryllium, aluminium, copper, tin, tantalum, tungsten and lead, entered as layers in any order and thickness, together with the air path from the focal spot to the point of test. Inherent filtration (the tube window and housing) and added filtration are separate entries, because one is fixed by the tube and the other is what the laboratory changes to build a quality.

From the filtered spectrum it reports the mean energy, the effective energy (the energy of a single line with the same first half-value layer), the first and second half-value layers in aluminium, copper or lead, the homogeneity coefficient h = HVL1/HVL2, and the relative air kerma. The homogeneity coefficient summarises how hard the beam is: a beam close to one line has h near 1, and an unfiltered beam much less, because the first layer strips its soft end. The raw spectrum without any of this filtration is what the X-ray tube spectrum tool on this site draws; Field Pro starts where that tool stops.

Matching a beam to an ISO quality

ISO 4037-1:2019 lists 50 qualities in four series. The narrow-spectrum series N, used for most dosimeter calibrations, and the low air-kerma series L have heavy filtration and high homogeneity; the wide series W gives more air kerma at the cost of a broader spectrum; and the high air-kerma series H is lightly filtered. Field Pro compares the computed beam with every quality and names the one whose first half-value layer, worked out in that quality's own material, is closest. If you choose a quality as the target instead, it sets the beam beside it characteristic by characteristic (HVL1, HVL2, homogeneity coefficient and mean energy) and flags each against an acceptance band you set, which is the usual way to check that a new filter set reproduces the intended beam.

The comparison is only as good as the inputs. The inherent filtration of a tube is often known only roughly, and at low voltages it moves the half-value layer noticeably. Measuring HVL1 once on the real beam and adjusting the inherent filtration until the model agrees pins it down.

Gamma fields from Cs-137 and Co-60

For the nuclide fields, Field Pro takes the source activity on its reference date, corrects it for decay to the day of the measurement, and gives the air kerma rate at the chosen distance treating the source as a point. The same calculation for any nuclide, with shielding, is in the gamma shielding tool, and the decay correction alone is in the decay tool. What Field Pro adds is the step from air kerma to the operational quantities with coefficients averaged over the source spectrum.

Operational dose conversion

Calibrations are stated in air kerma and then converted to the quantity the instrument is meant to show: Hp(10) for personal dosemeters worn on the body, H*(10) for area survey meters, and the directional dose equivalent H′(0.07) for skin and extremity dose. Field Pro averages the monoenergetic conversion coefficients over the computed spectrum, weighted by air kerma, and does so for an angle of incidence α from 0° to 75°. The angle matters for personal dosemeters, which are tested at several angles because their response falls away as the beam moves off the normal.

Correcting the reading

An open ionisation chamber responds to the mass of air in its volume, so its reading depends on the air density at the time of measurement. Field Pro computes the correction factor kρ from temperature, pressure and humidity, following the reading corrections in ISO 4037-4.

Uncertainty

The workbench builds an uncertainty budget in the form of the Guide to the Expression of Uncertainty in Measurement (GUM). The Type A component comes from repeated readings, corrected first, as u = s/√n; Type B components are entered by name as relative standard uncertainties. They are combined as a root sum of squares, the effective degrees of freedom νeff come from the Welch–Satterthwaite formula, and the expanded uncertainty is given with a coverage factor from the t-distribution for about 95% coverage, in a form that can go into a calibration certificate. Rerunning the beam model with the filtration at the ends of its plausible range shows how far the conversion coefficient can move, a sound basis for that Type B term.

How close it comes to the standard

The accuracy of the engine is checked against the standard by an automated test, and a published accuracy specification is generated from the engine itself, not typed in. For each of the 50 X-ray qualities, the reference beam (inherent and added filtration plus 2.5 m of air) is modelled and its first half-value layer compared with the value in ISO 4037-1:

For the gamma fields, the spectrum-averaged conversion coefficients for Hp(10), H*(10) and H′(0.07) at normal incidence agree with the values recommended in ISO 4037-3 to within 1.03% for both Cs-137 and Co-60, against an envelope of 2%. The specification lists the deviation for every quality but not the tabulated values themselves, which belong to the standard.

These figures say how well the model reproduces the nominal fields. They do not replace the measurements a laboratory makes to characterise its own beams, and a calibration should rest on those.

Finding your way around the workbench

The screen is in English and uses the quantity names, symbols and units of ISO 4037 and ICRU, so its labels read the same as the standard. The inputs sit in four numbered panels in the left column and the results in lettered panels on the right. On a phone, the Input and Results tabs switch between the two, and the bar fixed at the bottom of the screen runs the calculation.

Every calculation runs in the browser. Nothing you enter is sent to a server, and a saved session is a file on your own computer. The source code is public under the MIT licence.

Where to go next