X-ray tube voltage selection

Tube voltage sets the photon energy spectrum, and photon energy controls both penetration and contrast - in opposite directions. Raising the kV shortens the exposure dramatically, because more photons are produced and each is more penetrating. It also flattens the image, because the attenuation coefficient of steel falls and varies less with thickness as energy rises, so a given thickness change produces a smaller change in transmitted intensity.

Radiographic contrast is therefore best at the lowest voltage that will still penetrate the section in a workable time. That is the rule the standards codify: ISO 17636-1 sets a maximum tube voltage against penetrated thickness, not an optimum. Working at or near that maximum is legal but produces the flattest image the code will accept, and on thin sections it will fail the IQI. Working well below it gives better contrast at the cost of exposure time, until the exposure becomes impractical or scatter from the long shot degrades the image anyway.

For materials other than steel, convert to steel-equivalent thickness first using the radiographic equivalence factor for the intended energy - and note the circularity there, since the factor itself depends on kV. One iteration is normally enough: pick a kV, take the factor for that band, recompute, and check the answer has not moved into the next band.

Above roughly 100 mm of steel the required voltage runs into the practical limit of conventional tubes and gamma or linac sources take over. Iridium-192 covers 20 to 100 mm, cobalt-60 40 to 200 mm, and megavolt linacs beyond that.

t_steel_equivalent = t x equivalence factor
kV_max from ISO 17636-1 Figure 1 (steel)
kV_start = 0.85 x kV_max as a contrast-preserving starting point