Radiographic magnification

Any object held away from the film casts an enlarged shadow. The magnification is the ratio of the two distances from the source: M = (SOD + OFD) / SOD = SFD / SOD. In conventional contact radiography the film is pressed against the part, OFD is small, and M is a nuisance term - it is why an indication measured on the film is slightly larger than the real thing, and why the source-side weld cap images bigger than the film-side one on a double-wall shot.

Deliberate magnification is a different technique. With a microfocus source, whose focal spot is a few tens of micrometres or less, the object is moved well away from the detector so the image is projected several times life size. The geometric unsharpness Ug = F x OFD / SOD grows in proportion, but the detector's own unsharpness stays fixed in detector coordinates and is therefore divided by the magnification when referred back to the object. Below a certain magnification the detector limits resolution; above it the focal spot does.

The optimum sits where the two contributions balance, at M_opt = 1 + (U_d / F)^1.5, where U_d is the detector unsharpness. With a 3 mm gamma source and a 0.2 mm detector the optimum is essentially 1.0 - press the film against the part, magnification buys nothing. With a 5 micrometre microfocus spot and the same detector, the optimum is over 200 times, which is the basis of microfocus and computed tomography work.

Whatever the magnification, the IQI and the acceptance criteria refer to the object, not the image. Measurements taken off a magnified radiograph must be divided by M before they are compared to a code limit.

M = (SOD + OFD) / SOD = SFD / SOD
Ug = F x OFD / SOD  (measured at the film)
Ug referred to the object = Ug / M
M_opt = 1 + (U_detector / F)^1.5