Dimensional Measurements from Neutron Radiographs

Neutron radiography (NR) can solve many radiographic problems, which cannot be solved neither by X- or gamma-radiography. NR is mainly used for the control of explosives and nuclear fuel elements. In the latter field it is often necessary to determine the dimensions of the object under control from neutron
radiographs. There are three main techniques of NR: the direct and transfer technique (using silver halide films and metal converters) and track-etch technique (using nutrocellulose film with coated converters). To check the accuracy of dimensional measurements obtainable with all tehcniques used in NR, a test program was performed by the Euratom Neutron Radiography Working Group (NRWG). For that purpose a calibration fuel pin (CFP), designed and produced at Rise, was used. It contains UO₂ pellets (made of natural and enriched U) enclosed in a zircaloy cladding tube. The UO₂ pellets have different length and are reduced in diameter on half of it. Pellet-to-pellet gaps are formed by Al spacers and pellet-to-cladding gaps by the reduced diameter of half of the
pellet. All the CFP dimensions were calibrated during the assembly of the pin. The CFP was thereafter neutron radiographed at 11 NR facilities in 7 NR centers of the 6 countries participating in the NRWG test program. 30 different film/converter combinations were used. From each radiograph 61 different dimensions were measured (25 in the axial and 36 in the radial direction) using a
profile projector (projection microscope) and a travelling microdensitometer. Altogether about 25,000 dimensional measurements were made. The accuracy of those measurements was assessed by comparing the measured with the true (calibrated) dimensions and calculating standard deviations. This was done not only for all of the measurements but also separately for different recording techniques, NR facilities and different kinds of dimensions of the CFP. Conclusions were drawn about the accuracies of dimensional measurements which can be expected when using a particular recording technique, NR facilitiy and whrvn measuring a particular kind of dimensions.