Comparison of standard digital detector arrays and photon counting detectors with high-energy X-ray generators

Digital detector arrays (DDAs) have developed rapidly in some segments of the industry as a replacement for film radiography or in some cases computed radiography, especially in the aeronautical industry and for casting inspections. However, these detectors show limits when applied for inspection of high thicknesses of steel or other absorbing materials, linked to high levels of noise and scattered radiation. New technologies in the form of photon counting detectors (PCDs) are being developed for some specific inspections where usual DDAs are not proving effective enough. These detectors allow the choice of an energy threshold under which the X or gamma photons are not being considered when creating the digital image, allowing the user to basically cut off part of the scattered radiation from the signal received by the detector, hence improving the signal to noise ratio (SNR) and the image quality. Despite the drawback of a very small active area, such systems provide advantages over usual DDAs in specific settings such as: Inspection of high thicknesses; Inspections in environments where there is ambient radiation (nuclear power plants for example); Inspections where different materials need to be differentiated on the radiographic image (dual energy). This paper regards tests that have been performed jointly by CETIM and INTERCONTROLE agents at the request of CETIM and using a PCD developed by INTERCONTROLE especially to perform inspections using high-energy X or gamma sources. These tests were performed using X-ray generators of 450 and 530 kV, and their aim was to compare the performance of PCDs relative to standard DDAs on high thicknesses of steel blocks and samples. Similar exposures were made using the PCD and a DDA for steel thicknesses of 50 to 90 mm (see figure 1), and the comparison are based on measures of image quality index on wire-type image quality indicators (IQIs, see figure 2), signal to noise ratio, spatial resolution (SRb) and contrast to noise ratio (CNR) on the different images.