X-ray monolithic detectors for industrial CT scanning

Computed Tomography (CT) scanners have become indispensable in various industries due to their ability to perform non-invasive, real-time inspections. These systems utilize X-rays to generate detailed 3D reconstructions of scanned objects, enabling precise measurements and defects detection. Unlike traditional inspection methods that often require disassembly or destructive testing, CT scanners preserve the integrity of products, making them invaluable in fields like manufacturing, food safety, wood processing, and even cultural heritage preservation. The FLECT project (Food and Log Computed Tomography) aims to improve the performance of current industrial CT technology. CT scanner systems and reconstruction algorithms are in constant evolution to improve efficiency, cost effectiveness, and market competitiveness. In particular, improvements in X-ray detectors are of prime importance, as detection efficiency and resolution ultimately define the CT system overall reconstruction performances. Traditional detection systems, which rely on scintillators and photodiodes, are constrained by slow signal acquisition, high-energy requirements, and susceptibility to signal distortion caused by photon pile-up. Furthermore, these systems often face challenges in maintaining cost-effectiveness while delivering high-resolution results, especially in competitive markets. In response to these challenges, the FLECT project, in collaboration with MiCROTEC company and the INFN ARCADIA initiative, seeks to integrate advanced Depleted Monolithic Active Pixel Sensors (D-MAPS) into industrial CT scanners. These sensors, originally developed for High-Energy Physics (HEP) experiments, offer low power consumption, fast readout speed and high hit rates capability that would enable the use of lower intensity X-ray source. Consequently, the X-ray intensity reduction leads to a lower amount of radiation interacting with the products scanned, as well as an increase in the productivity and competitiveness while reducing the waste at the same time. This thesis initially gives an historical review of both the production and the detection mechanisms of X-ray particles, quantitatively modeling the phenomenon occurring when the inner structure of the substance is modified by the deposited energy. Then the sensors currently utilized in HEP are discussed, focusing on the chip designed by the INFN ARCADIA collaboration, which aims to develop a D-MAPS tailored towards large-area and low-power tracking applications. The GEANT4 simulation toolkit has been used to determine the feasibility of adapting such D-MAPS technology to industrial X-ray CT scanning applications. Moreover, scans of objects with simple geometries are simulated as in a tomography apparatus, and the image reconstruction is provided, showing the capability of distinguish materials of slightly different absorption coefficients. The risk for industrial employees in working with ionizing radiations are then considered, and geometries and materials that could be used as external shield of the apparatus to reduce the external dose rate to a safe level are explored. Finally, a prototype of the ARCADIA sensor is tested in two different environments: the Padova University laboratory with its X-ray tube source and the MiCROTEC company building with industrial CT scanner. The successful testing of D-MAPS sensors into existing MiCROTEC CT systems demonstrates their compatibility and potential for future industrial applications. By adapting this technology to industrial applications, the project envisions a new generation of CT scanners that are faster, safer, and more cost-effective.