From model to measurement: Development of a digital twin of the Dingo thermal neutron imaging beamline and real-time wide-band neutron beam-monitoring devices

This Thesis presents the development and validation of a comprehensive Monte Carlo model of the Dingo thermal neutron imaging beamline, Australia’s sole thermal neutron source, located at the Australian Nuclear Science and Technology Organisation (ANSTO). The model, constructed using CAD drawings of the entire beam transport path and shielding structures, is designed to precisely predict the in-beam neutron field at the sample irradiation stage positions.

The research demonstrates methods for importing complex CAD structures into Geant4 and modelling neutron interactions with a sapphire crystal thermal neutron filter using an external package. The model’s performance was validated using multiple techniques, including Bonner sphere spectroscopy (BSS), neutron activation analysis (NAA) and microdosimetry. A methodology to model and validate the neutron spectrum was developed, rather than relying solely on unfolding from spectroscopic results.

Characterisation of different beam configurations available at Dingo are presented, including neutron planar distribution, fluxes, and spectrum. The research reveals that the Dingo neutron beam consists of approximately 63.5% thermal neutrons, 16.3% epithermal neutrons, and 20.2% fast neutrons. The spatial distribution of the beam was mapped with high resolution, showing good agreement between simulated and experimental results within 2.3% for most data points at the centre of the field.

This thesis also presents the computational design and optimisation of novel instrumentation for real-time quality control of high-intensity mixed gamma/neutron fields. Specifically, a quad-MOSFET device and an array of planar diodes were developed for the real-time measurement of individual beam components. Preliminary experimental validation of these devices at Dingo demonstrates their potential for enhancing beam monitoring capabilities.

The developed model provides a platform for for virtual optimisation of experimental parameters, computational design and testing of new instrumentation, shielding materials, neutron filters, and beamline upgrades. It also enables estimation of sample activation and offers insights into phenomena that are challenging to measure experimentally. This work contributes to the establishment of a framework for detector research and development and radiological research in Australia, expanding the capabilities and applications of the Dingo beamline.