Study on Neutron Bragg edge Imaging Technology Using Compact Accelerator-Driven Neutron Sources (CANS)

Abstract

Bragg edge imaging has become one of the powerful tools in many research areas such as material characterization. This technique utilizes a neutron time-of-flight (TOF) method and a two-dimensional detector. Using Bragg edge imaging, the crystallographic information of structural materials is obtained for bulky samples. Another advantage of this method is that no destructive processes are needed during sample preparation. To perform Bragg edge imaging, a source of neutron with the flux of ~ 10^4 n/cm^2/s at 5 meV (λ ∼ 0.4 nm) and a wavelength resolution (∆λ/λ) of ~ 1% are required. The reason for using 5 meV (0.4 nm) neutrons is that the main Bragg edge of structural materials is appeared at this neutron energy (wavelength).

There are several methods of neutron production such as nuclear reactors, spallation sources, and small scale accelerators using charge particles e.g. electron or protons. Due to limitation to access the large scale neutron facilities, compact accelerator-driven neutron sources (CANSs) are gaining a lot of interest globally for their structure simplicity and construction cost compared to large accelerator facilities or a nuclear reactor.

There is a lack of neutron source for performing Bragg edge imaging in Korea. This thesis aims to provide fundamental steps towards developing a neutron source based on an electron linac suitable for the Bragg edge imaging.

A target-moderator-reflector (TMR) was optimized by the Monte Carlo calculations using the PHITS-3.1 code. The aim of the optimized TMR was to achieve the neutron flux of 10^4 n/cm^2/s and ∆λ/λ of 1% for neutron energy of 5 meV (λ = 0.4 nm) and at a distance of 1000 cm. The target material, target dimensions, electron beam energy, moderator and reflector were optimized. Two important parameters were considered during TMR optimization, neutron flux and neutron wavelength resolution (∆λ/λ). Different target-moderator geometries were considered and a wing geometry was selected as the optimum configuration. Polyethylene (PE) and graphite were selected as the moderator and reflector, respectively, and their dimensions were optimized. The PE moderator was considered not only at 296 K, but also at 77 K in the PHITS calculations using proper thermal scattering law (TSL) data to increase the cold neutron flux. The TMR including a natW target and a PE moderator at 77 K surrounded by a graphite reflector yielded a neutron flux of 1.16 × 10^4 n/cm^2/s and ∆λ/λ of 1.05% at a distance of 1000 cm for 5 meV neutrons. The calculations were also performed using the FLUKA Monte Carlo code and the results were consistent with the PHITS results. The Bragg edges of an α-Fe sample were reproduced using the PHITS simulations by the wavelength resolution value obtained from the optimized TMR, concluding that the designed TMR is suitable for the Bragg edge imaging.

A neutron facility was constructed to experimentally verify the photoneutrons production using a small scale electron linac. The Electron Linac for Basic Science (e-LABs) at Pohang Accelerator Laboratory was utilized for this purpose. Due to the geometrical constraint in the e-LABs, it was not feasible to examine the whole optimized TMR. A natW target irradiated with the electron beam and the photoneutron production yields from 1.4 to 11 MeV were successfully measured by the TOF method. The measured data were also compared with the PHITS-3.1 and FLUKA 4-3.0 calculations. The results showed that PHITS (default mode), FLUKA and measured data were in a better agreement for neutron energies below 4 MeV, while PHITS underestimated the measured data above 4 MeV more than the FLUKA calculated results. The JENDL-5 and TENDL-2019 photonuclear libraries were also used in the PHITS code and the calculated photoneutron production yields improved above 4 MeV. However, they still underestimated the measured data.

For enhancing cold neutron flux, a PE moderator block of 20 × 20 × 5 cm^3 was placed in a Styrofoam next to the target and it was cooled down to 87.3 K using liquid nitrogen. The integrated cold neutron flux over 4 to 16 meV was measured and compared with that of the PE moderator at 296 K. The results showed that the cold neutron flux over 4 to 16 meV increased by a factor of 1.2 by cooling down the PE moderator to 87.3 K. The experiment setup was simulated in the PHITS code and the cold neutron flux was calculated over 4 to 16 meV. The PHITS calculations showed that a thick layer of liquid nitrogen absorbs cold neutrons. Thus, the PE moderator needs to be cooled down in a thin layer of liquid nitrogen. In this way the absorption of cold neutrons by nitrogen is effectively mitigated and the cold neutron flux is significantly enhanced.

The capability of the Bragg edge imaging technique in material characterization was experimentally validated. As the electron beam current of e-LABs was very low, the cold neutron flux was not sufficient to perform the Bragg edge imaging experiment. The experiment was performed at Hokkaido University Neutron Source (HUNS) in Japan. This experiment was performed at HUNS because our designed TMR parameters (neutron flux and wavelength resolution) were quite similar to HUNS parameters. It was also important to show the capability of a CANS in Bragg edge imaging in measuring microstructures and texture of the industrially produced steel samples. The microstructures of two sets of austenite stainless steel type 304 were measured. The samples included cold-rolled, hot-rolled, heat treated and solidified cast slab samples. The crystallite size, phase volume fraction and texture of the samples were successfully measured by the Bragg edge imaging and the Rietveld Imaging of Transmission Spectra (RITS) code. The results of the Bragg edge imaging were confirmed and validated by the Electron Backscatter Diffraction (EBSD) which is a conventional method in material characterization.

It is concluded that the results presented in this thesis provide important advancements towards a neutron source development based on an electron linac to be suitable for the Bragg edge imaging. The Bragg edge imaging technology was experimentally validated by measuring the microstructures of type 304 stainless steel samples.