Speaker
Description
With the growing interest in Generation-IV reactor designs, such as the molten salt fast reactor, new materials are envisioned for use in reactor components, including fuel, structural materials, and coolants. Accurate knowledge of the nuclear properties of these materials is crucial not only for assessing the safety of these designs but also because many of these reactors will operate with neutron energy spectra that differ significantly from those of conventional light water reactors. For example, in chloride-based molten salt fast reactors, various fuel compositions are being considered, with varying uranium enrichment and even potential enrichment of chlorine-37. Once fabricated, these fuel compositions require testing in experimental facilities to validate the predictions of nuclear codes and to ensure the safe operation of the reactor design. To address this need, the NAUTILUS project was initiated at the Chair of Hydrogen and Nuclear Technology using the AKR-2 training and research reactor at TU Dresden. This project aims to develop experimental and simulation expertise to establish an experimental platform for validating nuclear codes and measuring nuclear data. Among the project's key objectives is the development of the pile oscillation method for measuring integral cross section data. This contribution presents advancements in developing this method at the AKR-2 training and research reactor. The experimental design of the pile oscillation device, driven by a linear motor axis, and the detector setup, comprising four pairs of He-3 proportional counters, are described. The drive mechanism, being fully customizable, is restricted in this contribution to a trapezoidal profile, oscillating the sample between the core center and a position 20 cm outside the core, which is well-suited for measuring absorption cross-sections. Signals from the detectors are acquired using a custom ZYNQ FPGA-based data acquisition system at AKR-2. This system registers individual neutron detection events, enabling statistical analysis and user-defined dwell times during post-processing. The signals are then transformed to the frequency domain and processed to obtain the cross-power spectral density between different detectors to reduce biases. To validate the feasibility of this method at AKR-2, experimental data obtained through pile oscillation measurements on foils of natural materials were compared with simulation data generated using the Monte Carlo transport code Serpent 2.2.1. The predicted self-shielding factors for an increasing number of stacked foils of natural indium and iridium show good agreement between simulation and experiment. Furthermore, results for samples of natural gold, indium, iridium, copper, zirconium, and carbon are presented. As expected, the method demonstrates good performance with strong absorbers, yielding results that agree well with simulations, with C/E-1 being below 6 %. For weakly absorbing materials, the method provides generally good results, with C/E-1 ranging between 40 % for copper and 1000 for carbon, where no good agreement was expected, due to the low absorption cross section. These results demonstrate that the developed experimental design, data acquisition, and processing methods for the pile oscillation experiment can be successfully employed at AKR-2. This provides a valuable tool for obtaining integral cross section data and validating numerical models of the AKR-2 reactor.
