Study Ray Detector Deuterated Liquid Scintillation Detector Based Digital γ

With the fast development of nuclear engineering, more precise nuclear data are required. Integral experiment is an important element of the work of nuclear data. To check the related nuclear data, it is of great significance to detect the γ-rays in the mixed neutron/γ-ray radiation fields in the neutron integral experiments. To obtain excellent pulse shape discrimination performance, accurate y-ray spectrum and original data subsequent analysis property in mixed neutron/γ-ray radiation fields, a digital γ-ray spectroscopy system based on a deuterated liquid scintillator and an8-bit resolution and2GSamples/s sampling frequency high-speed digitizer was set up.The neutron/γ-ray discrimination properties of three algorithms based on the digital spectroscopy system were investigated in a mixed neutron/γ-ray Am-Be source field. The discrimination algorithms utilized are the digital charge comparison (CC), simplified digital charge collection (SDCC) and pulse gradient analysis (PGA). The effects of both digital filters and simulated pile-up events on the performance of these techniques were investigated. The results show that all these discrimination algorithms are potentially suitable for developing a digital γ-ray spectroscopy system in mixed neutron/γ-ray radiation fields. In addition, the SDCC and CC methods are preferred choices when applied in low count-rates fields, whereas PGA method with a seven-point average running filter is favorable when pile-up signals constitute a large fraction of the detected events.The BC537presents two basic advantages with respect to the BC501A, namely lower neutron detection efficiency and comparatively better pulse shape discrimination in the MeV energy region. Moreover, the BC537has fewer parasitic and contamination γ-rays than BC501A when used to measure γ-ray spectra in mixed neutron/γ-ray radiation fields, which confirm that the BC537is to be preferred to BC501A for γ-ray detection purposes.To determine the incident γ-ray spectra from the measured BC537recoil electron spectrum, a set of detector response functions at appropriate energy intervals are needed. The Geant4Monte Carlo simulation program has been used to generate γ-ray response functions, which are in good agreement with the results of the EGSnrc codes. The response functions of the simulation are in good agreement with the experimental measurement results, which validate the rationality and reliability of simulation parameters. The effects of the aluminum container and the beam radius on the detection efficiency were investigated as a function of the beam energy by using Geant4simulation.With the recoil electron spectrum detected by the digital γ-ray spectroscopy system and the response functions calculated by the Geant4codes, a neural network unfolding technique has been used to unfold the y-ray response of the BC537detector, and the results are compared with those obtained using an iterative unfolding algorithm. The iterative solution, a time consuming method, has the potential advantages of stable and non-negative. On the other hand, the neural network undertakes the task of determining the weight matrix during training and performs only very simple calculations during unfolding, which make it much more adequate for a real-time spectrometry system.