Chihiro Ichihara

Preliminary Experiments on Accelerator-Driven Subcritical Reactor with Pulsed Neutron Generator in Kyoto University Critical Assembly

A series of preliminary experiments on an accelerator-driven subcritical reactor (ADSR) with 14 MeV neutrons were conducted at Kyoto University Critical Assembly (KUCA) with the prospect of establishing a new neutron source for research. A critical assembly of a solid-moderated and -reflected core was combined with a Cockcroft-Walton-type accelerator. A neutron shield and a beam duct were installed in the reflector region for directing as large a number as possible of the high-energy 14MeV neutrons generated by deuteron-tritium (D-T) reactions to the fuel region, since the tritium target is located outside the core. And then, neutrons (14MeV) were injected into a subcritical system through a polyethylene reflector. The objectives of this paper are to investigate the neutron design accuracy of the ADSR with 14MeV neutrons and to examine experimentally the neutronic properties of the ADSR with 14MeV neutrons at KUCA. The reaction rate distribution and the neutron spectrum were measured by the foil activation method for investigating the neutronic properties of the ADSR with 14 MeV neutrons. The eigenvalue and fixed-source calculations were executed using a continuous-energy Monte Carlo calculation code MCNP-4C3 with ENDF/B-VI.2 for the subcriticality and the reaction rate distribution, respectively; the unfolding calculation was done using the SAND-II code coupled with JENDL Activation Cross Section File 96 for the neutron spectrum. The values of the calculated subcriticality and the reaction rate distribution were in good agreement with those of the experiments. The results of the experiments and the calculations demonstrated that the installation of the neutron shield and the beam duct was experimentally valid and that the MCNP-4C3 calculations were accurately carried out for analyzing the neutronic properties of the ADSR with 14MeV neutrons at KUCA.

Experimental investigation of variance-to-mean formula for periodic and pulsed neutron source

Abstract A series of experiments was carried out by using the Kyoto University Critical Assembly and a pulsed neutron generator attached. The experiments were intended to investigate a variance-to-mean technique for measuring the prompt neutron decay constant α of the subcritical reactor system that was driven by the periodic and pulsed neutron source. As the results, it was confirmed that the α value measured by the present technique agreed with a reference value evaluated by the pulsed neutron technique. Furthermore, it was demonstrated that the present technique was available to measure the α value even when the product of α and the repetition period of pulsed neutrons was smaller than unity. Since measurement of α by the pulsed neutron technique is generally difficult under such conditions, it was concluded that the present technique was convenient for measuring the α value.

Application of variance-to-mean technique to subcriticality monitoring for accelerator-driven subcritical reactor

To investigate the applicability of the Feynman- (or variance-to-mean) and Rossi-α techniques to subcriticality monitoring of the future accelerator-driven subcritical reactor (ADSR), a series of experiments was performed at the Kyoto University Critical Assembly (KUCA). To simulate the pulse mode operation of the ADSR, a D-T pulsed neutron generator was employed. As a result, it was observed that the α-values measured by these techniques clearly depended on the subcriticality, ranging from

Measurement and analysis of leakage neutron spectra from spherical assemblies of chromium, manganese and copper with 14 MeV neutrons

0.65 to

Measurement of Leakage Neutron Spectra from a Spherical Pile of Zirconium Irradiated with 14MeV Neutrons and Validation of Its Nuclear Data

2.72. Furthermore, through measurements in transient states that were realised by operating control rods, it was found that one can detect subcriticality changes in quasi-real-time.

EXPERIMENTAL STUDY ON ACCELERATOR DRIVEN SUBCRITICAL REACTOR BY USING THE KYOTO UNIVERSITY CRITICAL ASSEMBLY (KUCA)

In order to make benchmark validation of the nuclear data for Cr, Mn and Cu in existing modern evaluated nuclear data files, leakage neutron spectra in the energy range from 0.1 to 15 MeV from spherical sample piles were measured by a time-of-flight technique using intense 14 MeV neutron source, OKTAVIAN. To make a qualitative understanding of the dependence of the spectra on individual partial cross sections, a trial calculation was performed by Monte Carlo neutron transport code MCNP-4A along with several kinds of modified cross sections for Mn. The measured spectra were compared with those by theoretical calculation using MCNP-4A with the evaluated nuclear data files. The comparison was made by the spectrum shape and by the C/E values in four energy regions. The calculations for Cr, Mn and Cu using JENDL-3.2, JENDL Fusion File, FENDL-1.0 and EFF-2 give satisfactory predictions with a few exceptions. By checking the result of the trial calculation, it is considered that these disagreement is caused by improper evaluation either of inelastic scattering cross section values and/or energy distribution of secondary neutrons.

Integral Test of JENDL-3.3 with Shielding Benchmarks

For validation of nuclear data files for Nb, the absolute leakage neutron spectrum from a spherical pile of Nb was measured in the energy range between 0.1 and 16MeV at OKTAVIAN. A sensitivity analysis was performed with MCNP-4A and the modified JENDL-3.2 libraries. The measured spectrum was compared with the MCNP-4A calculation with two nuclear data libraries processed from JENDL-3.2 and ENDF/B-VI. The comparison was made by the spectrum shape and the C/E values in four energy regions. The JENDL-3.2 calculation gives satisfactory prediction except below 0.8 MeV, but that with ENDF/B-VI considerably overestimates the spectrum above 1 MeV. From the observation of the calculated and measured spectra and of the secondary energy distribution spectra (SED) of the (n, 2n) and inelastic scattering in JENDL-3.2 and ENDF/B-VI, the discrepancies in the calculations are probably due to improper SED of the (n, 2n) reaction for that below 0.8 MeV in JENDL-3.2 and that between 0.8 and 6 MeV and SED of inelastic scattering for the secondary neutron energy between 5 and 10 MeV in ENDF/B-VI. Therefore, the (n, 2n) reaction values in JENDL-3.2 library was reduced by 20% and the SED below 0.8 MeV was replaced with that of ENDF/B-VI, and this library reproduced the experiment within about 10% discrepancy over all energy regions.

Measurement and Analysis of Leakage Neutron Spectrum from a Spherical Pile of Silicon with Incident 14 MeV Neutrons

A series of basic experiments for an accelerator driven subcritical reactor (ADSR) has been performed at the Kyoto University Critical Assembly (KUCA) by combining a critical assembly with a Cockcroft-Walton type accelerator in view of a future plan to establish a new neutron source for research. Fourteen MeV neutrons were injected into a subcritical system through the polyethylene reflector. By varying subcriticality, the neutron multiplication and the prompt neutron decay constant as well as the subcriticality were measured mainly by an optical fiber detector system. Calculations were executed by a continuous energy Monte Carlo code MVP on the basis of JENDL-3.2 to examine the accuracy of neutronics design for the ADSR at the present stage. A large discrepancy was observed between the measured subcriticality and the calculated one mainly because of an inadequate evaluation of 235 U nuclear data compiled in JENDL-3.2, which leads to a large difference among the measured and calculated nuclear parameters of the ADSR. Through the present study, it was strongly recognized that the present tools for the neutronics design calculation of ADSR are not accurate enough especially to predict the neutron multiplication in the ADSR. Although the accuracy of keff calculation is becoming better and better in the current neutronics design tools for nuclear reactors, one should be careful that the neutron multiplication in the ADSR does not depend on keff itself, but approximately on 1/(1-keff).

Integral Experiments with Regard to the Thorium-Based Hybrid Fusion Blanket

Integral test of neutron and gamma-ray production data for the latest version of Japanese Evaluated Nuclear Data Library, Version 3.3 (JENDL-3.3) has been performed by using shielding benchmarks. An evaluation scheme for shielding benchmark analysis established in Japanese Nuclear Data Committee (JNDC) was applied to the integral test for medium-heavy nuclei such as Aluminum, Sodium, Titanium, Vanadium, Chromium, Iron, Cobalt, Nickel, Copper, Niobium and Tungsten. Calculations were made based on a continuous-energy Monte Carlo code MCNP4B/4C and multi-group discrete ordinates codes ANISN and DORT. The latest version of NJOY99 was employed to generate cross-section libraries for these transport codes. Calculations with JENDL-3.2, ENDF/B-VI, FENDL-1 and FENDL-2 were also made for comparison. In the present study, benchmark results of neutron and gamma-ray production data are shown for Sodium, Iron, Vanadium, Tungsten, Nickel, Titanium, Chromium, Niobium and Cobalt

Measurement of Leakage Neutron Spectra from Various Spherical Piles of Five Elements with 14 MeV Neutrons

Absolute leakage neutron spectrum in the energy range between 0.1 and 15 MeV from a silicon spherical pile of 60-cm diameter was measured using an intense 14 MeV pulsed neutron source, OKTAVIAN, for the purpose of the benchmark validation of the nuclear data of silicon in JENDL-3.2, JENDL-3.3, ENDF/B-VI (release 2), ENDF/B-VI (release 5) and EFF-2.4. The neutron spectrum was compared with the calculated results using MCNP with different nuclear data. For estimating sensitivity of the individual cross section, the calculations were also performed using modified JENDL-3.2 based data libraries. It was found that the all nuclear data could predict the measured spectrum with the measurement fairly well. ENDF/B-VI (release 2) gave considerably large discrepancy over the whole energy region presumably caused by the inadequate inelastic scattering cross section values. The discrepancy found in the JENDL-3.2, JENDL-3.3 and EFF-2.4 between 5 and 10 MeV could be attributed to the too low values of the discrete inelastic scattering cross section values. The prediction with ENDF/B-VI (release 5) was almost satisfactory over whole energy region.