Analyses of neutron multiplication integral experiments on beryllium

摘要

The work is devoted to analyses of neutron multiplication integral experiments on beryllium, carried out at Southwest Institute of Nuclear and Chemistry (SWINPC), Chengdu, China. Four parts are included. Part II gives a brief description of the SWINPC beryllium experiments. In Part III, Monte Carlo transport code MCNP is used to simulate the experiments in every detail, so as to analyze the questions concerning the experimental method and procedure, including validity of measurement by an ideal U-235 fission chamber detector, influence of the impurities in beryllium samples and D-T source structure materials, anisotropic effect of the experiment system, calculational factors and their errors. Part IV puts emphases on perturbation of the detecting system to measured values, caused by the in-system detecting procedure. ANISN-MORSE coupled method is developed to deal with the problem. Calculational results give a disappointing conclusion on the detecting system. Up to 6% integral deviations occur to measured results, becoming experimental system errors. Corrections need to be made. In Part V, the beryllium experiments are simplified to a one-dimensional benchmark problem. The final experimental neutron leakage multiplications are compared with the benchmark calculations by using ENDF/B-IV, ENDF/B-VI, JENDL-3 and LASL-Sub beryllium data. Large differences are found between the neutron leakage multiplications by experiment and calculation, far beyond the experimental errors reported by experimenters, suggesting the inefficiency of the current beryllium evaluated data. Further studies on beryllium nuclear data are recommended. (author). 11 refs, 9 figs, 9 tabs. (Atomindex citation 26:068507)