[en] Neutron transport in nuclear reactors is quite well modelled by the linear Boltzmann transport equation. Its solution is relatively easy, but unfortunately too expensive to achieve whole core computations. Thus, we have to simplify it, for example by homogenizing some physical characteristics. However, the solution may then be inaccurate. Moreover, in strongly homogeneous areas, the error may be too big. Then we would like to deal with such an inconvenient by solving the equation accurately on this area, but more coarsely away from it, so that the computation is not too expensive. This problem is the subject of a thesis. We present here some results obtained for slab geometry. The couplings between the fine and coarse discretization regions could be conceived in a number of approaches. Here, we only deal with the coupling at crossing the interface between two sub-domains. In the first section, we present the coupling of discrete ordinate methods for solving the homogeneous, isotropic and mono-kinetic equation. Coupling operators are defined and shown to be optimal. The second and the third sections are devoted to an extension of the previous results when the equation is non-homogeneous, anisotropic and multigroup (under some restrictive assumptions). Some numerical results are given in the case of isotropic and mono-kinetic equations. (author)

[en] To calibrate theoretical calculation result of reflecting neutrons effects for carbon nucleus, the reflecting neutrons effects of 14 MeV neutrons after penetrating three composed materials were measured with capture detector. The errors of experimental measurement is better than 10%. The experimental results accord with the values of theoretical calculation (Monte Carlo Method) in the experimental errors

[en] Three moderator materials, AlF₃/Al, D₂O and LiF, have been analyzed for clinical usefulness using the reaction ⁷Li(p,n) as an accelerator driven neutron source. Proton energies between 2.1 MeV and 2.6 MeV have been investigated. Radiation transport in the reflector/moderator assembly is simulated using the MCNP program. Depth-dose distributions in a head phanton are calculated with the BNCT-RTPE patient treatment planning program from INEEL using the MCNP generated neutron and photon spectra as the subsequent source. Clinical efficacy is compared using the current BMRR protocol for all designs. Depth-dose distributions are compared for a fixed normal tissue tolerance dose of 12.5 Gy-Eq. Radiation analyses also include a complete anthropomorphic phantom. Results of organ and whole body dose components are presented for several designs. Results indicate that high quality accelerator beams may produce clinically favorable treatments to deep-seated tumors when compared to the BMRR beam. Also discussed are problems identified in comparing accelerator and reactor based designs using in-air figures of merit as well as some results of spectrum-averaged RBE's

[en] A Monte Carlo neutron transport kernel capability primarily for ex-core neutron response is described. The capability consists of the generation of a set of response kernels, which represent the neutron transport from the core to a specific ex-core volume. This is accomplished by tagging individual neutron histories from their initial source sites and tracking them throughout the problem geometry, tallying those that interact in the geometric regions of interest. These transport kernels can subsequently be combined with any number of core power distributions to determine detector response for a variety of reactor Thus, the transport kernels are analogous to an integrated adjoint response. Examples of pressure vessel response and ex-core neutron detector response are provided to illustrate the method

[en] The spectrometer of polarized neutrons aimed for investigations of interactions of polarized neutrons with nuclei and condensed matters over the energy range from thermal to several eV has been created on the IBR-2 pulsed reactor in JINR (Dubna). The magnetized Co(92%)-Fe(8%) monocrystals are used for neutron polarization and polarization analysis. Neutron polarization is approximately equal to 95% over the whole range of energy. (author)

[en] Investigation and assessment are made for approximate design calculation methods of radiation streaming in shield irregularities. Investigation is made for (1) source, (2) definition of streaming radiation components, (3) calculation methods of streaming radiation, (4) streaming formulas for each irregularity, (5) difficulties in application of streaming formulas, etc. Furthermore, investigation is made for simple calculation codes and albedo data. As a result, it is clarified that streaming calculation formulas are not enough to cover various irregularities and their accuracy or application limit is not sufficiently clear. Accurate treatment is not made in the formulas with respect to the radiation behavior for slant incidence, bend part, offset etc., that results in too much safety factors in the design calculation and distrust of the streaming calculation. To overcome the state and improve the accuracy of the design calculation for shield irregularities, it is emphasized to assess existing formulas and develop better formulas based on systematic experimental studies. (author)

[en] This report describes the philosophy and structure of MCLIB, A Fortran library of Monte Carlo subroutines which has been developed to test designs of neutron scattering instruments. Emphasis is placed on new features added to the library since the previous presentation of MCLIB at ICANS-XIII in October, 1995. These new features include toroidal mirrors, writing and reading source files, splitting and banking of histories, and a Maxwellian probability distribution. The only change of a program structure has been to include charge and polarization vector in the description of a particle. The latest release of the source code and documentation may be obtained by anonymous ftp. Work is also continuing on a more friendly web-based user interface, and user input is requested for additional features to be added to the library

[en] In Czubek's method of measurement of the thermal neutron macroscopic absorption cross section a two-region geometry is applied where the investigated sample is surrounded by an external moderator. Both regions in the measurements made up till now were hydrogenous, which means the same type of the thermal neutron transport properties. In the paper a theoretical consideration to use non-hydrogenous materials as the samples is presented. Pulsed neutron measurements have been performed on homogeneous material in a geometry of the classic experiment with the variable geometric buckling. Two decay constants have been measured for different cylindrical samples of small lead grains (a lead shot). (author)

[en] Neutron transports in different diffusing and thermalising media were studied within one dimensional theory. Macroscopic cross section libraries for each medium or region were generated by one dimensional models that represent the geometry of the surrounding regions. Few group total and angular fluxes are computed. Especially, determination of angular fluxes at some points and directions are focused on. The results are compared with other computed and experimental values

[en] The data for 340 nuclides will be stored in the second revised version of JENDL-3, JENDL-3.2, and about half of them will be modified more or less. An outline of modifications for JENDL-3.2 is described here in particular on double-differential cross sections, total cross sections and inelastic scattering cross sections of structural materials, data of fission product nuclides, resonance parameters and fission spectra of heavy nuclides, and γ-ray production data. (author)