No abstract available

No abstract available

[en] Short communication

[en] Spectral indices have been measured by foil activation technique in a number of different D₂O-moderated lattices in the Swedish zero power reactor R0 and the pressurized exponential assembly TZ. In most cases the fuel was in the form of single rods, distributed uniformly in the lattice. Parameters in these cases were lattice pitch and fuel composition. A 31-rod cluster lattice was also investigated, with the moderator temperature varying up to 210 deg C. On the basis of these measurements, as well as measurements on cluster lattices, reported by other investigators, it has been possible to derive simple correlations for the spectral indices, which seem to be of fairly general validity for D₂O lattices. The experimental results have also been compared to calculations with the multigroup collision probability program FLEF

[en] The integral transport equation will be solved in square unit cells by assuming the existence of a fundamental mode. The equations governing the B_n method are given without making the small buckling approximation. First, the angular flux is factorized into two parts: a periodic microscopic fine structure flux, and a macroscopic form with no angular dependence. The macroscopic form only depends on a buckling vector with a given orientation. The critical buckling norm along with the corresponding fine structure flux are obtained using collision probability (CP) calculations which are repeated until criticality is achieved. This procedure allows the boundary conditions of the unit cell to be taken into account using closed-form contributions obtained from the cyclic tracking technique. Numerical results are presented for heterogeneous problems with isotropic scattering kernels. The applicability of this heterogeneous B_n model to accurately follow voiding effects is also considered

[en] In this paper we develop a methodology based on the use of the Fuzzy Logic technique to build multi-objective functions to be used in optimization processes applied to in-core nuclear fuel management. As an example, we selected the problem of determining optimal radial fuel enrichment and gadolinia distributions in a typical 'Boiling Water Reactor (BWR)' fuel lattice. The methodology is based on the use of the mathematical capability of Fuzzy Logic to model nonlinear functions of arbitrary complexity. The utility of Fuzzy Logic is to map an input space into an output space, and the primary mechanism for doing this is a list of if-then statements called rules. The rules refer to variables and adjectives that describe those variables and, the Fuzzy Logic technique interprets the values in the input vectors and, based on the set of rules assigns values to the output vector. The methodology was developed for the radial optimization of a BWR lattice where the optimization algorithm employed is Tabu Search. The global objective is to find the optimal distribution of enrichments and burnable poison concentrations in a 10*10 BWR lattice. In order to do that, a fuzzy control inference system was developed using the Fuzzy Logic Toolbox of Matlab and it has been linked to the Tabu Search optimization process. Results show that Tabu Search combined with Fuzzy Logic performs very well, obtaining lattices with optimal fuel utilization. (authors)

[en] A coarse-mesh finite difference method has been developed for multidimensional, mixed-lattice reactor diffusion calculations, both statics and kinetics, in hexagonal geometry. Results obtained with the coarse-mesh (CM) method have been compared with a conventional mesh-centered finite difference method and with experiment. The results of this comparison indicate that the accuracy of the CM method for highly heterogeneous (mixed) lattices using one point per hexagonal mesh element (''hex'') is about the same as the conventional method with six points per hex. Furthermore, the computing costs (i.e., central processor unit time and core storage requirements) of the CM method with one point per hex are about the same as the conventional method with one point per hex

[en] The present report is based on lectures delivered at the Workshop on Nuclear Reaction Data and Nuclear Reactors: Physics, Design and Safety hold in International Centre of Theoretical Physics, Trieste, in March 1998. The main goal of the set of lectures was to give the basis of reactor physics calculations for participants working on nuclear data.The last lectures, devoted to WIMS including the WIMSD code users. Following this general line the material is divided into three parts: The first part includes a short description of neutron transport phenomena limited to those definitions that are necessary to understand the approach to practical solution of the problem given in the second part on reactor lattice transport calculations. The detailed discussion of the neutron cross sections has been skipped as this subject has been treated in detail by other lectures. In the third part those versions of the well-known WIMSD code which are distributed by NEA Data Bank are described. The general structure of the code is given supplied in a more detailed description of aspects being the most common points of misunderstanding for the code users. (author)

No abstract available

[en] A cylindrical water-uranium lattice depending on three parameters is optimized taking into consideration seven performance criteria. Actually all optimum design problems are multicriterial, and it is not clear what is an optimum solution. A traditional approach to these problems is to convert them into problems with a single criterion or sequences of such problems. However, all conversion procedures require additional data that are assigned arbitrarily and more than often optimum solutions of converted problems turn out impractical. A method called Parameter Space Investigation was developed earlier. In the Soviet Union more than 50 reports have been published on applications of this method in computer aided design but only two papers have been concerned with reactor problems. The main aim of the present paper is to demonstrate how the method works. We assume that a Monte Carlo program is available that enables us to compute the performance of the system under investigation for an arbitrary fixed set of the decision variables that are called parameters. (author)