摘要:

AbstractThe residual method of orthogonal collocations (OCs) is evaluated on the basis of three problems of increasing degree of complexity. Two model problems, a Poisson equation and a wave front propagation problem, allow a comparison with known analytical solutions and other numerical results obtained with finite differences or with the classical Galerkin method. The third problem consists of the numerical solution of the equations describing a one-dimensional sodium vapor flow, obtained using a variant of the BL0W-3A computer program developed for this purpose. Shape functions of second degree are used throughout the analysis. The results show the applicability of the OC technique to two-phase flow problems.

ISSN:0029-5639    

DOI:10.13182/NSE82-A19024

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

AbstractAnalytical techniques were developed to analyze pulsed neutron activation measuements in large pipes leading to a determination of the fluid velocity in the pipe. Neutron and gamma-ray Monte Carlo transport calculations were carried out at the neutron tagging and gamma-ray detector positions, for the piping sizes typical of the loss-of-fluid test (LOFT) experiment. Dispersion models were developed, to describe the transport and mixing of the irradiated fluid from the source to the detector location, and the L3-7 LOFT small break neutron activation test data were analyzed. The values of the fluid transport velocity obtained by a phenomenological model based on finite difference equations agree with those found from experiment.

ISSN:0029-5639    

DOI:10.13182/NSE82-A19025

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

AbstractAlcouffe’s diffusion-synthetic acceleration scheme for one-dimensional discrete ordinates calculations is extended to accelerate both the zero’th and first moments of the scattering source. The extended scheme is found to be significantly more effective than the standard scheme for problems with highly forward-peaked scattering. A new diffusion theory is derived directly from the discrete ordinates equations, which varies from the standard theory only in the definition of the diffusion coefficient. When employed in the standard diffusion-synthetic acceleration scheme, the new theory is found to perform slightly better than the standard diffusion theory.

ISSN:0029-5639    

DOI:10.13182/NSE82-A19026

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

AbstractWe consider the slab geometry discrete ordinates equations, with the weighted diamond, linear characteristic, linear discontinuous, and linear moments spatial differencing schemes. For each differencing scheme we derive a diffusion-synthetic, source-correction acceleration method which, for model (infinite medium, isotropic scattering, constant cross section, constant mesh spacing) problems, unconditionally reduces the spectral radius of the iteration method from the unaccelerated value of c (the scattering ratio) to

ISSN:0029-5639    

DOI:10.13182/NSE82-1

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

AbstractDiffusion-synthetic acceleration methods that have been proven analytically to be stable for model discrete ordinates problems (for infinite media, with isotropic scattering, constant cross sections, and a uniform spatial mesh) are shown to be experimentally stable for realistic problems (for finite media, with anisotropic scattering, variable cross sections, and a nonuniform spatial mesh). Also, the effect of negative flux fixups on the acceleration methods is discussed.

ISSN:0029-5639    

DOI:10.13182/NSE82-2

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

AbstractThe homogenized neutron diffusion equation for Benoist’s uncorrected diffusion coefficients is derived. The approximations in the asymptotic and buckling methods are analyzed to show that both these methods are identical and that Benoist’s corrected diffusion coefficients should not be used in the diffusion equation. Furthermore, the practical limitations of the homogenization methods are discussed, and it is pointed out that the use of Benoist’s uncorrected coefficients in the diffusion equation may be superior to the use of the coefficients of Larsen or of Deniz and Gelbard. It is also recommended, in view of the severe approximations made in the transport theory to arrive at the homogenization prescriptions, that the accuracy in the results of the homogenization methods should be examined for a few benchmarks in slab geometry which are amenable to transport theory solution.

ISSN:0029-5639    

DOI:10.13182/NSE82-A19029

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

AbstractThe application of the cell discretization (CD) method to a class of nuclear reactor problems is described. The CD method is based on partitioning the domain in which the diffusion equations are to be solved into a set of subdomains, or“cells.”This approach, which resembles that used in the finite element method, nevertheless differs from it in certain important respects, some of which are mentioned in the course of describing CD. A FORTRAN program has been written that implements many of the features of the CD method, but is restricted to rectangular geometry. Several representative problems from the literature are solved numerically with CD, and the results are compared with the published ones. The central processor unit times are given for solution of these problems on the IBM 370/158 under VM, a time-sharing system. All results, including keff, peak-to-average-power ratios, integrated fluxes, etc. are listed in tables in such a way as to make comparison convenient. Flux plots are also shown for those cases where they were given in the literature.

ISSN:0029-5639    

DOI:10.13182/NSE82-A19030

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

AbstractA new neutron scattering kernel for graphite has been developed with the frequency distribution function generated by the authors using the unfolding technique. This has been used to study the decay of neutron pulses in different graphite assemblies. This kernel (withθD= 2000 K) can give a slightly better explanation of the experimental results than those based on the Krumhansl and Brooks model or the Young and Koppel model of lattice vibrations.

ISSN:0029-5639    

DOI:10.13182/NSE82-A19031

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

AbstractCross-section libraries and input neutron spectrum data are generally required in the application of activation neutron spectrometry and are usually given in the form of group values. Fine structures (620 groups for the energy range from 10−10to 18 MeV) are often used, but problems arise if adjustment codes implying matrix calculations are adopted. In this Note, the errors deriving from the use of a coarse (40-group) structure are estimated and discussed making special reference to a fast neutron spectrum.

ISSN:0029-5639    

DOI:10.13182/NSE82-A19032

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

AbstractThe critical problem in a finite plane slab is studied by an appropriate extension of the integral transform method. For a given optical thickness, the value of the scattering ratio (or mean number of secondaries per collision) that makes the system critical is determined as an eigenvalue of a Fredholm integral equation, in which a continuous position dependence of the scattering ratio itself has been taken into account. The integral equation is solved by a suitable projection technique, and numerical results are presented and briefly discussed for a parabolic trend of the fuel enrichment inside the slab.

ISSN:0029-5639    

DOI:10.13182/NSE82-A19033

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor