Filters

Results

**1**-**10**of**1844** Results

**1**-**10**of**1844**. Search took:**0.027**secondsSort by: date | relevance |

Ragheb, M.M.H.

Wisconsin Univ., Madison (USA)

Wisconsin Univ., Madison (USA)

AbstractAbstract

[en] New estimation approaches for nuclear reactor calculations by Monte Carlo are developed and investigated, with the purpose of surmounting or alleviating existing difficulties facing the application of the Monte Carlo Method. Bias-free estimators which extract more information on a given particle tract than currently used estimators are deduced. These depend on estimating separately the different Neumann series terms of the solution to the Boltzmann Transport Equation, and are based on underlying absorbing or nonabsorbing Markov Chain random walk models. They are applied to representative slowing-down and deep penetration problems. Effective biases arising from application of currently used methods are highly suppressed. Comparison to analytical solutions and to currently used methods shows their distinctive advantages. An approach for the systematic determination of optimal biasing parameters in importance sampling calculations, depending on particle tracks scaling; which avoids the effective biases and infinite variances in these calculations, is developed and applied to the deep penetration problem of particle transport. The use of stationary functionals from variational theory for variance reduction in Monte Carlo calculations is discussed. Suggestions for alternative error estimation methods in Monte Carlo calculations, based on functionals of the second moment for the collision and last collision estimators, or on reconstructing the sample distributions from sample moments, are exposed. Implementation of the suggested new approaches in future Monte Carlo calculations is discussed

Primary Subject

Secondary Subject

Source

1978; 503 p; University Microfilms Order No. 78-23,084; Thesis (Ph. D.).

Record Type

Report

Literature Type

Thesis/Dissertation

Country of publication

Publication YearPublication Year

Reference NumberReference Number

INIS VolumeINIS Volume

INIS IssueINIS Issue

Hong, K.J.

Kansas State Univ., Manhattan (USA)

Kansas State Univ., Manhattan (USA)

AbstractAbstract

[en] Several computational techniques for treating highly anisotropic transport problems are investigated. A detailed analysis of the neutron scattering process indicates that the multigroup transfer cross sections typically have very small ranges of scattering angular support for fine group structure and light nuclei. These elastic and inelastic multigroup transfer cross sections can generally be well represented by low-order polynominals between the natural break points within the ranges of scattering angular support. It is shown that the existing techniques used for evaluating the multigroup transfer coefficients for neutron elastic and inelastic scattering can be significantly improved by taking advantage of inherent features of the exact transfer cross section. In addition, an approximate analytical technique is developed to calculate the Legendre expansion coefficients accurately and efficiently for fine energy group structure. The exact kernel method, which is suited to highly anisotropic problems, is generalized to solve spherical, cylindrical, and azimuthally dependent plane geometry transport problems. The EXAKER1N computer code developed in this study optimizes the use of the pheripheral storage units and computer core memory such that the inherent severe storage requirements of the exact kernel method is eased. Codes have also been developed to generate from ENDF data tapes the exact transfer cross sections required by the exact kernel method. A low-order, piecewise polynomial representation of the cross section is shown to yield analytical results for the azimuthal averaged transfer cross section. This approximation is found to give accurate cross section values with far less computational effort than previously needed. For extremely high anistropic problems, the kernel method requires very high quadrature order to ensure accurate results. A new technique is developed to handle this type of extreme problem

Original Title

EXAKER1N

Primary Subject

Source

1980; 268 p; University Microfilms Order No. 80-15,251; Thesis (Ph. D.).

Record Type

Report

Literature Type

Thesis/Dissertation

Country of publication

Publication YearPublication Year

Reference NumberReference Number

INIS VolumeINIS Volume

INIS IssueINIS Issue

Cardona, Augusto V.

Rio Grande do Sul Univ., Porto Alegre, RS (Brazil). Escola de Engenharia

Rio Grande do Sul Univ., Porto Alegre, RS (Brazil). Escola de Engenharia

AbstractAbstract

[en] In this work it is presented a generic method of analytical solution to the one-dimensional S

_{N}, P_{N}, W_{N}, Ch_{N}, A_{N}and L D_{N}approximations of the linear transport equation. The main idea of this method consists in the application of the Laplace transform to solve the differential equation system related to the considered approximations and solution of the resultant algebraic system by Trzaska's algorithm. (author). 46 refs., 3 figs., 15 tabsOriginal Title

Metodo generico de solucao analitica para aproximacoes da equacao linear de transporte

Primary Subject

Source

May 1996; 96 p; Available from the Library of Brazilian Nuclear Energy Commission, Rio de Janeiro; Tese (Ph.D.).

Record Type

Miscellaneous

Literature Type

Thesis/Dissertation

Country of publication

Publication YearPublication Year

LanguageLanguage

Reference NumberReference Number

INIS VolumeINIS Volume

INIS IssueINIS Issue

El-Hajjami, Said

Ecole Polytechnique, Montreal, Quebec (Canada)

Ecole Polytechnique, Montreal, Quebec (Canada)

AbstractAbstract

[en] Many physical phenomena treatments, such as Doppler broadening and self shielding, are not included in the microscopic cross sections on ENDF files. These treatments affect largely the cell calculations. Moreover, the majority of the cell codes are based on deterministics approach, which resolve the multigroup transport equation and thus require multigroup constants. The quality of multigroup cross sections cannot be judged independently without the particular combination of data and methods in the complex environment of reactor analysis. In order to validate the adequacy of the newly generated multigroup data library for design analysis, benchmarking were performed using the DRAGON code. Using the NJOY code, a new 69 groups library, based on the WIMS-D4 format was produced. The newly generated library was then validated by three benchmarks, Mosteller, Rowlands and feedback effects in CANDU-6 cell. The feedback effects analysed in this study are void and temperature effects. The multigroup data recently produced were compared to WIMS-AECL, Monte-Carlo results, and to the libraries available at the IGN. (author)

Original Title

Production d'une interface-librairie de sections efficaces integree au logiciel DRAGON

Primary Subject

Source

2001; 181 p; ISBN 0-612-73407-2; ; Available from University Microfilms International-UMI, 300 North Zeeb Road, PO Box 1346, Ann Arbor, Michigan (United States) under document order no. MQ73407; Thesis (M.Sc.A.)

Record Type

Miscellaneous

Literature Type

Thesis/Dissertation

Country of publication

Publication YearPublication Year

Reference NumberReference Number

INIS VolumeINIS Volume

INIS IssueINIS Issue

AbstractAbstract

[en] The principle of the curved neutron guide is to transport neutrons far away from the reactor core with as minimum particle loss as possible. After a series of total reflection,the neutron beam is no longer visible from the reactor core and consequently, gamma radiations and fast neutrons emitted from the core are scattered by the walls of the guide and absorbed by the biological shielding set around the guide. The curved neutron guide provides a high-quality beam of slow neutrons. The first chapter deals with the theoretical concept of curved guide, we have determined the parameters for the setting of such a guide in the EL3 reactor at Saclay (France). The different tolerances on the state the surface, on the alignment of the different parts of the guide, on the waving of the guide wall have been assessed. The second chapter presents the technical solution chosen that complies to all the required specifications. The curved neutron guide has been designed for neutrons with wavelength of 4 Angstroms, it is 29 m long, has a bending radius of 835 m and is composed of 87 rectangular components made of glass plates on which a 1500 angstrom thick layer of nickel has been deposited. Each component is set with a fixed angle of (4±0.25)*10

^{-4}radians from the previous component in order to form the bending radius. The last chapter is dedicated to the neutron flux measurement made at the end of the neutron guideOriginal Title

Guide courbe conducteur de neutrons

Primary Subject

Source

4 Mar 1969; 110 p; 8 refs.; Available from the INIS Liaison Officer for France, see the 'INIS contacts' section of the INIS website for current contact and E-mail addresses: http://www.iaea.org/inis/Contacts/; These Docteur de l'Universite, Mention Sciences

Record Type

Report

Literature Type

Thesis/Dissertation

Report Number

Country of publication

Publication YearPublication Year

Reference NumberReference Number

INIS VolumeINIS Volume

INIS IssueINIS Issue

Alonso-Vargas, G.

Instituto Politecnico Nacional, Mexico City (Mexico). Escuela Superior de Fisica y Matematicas

Instituto Politecnico Nacional, Mexico City (Mexico). Escuela Superior de Fisica y Matematicas

AbstractAbstract

[en] A computer program has been developed which uses a technique of synthetic acceleration by diffusion by analytical schemes. Both in the diffusion equation as in that of transport, analytical schemes were used which allowed a substantial time saving in the number of iterations required by source iteration method to obtain the K

_{e}ff. The program developed ASD (Synthetic Diffusion Acceleration) by diffusion was written in FORTRAN and can be executed on a personal computer with a hard disc and mathematical O-processor. The program is unlimited as to the number of regions and energy groups. The results obtained by the ASD program for K_{e}ff is nearly completely concordant with those of obtained utilizing the ANISN-PC code for different analytical type problems in this work. The ASD program allowed obtention of an approximate solution of the neutron transport equation with a relatively low number of internal reiterations with good precision. One of its applications would be in the direct determinations of axial distribution neutronic flow in a fuel assembly as well as in the obtention of the effective multiplication factor. (Author)Original Title

Solucion numerica de la ecuacion de transporte de neutrones en geometria plana mediante esquemas analiticos empleando aceleracion por difusion sintetica

Primary Subject

Source

1991; 136 p; Thesis (M. Sc.).

Record Type

Miscellaneous

Literature Type

Thesis/Dissertation

Report Number

Country of publication

Publication YearPublication Year

Reference NumberReference Number

INIS VolumeINIS Volume

INIS IssueINIS Issue

Fan, C.P.W.

Texas A and M Univ., College Station (USA)

Texas A and M Univ., College Station (USA)

AbstractAbstract

[en] An approximate solution method has been developed to solve the one-dimensional, steady-state neutron transport problems in plane and spherical geometries. The spherical harmonics expansion and the multigroup approximation are employed to represent the angular- and energy-dependence. The angular moments are replaced by a set of transformation functions that leads to the second-order form of the multigroup P/sub N/ equations. The approximate solutions of the transformation functions are formulated by a variation principle in conjunction with the cubic Hermite polynomials. Conservation constraints are imposed by the usage ofLagrange multipliers. In order to validate the numerical solutions, the analytical expressions of criticality conditions and angular moments to the multigroup P/sub N/ equations are constructed by applying the eigenfunction expansion technique. This analytical approach is further extended for problems in cylindrical geometry. In this study, both external sources and criticality problems are addressed. Accuracy and reliability of the approximate solution methods are investigated by comparing with the benchmark calculations or other conventional methods. Preliminary results are reported, and recommendations for future research are made

Primary Subject

Source

1984; 239 p; University Microfilms Order No. 84-28,753; Thesis (Ph. D.).

Record Type

Report

Literature Type

Thesis/Dissertation

Country of publication

Publication YearPublication Year

Reference NumberReference Number

INIS VolumeINIS Volume

INIS IssueINIS Issue

Ahmad, S.A.

California Univ., Los Angeles (USA)

California Univ., Los Angeles (USA)

AbstractAbstract

[en] This research focuses on the determination of effective multiplication and accurate flux distributions in three dimensions. Since the direct solution is very time consuming, a method for expediting the solution is proposed. Specifically, the objectives of the research were to: (1) develop a technique for accelerating the solution of the particle transport equation in three dimensional, cartesian geometry for determining the effective multiplication and the particle flux distribution. (b) apply this technique to sample problems. The technique developed was based on the fact that the numerical solutions of the corresponding diffusion equation converge much faster. The difference lies mainly in the treatment of particle leakage. Hence, if the diffusion coefficient in the diffusion leakage term were computed from the transport angular fluxes, rather than from the nuclear properties of the medium, then the resulting equation would have the desirable properties of a solution that matches the transport solution, but is computationally cost effective. The multigroup Discrete Ordinates finite difference technique in three dimensional cartesian geometry is used for the transport equation. This method was then applied to solve five nuclear reactor criticality problems. It was observed that the solutions are accurate and decrease computing time 3-11, fold depending upon the nature of the problem

Primary Subject

Secondary Subject

Source

1987; 145 p; University Microfilms Order No. 87-19,912; Thesis (Ph.D).

Record Type

Report

Literature Type

Thesis/Dissertation

Country of publication

Publication YearPublication Year

Reference NumberReference Number

INIS VolumeINIS Volume

INIS IssueINIS Issue

Flores Calderon, J.E.

Instituto Politecnico Nacional, Mexico City. Escuela Superior de Fisica y Matematicas

Instituto Politecnico Nacional, Mexico City. Escuela Superior de Fisica y Matematicas

AbstractAbstract

[en] Experiments with non-stationary neutron transport in large cavity moderators (l>>Σsub(tr)

^{-1}) (where l is the characteristic cavity length and Σsub(tr)^{-1}the macroscopic transport section of the moderator) led to the method reported in this study which, based on neutron impulses for determining albedo of thermal neutrons, gave a precision greater by an order of magnitude over previous methods. A sufficient time interval after introduction of the neutron flux into the moderator chamber decreased exponentially the decay constant L, which was itself related to albedo by a function called f. Numerical calculations of albedo were assisted. (author)Original Title

Determinacion del albedo por el metodo del impulso neutronico

Primary Subject

Source

1982; 85 p; Tesis (M. in Sci.).

Record Type

Miscellaneous

Literature Type

Thesis/Dissertation

Report Number

Country of publication

Publication YearPublication Year

Reference NumberReference Number

INIS VolumeINIS Volume

INIS IssueINIS Issue

Harvey, J.T.

Arkansas Univ., Little Rock (USA)

Arkansas Univ., Little Rock (USA)

AbstractAbstract

[en] Neutron spectra have been measured by the threshold foil activation technique for the White Sands Missile Range Fast Burst Reactor. The neutron spectra for free-field, free-field with the experimenters table in place, for the in-core irradiation port (Glory Hole) and with a one-half inch section of aluminum in place are reported. Neutron transport calculations are also performed for the above mentioned spectra and for six other geometrics reported elsewhere. The absolute values of the spectral parameters derived from calculations do not agree with the spectral parameters obtained by the foil activation method, but the same trends are observed. Transport calculation results are combined with the foil activation results to give ''best value'' spectral parameters. It was found that the free-field spectrum was moderately hard and the Glory Hole spectra to be softer by approximately 18 percent when compared to the free-field spectrum. The experimenters table and two sections of aluminum of differing thicknesses were found to have no major effect on the neutron spectrum when comparing spectral parameters. The neutron spectrum behind a section on plexiglas was found to be considerably harder than the free-field spectrum. The differences between the results of the neutron transport calculations and the foil activation method are discussed

Primary Subject

Source

1977; 152 p; University Microfilms Order No. 77-23,339; Thesis (Ph. D.).

Record Type

Report

Literature Type

Thesis/Dissertation

Country of publication

Publication YearPublication Year

Reference NumberReference Number

INIS VolumeINIS Volume

INIS IssueINIS Issue

1 | 2 | 3 | Next |