摘要:

AbstractA perturbation method is developed to find solutions of sloshing ion distributions. This method uses an expansion in the ratio of electrostatic potential to average ion energy to simplify the bounce-averaged Fokker-Planck equation. Finite element techniques, which provide rapid numerical solutions for parametric studies of sloshing ions, are used to derive the zeroth-order angular and velocity equations. The first-order two-dimensional equation was also expanded into finite element“hat functions.”Application of Galerkin’s method gives a linear system of equations where all matrix and source elements are calculated analytically. The density ratio and the potential profiles as functions of axial distance are computed. There is excellent agreement with results from the Lawrence Liver-more National Laboratory bounce-averaged Fokker-Planck code with as much as 500 times and 50 times less Cray-1 computer time for the zeroth- and the first-order solutions, respectively.

ISSN:0029-5639    

DOI:10.13182/NSE86-A17868

出版商:Taylor&Francis    

年代:1986

数据来源: Taylor

摘要:

AbstractA one-dimensional radial hybrid code was written to study the start-up of the field-reversed mirror via neutral beam injection. This code, named FROST (Field-ReversedOne-dimensionalSTart-up), models the plasma as azimuthally symmetric with no axial dependence. A multigroup method in energy and canonical angular momentum describes the large orbit ions from the beam. This method is designed to be more efficient than those employing particle tracking, since the characteristic time scale of the simulation is the ion slowing down time, rather than the much shorter cyclotron period. A time-differentiated Grad-Shafranov equation couples the ion current to massless fluid equations describing electrons and low-energy ions. Flux coordinates are used in this fluid model, in preference to a Eulerian framework, so that coupling of plasma at the two different radii of a closed flux surface can be treated with ease. Since a fluid treatment for electrons is invalid near afield null, a separate model for the electron current was included for this region, which is a unique feature. Results of simulation of injection into a 2XIIB-like plasma are discussed. Electron currents are found to retard, but not prevent, reversal of the magnetic field at the plasma center.

ISSN:0029-5639    

DOI:10.13182/NSE86-A17869

出版商:Taylor&Francis    

年代:1986

数据来源: Taylor

摘要:

AbstractA linear discontinuous Galerkin treatment is developed for the continuous slowing down operator. Multigroup Legendre coefficients are derived that allow this treatment to be implemented in standard Sncodes, A synthetic acceleration scheme is developed to accelerate the outer iterations required by this treatment. A computational comparison between the diamond difference treatment and the linear discontinuous treatment is given.

ISSN:0029-5639    

DOI:10.13182/NSE86-A17870

出版商:Taylor&Francis    

年代:1986

数据来源: Taylor

摘要:

AbstractTwo discrete ordinates methods, the Snand streaming ray methods, are applied to electron transport problems. Calculational results in the form of energy deposition profiles are compared with those obtained by the method of moments for the case of a 200-keV plane perpendicular source embedded in infinite aluminum. In a second set of calculations, Snand streaming ray flux data are compared with results of a recently developed analytical benchmark technique applied to the solution of the energy-independent form of the Spencer-Lewis equation for electron transport. Satisfactory agreement with moments and analytical benchmark calculations is found. Areas of divergence among the various calculational methods are examined.

ISSN:0029-5639    

DOI:10.13182/NSE86-A17871

出版商:Taylor&Francis    

年代:1986

数据来源: Taylor

摘要:

AbstractA semi-implicit method for solving the full compressible resistive magnetohydrodynamic equations in three dimensions is presented. The method is designed for use in the modeling of fusion plasmas in magnetic confinement devices. The method is unconditionally stable with respect to the fast compressional modes. The time step is limited instead by the slower shear Alfvén motion. The computing time required for one time step is essentially the same as that for explicit methods. The code is applied to resistive instabilities in cylindrical tokamak equilibria.

ISSN:0029-5639    

DOI:10.13182/NSE86-A17872

出版商:Taylor&Francis    

年代:1986

数据来源: Taylor

摘要:

AbstractAn experience with the vectorization of the light water reactor transient analysis code RELAP5/MODI on a vector supercomputer FACOM VP-100 (peak speed 250 million floating point operations/s, clock period 7.5 ns) is described. The approach to the vectorization is based on the junction and volume level parallelisms for the hydrodynamic model, and the heat structure and heat mesh levels for the heat transfer model. The VP-100 vectorized code version yields a 2.4 to 2.8 factor speed increase over the FACOM M-380 computer, depending on the number of spatial cells being used. The M-380 is an IBM-type computer with the same speed as the VP-100 in scalar mode.

ISSN:0029-5639    

DOI:10.13182/NSE86-A17873

出版商:Taylor&Francis    

年代:1986

数据来源: Taylor

摘要:

AbstractNumerical experiments performed on a single instruction multiple data-pipeline vector parallel (SIMD-PVP) architecture computing machine, e.g., a CRAY X-MP/48, demonstrate that current nuclear reactor systems codes can be restructured for concurrent multiprocessing and show wall clock performance improvements of 1.5 to 3.0 on a 4-CPU machine, depending on plant model, problem type, and problem length. In addition, algorithm development studies indicate that up to a 20% speedup can be obtained by a new class of parallel numerical methods. Faster-than-real-time simulation has been demonstrated utilizing RELAP5/MOD1 and a pressurized water reactor plant model characteristic of licensing and/or safety analysis calculations. A theoretical analysis indicates that five to ten times faster than real-time computation may be possible for this class of problems utilizing this or the next generation of SIMD-PVP architecture machines, such as the CRAY X-MP/48, and new computer codes optimized for such machines.

ISSN:0029-5639    

DOI:10.13182/NSE86-A17874

出版商:Taylor&Francis    

年代:1986

数据来源: Taylor

摘要:

AbstractAdvanced technology for high-speed interactive nuclear power plant simulations is of great value for timely resolution of safety issues, for plant monitoring, and for computer-aided emergency responses to an accident. Presented is the methodology employed at Brookhaven National Laboratory to develop a boiling water reactor (BWR) plant analyzer capable of simulating severe plant transients at much faster than real-time process speeds.Five modeling principles are established and a criterion is given for selecting numerical procedures and efficient computers to achieve the very high simulation speeds. Typical results are shown to demonstrate the modeling fidelity of the BWR plant analyzer.

ISSN:0029-5639    

DOI:10.13182/NSE86-A17875

出版商:Taylor&Francis    

年代:1986

数据来源: Taylor

摘要:

AbstractThe finite element solution of fluid/structure interaction problems is considered for a class of acousto-elastic problems where the fluid is linear acoustic and the structure is linear elastic. The finite element formulation in terms of fluid pressure and structural displacement results in a system of unsymmetric equations. Due to the complexities of eigensolution for large systems involving unsymmetric matrices, little progress has been reported. Recently, Yu showed that the real form of QZ algorithm can be used for solving small unsymmetric eigenproblems of fluid/structure interaction, and, as a major advance, now presents the use of the subspace iteration method, in conjunction with the QZ algorithm, for solving large fluid/structure systems. The computational procedure is similar to that for the real symmetric case, and the procedure can easily be adopted by any finite element code.

ISSN:0029-5639    

DOI:10.13182/NSE86-A17876

出版商:Taylor&Francis    

年代:1986

数据来源: Taylor

摘要:

AbstractA method of computing hydraulic jet reaction force at a 100% steamline break is presented. The steamline break analysis is performed by the use of the NOTRUMP code. Formation of supersonic flow at the converging-diverging flow restrictor is computed with a homogeneous equilibrium model. Consideration is given, however, to the slip ratio of the liquid and steam phases, thermal nonequilibrium effects, and normal and oblique shocks.

ISSN:0029-5639    

DOI:10.13182/NSE86-A17877

出版商:Taylor&Francis    

年代:1986

数据来源: Taylor