Refine your search:     
Report No.
 - 
Search Results: Records 1-2 displayed on this page of 2
  • 1

Presentation/Publication Type

Initialising ...

Refine

Journal/Book Title

Initialising ...

Meeting title

Initialising ...

First Author

Initialising ...

Keyword

Initialising ...

Language

Initialising ...

Publication Year

Initialising ...

Held year of conference

Initialising ...

Save select records

Journal Articles

Fast vector computation of the characteristics method

Kugo, Teruhiko

Journal of Nuclear Science and Technology, 39(3), p.256 - 263, 2002/03

 Times Cited Count:3 Percentile:23.39(Nuclear Science & Technology)

Two vector computation algorithms; an odd-even sweep (OES) method and an independent sequential sweep (ISS) method, have been developed for the characteristics method to solve the neutron transport equation in a heterogeneous geometry. They realize long vector lengths without recursive operations for effective use of vector computers. Their efficiency has been investigated to a realistic fuel assembly calculation. For both methods, a vector computation is 15 times faster than a scalar computation. From a viewpoint of a comparison between the OES and ISS methods, the ISS method is superior to the OES method because the ISS method shows a faster convergence and saves a computer memory without reducing a computation speed.

JAEA Reports

Fast computation of the characteristics method on vector computers

Kugo, Teruhiko

JAERI-Research 2001-051, 39 Pages, 2001/11

JAERI-Research-2001-051.pdf:2.04MB

Fast computation of the characteristics method to solve the neutron transport equation in a heterogeneous geometry has been studied. Two vector computation algorithms; an odd-even sweep (OES) method and an independent sequential sweep (ISS) method have been developed and their efficiency to a typical fuel assembly calculation has been investigated. For both methods, a vector computation is 15 times faster than a scalar computation. From a viewpoint of comparison between the OES and ISS methods, the ISS method is superior to the OES method because the ISS method shows a faster convergence and saves a computer memory without reducing a computation speed. In the vector computation, a table-look-up method to reduce computation time of an exponential function saves only 20% of a whole computation time. Both the coarse mesh rebalance method and the Aitken acceleration method are effective as acceleration methods for the characteristics method, a combination of them saves 70-80% of outer iterations compared with a free iteration.

2 (Records 1-2 displayed on this page)
  • 1