NAME OR DESIGNATION OF PROGRAM, COMPUTER, DESCRIPTION OF PROGRAM OR FUNCTION, METHODS, RESTRICTIONS ON THE COMPLEXITY OF THE PROBLEM, TYPICAL RUNNING TIME, UNUSUAL FEATURES, RELATED OR AUXILIARY PROGRAMS, STATUS, REFERENCES, HARDWARE REQUIREMENTS, LANGUAGE, SOFTWARE REQUIREMENTS, OTHER PROGRAMMING OR OPERATING INFORMATION OR RESTRICTIONS, NAME AND ESTABLISHMENT OF AUTHORS, MATERIAL, CATEGORIES

[ top ]

[ top ]

To submit a request, click below on the link of the version you wish to order.
Only liaison officers are authorised to submit online requests. Rules for requesters are
available here.

Program name | Package id | Status | Status date |
---|---|---|---|

SUSD3D | NEA-1628/03 | Tested | 13-MAR-2008 |

Machines used:

Package ID | Orig. computer | Test computer |
---|---|---|

NEA-1628/03 | Many Computers | Linux-based PC |

[ top ]

3. DESCRIPTION OF PROGRAM OR FUNCTION

SUSD3D calculates sensitivity coefficients and standard deviation in the calculated detector responses or design parameters of interest due to input cross sections and their uncertainties. One-, two- and three-dimensional transport problems can be studied. Several types of uncertainties can be considered, i.e. those due to: (1) neutron/gamma multigroup cross sections, (2) energy-dependent response functions, (3) secondary angular distribution (SAD) or secondary energy distribution (SED) uncertainties.

SUSD3D development was started from the SUSD [6] code. Besides several minor modifications and extensions SUSD3D differs from SUSD in particular that:

- Three-dimensional analysis is possible,

- Flux moment files are used to evaluate the sensitivity profiles, instead of angular flux files; substantially reducing in this was the computer space requirements. SUSD3D can use the flux moment files produced by the DORT, TORT [7], ONEDANT, TWODANT and THREEDANT [8] discrete ordinates codes. The method used in the SUSD code based on the angular flux files from the ANISN [9] and DOT-III codes was kept for comparison,

- Processing codes were updated to the ENDF-6 format,

- Processing of SAD covariance matrices was programmed,

- Complete SAD covariance matrices can be taken into account in SUSD3D to calculate the variance.

SUSD3D calculates sensitivity coefficients and standard deviation in the calculated detector responses or design parameters of interest due to input cross sections and their uncertainties. One-, two- and three-dimensional transport problems can be studied. Several types of uncertainties can be considered, i.e. those due to: (1) neutron/gamma multigroup cross sections, (2) energy-dependent response functions, (3) secondary angular distribution (SAD) or secondary energy distribution (SED) uncertainties.

SUSD3D development was started from the SUSD [6] code. Besides several minor modifications and extensions SUSD3D differs from SUSD in particular that:

- Three-dimensional analysis is possible,

- Flux moment files are used to evaluate the sensitivity profiles, instead of angular flux files; substantially reducing in this was the computer space requirements. SUSD3D can use the flux moment files produced by the DORT, TORT [7], ONEDANT, TWODANT and THREEDANT [8] discrete ordinates codes. The method used in the SUSD code based on the angular flux files from the ANISN [9] and DOT-III codes was kept for comparison,

- Processing codes were updated to the ENDF-6 format,

- Processing of SAD covariance matrices was programmed,

- Complete SAD covariance matrices can be taken into account in SUSD3D to calculate the variance.

NEA-1628/03

This version differs from the previous one in the following points:------------------------------------------------------------------

Modifications are relevant for the sensitivity calculations of the critical systems and include:

- Correction of the sensitivity calculation for prompt fission and number of delayed neutrons per fission (MT=18 and MT=455).

- An option allows the re-normalisation of the prompt fission spectra covariance matrices to be applied via the "normalisation" of the sensitivity profiles. This option is useful in case if the fission spectra covariances (MF=35) used do not comply with the ENDF-6 Format Manual rules.

- For the criticality calculations the normalisation can be calculated by the code SUSD3D internally. Parameter NORM should be set to 0 in this case. Total number of neutrons per fission (MT=452) sensitivities for all the fissile materials must be requested in the SUSD3D OVERLAY-2 input deck in order to allow the correct normalisation.

[ top ]

4. METHODS

First-order perturbation theory is used to obtain sensitivity coefficients. They are derived from the direct and adjoint flux moments (or angular fluxes) calculated by the discrete ordinates codes listed above. The sensitivity profiles are folded with the cross section covariance matrices to determine the variance and standard deviation in an integral response of interest.

First-order perturbation theory is used to obtain sensitivity coefficients. They are derived from the direct and adjoint flux moments (or angular fluxes) calculated by the discrete ordinates codes listed above. The sensitivity profiles are folded with the cross section covariance matrices to determine the variance and standard deviation in an integral response of interest.

[ top ]

5. RESTRICTIONS ON THE COMPLEXITY OF THE PROBLEM

Variable dimensioning is used providing flexibility to adjust the storage requirements. Core storage is reserved for a particular dimensional array only during the time the corresponding data are needed in the calculation, afterwards the array is released for other data.

Variable dimensioning is used providing flexibility to adjust the storage requirements. Core storage is reserved for a particular dimensional array only during the time the corresponding data are needed in the calculation, afterwards the array is released for other data.

[ top ]

6. TYPICAL RUNNING TIME

Highly problem dependent, running time is affected by the parameters like the number of energy groups, number of dimensions (1, 2, 3), number of spatial intervals and PN approximation order used in the discrete ordinates transport calculations. The most complex case studied (VENUS-3 benchmark analysis [4] based on TORT 3D calculation using P-3/S-8 and 51/52/22 X/Y/Z intervals) took 2h40' on a PC Pentium.

Highly problem dependent, running time is affected by the parameters like the number of energy groups, number of dimensions (1, 2, 3), number of spatial intervals and PN approximation order used in the discrete ordinates transport calculations. The most complex case studied (VENUS-3 benchmark analysis [4] based on TORT 3D calculation using P-3/S-8 and 51/52/22 X/Y/Z intervals) took 2h40' on a PC Pentium.

[ top ]

7. UNUSUAL FEATURES

Although the sensitivities are calculated from the flux moment files, much smaller in size than the corresponding angular flux files, these files can still be voluminous for large 3D problems (see example in Section 11).

To simplify input preparation in 2D and in particular in 3D calculations the geometry description is read directly from the files produced by the discrete ordinates codes (see Section 8).

Modular structure of the code permits easy restart and supplementary calculations.

Although the sensitivities are calculated from the flux moment files, much smaller in size than the corresponding angular flux files, these files can still be voluminous for large 3D problems (see example in Section 11).

To simplify input preparation in 2D and in particular in 3D calculations the geometry description is read directly from the files produced by the discrete ordinates codes (see Section 8).

Modular structure of the code permits easy restart and supplementary calculations.

[ top ]

8. RELATED OR AUXILIARY PROGRAMS

SUSD3D is loosely based on the SUSD code [6].

SUSD3D is coupled to several discrete-ordinates codes via binary interface files. For 1D analysis angular flux files from ANISN or flux moment files from ONEDANT can be used, for 2D analysis angular flux files from DOT-III or flux moment files produced by DORT or TWODANT, and for 3D analysis flux moment files produced by the TORT or THREEDANT codes. In some of these codes minor modifications are required. Variable dimensions used in the TORT-DORT system are supported. In 3D analysis the geometry and material composition is taken directly from the TORT produced VARSCL binary file, reducing in this way the user's input to SUSD3D.

Multigroup cross-section sets are read in the GENDF format of the NJOY/GROUPR [10] code system, and the covariance data are expected in the COVFIL format of NJOY/ERRORR or the COVERX format of PUFF-2.

NJOY-SUSD3D (SUNJOY): The following cross section processing modules to be added to the NJOY-94 code system are included in the package:

ERR34: an extension of the ERRORR module of the NJOY code system for the File-34 processing. It is used to prepare multigroup SAD cross sections covariance matrices.

GROUPS: An additional code module for the preparation of partial cross sections for SAD sensitivity analysis. Updated version of the same code from SUSD, extended to the ENDF-6 format.

SEADR: An additional code module to prepare group covariance matrices for SAD/SED uncertainty analysis. As above.

Not included:

Auxiliary codes: NJOY, ERRORJ, BOT3P.

The ZZ-VITAMIN-J/COVA [11] cross section covariance matrix library can be used as an alternative to the NJOY code system. The package includes the ANGELO code to produce the covariance data in the required energy structure in the COVFIL format.

SUSD3D is loosely based on the SUSD code [6].

SUSD3D is coupled to several discrete-ordinates codes via binary interface files. For 1D analysis angular flux files from ANISN or flux moment files from ONEDANT can be used, for 2D analysis angular flux files from DOT-III or flux moment files produced by DORT or TWODANT, and for 3D analysis flux moment files produced by the TORT or THREEDANT codes. In some of these codes minor modifications are required. Variable dimensions used in the TORT-DORT system are supported. In 3D analysis the geometry and material composition is taken directly from the TORT produced VARSCL binary file, reducing in this way the user's input to SUSD3D.

Multigroup cross-section sets are read in the GENDF format of the NJOY/GROUPR [10] code system, and the covariance data are expected in the COVFIL format of NJOY/ERRORR or the COVERX format of PUFF-2.

NJOY-SUSD3D (SUNJOY): The following cross section processing modules to be added to the NJOY-94 code system are included in the package:

ERR34: an extension of the ERRORR module of the NJOY code system for the File-34 processing. It is used to prepare multigroup SAD cross sections covariance matrices.

GROUPS: An additional code module for the preparation of partial cross sections for SAD sensitivity analysis. Updated version of the same code from SUSD, extended to the ENDF-6 format.

SEADR: An additional code module to prepare group covariance matrices for SAD/SED uncertainty analysis. As above.

Not included:

Auxiliary codes: NJOY, ERRORJ, BOT3P.

The ZZ-VITAMIN-J/COVA [11] cross section covariance matrix library can be used as an alternative to the NJOY code system. The package includes the ANGELO code to produce the covariance data in the required energy structure in the COVFIL format.

[ top ]

10. REFERENCES

- I. Kodeli, L. Petrizzi, P. Batistoni:

Transport, Sensitivity and Uncertainty Analysis of FNG 14 MeV

Neutron Bulk Shield Experiment, ICRS 9 conference, Tsukuba,

Ibaraki, Japan (October 17-22, 1999)

- K. Furuta, Y. Oka, S. Kondo, SUSD: A Computer Code for

Cross-Section Sensitivity and Uncertainty Analysis Including

Secondary Neutron Energy and Angular Distributions, Draft of

UTNL-R0185 (English Translation, August 1986)

- W. A. Rhoades, R. L. Childs, TORT-DORT:

Two- and Three-Dimensional Discrete Ordinates Transport,

Version 2.7.3, RSIC-CCC, ORNL RSIC, Oak Ridge, TN (1993)

- R. E. Alcouffe et al.:

DANTSYS 3.0 - A Diffusion-Accelerated, Neutral-Particle

Transport Code System, LA-12969-M, LANL (1995)

- W. W. Engle:

A User's Manual for ANISN, A One-Dimensional Discrete Ordinates

Transport Code with Anisotropic Scattering, K-1693, Union Carbide

Corporation (1967)

- R. E. MacFarlane, D. W. Muir:

The NJOY Nuclear Data Processing System, Manual LA-12740-M (1994).

- I. Kodeli, E. Sartori:

ZZ-VITAMIN-J/COVA - Covariance Data Library, OECD/NEA-DB,

NEA 1264/03 package (1990).

- I. Kodeli:

Multidimensional Deterministic Nuclear Data Sensitivity

and Uncertainty Code System: Method and Application

Nucl. Sci. and Eng. 138, pp45-66 (2001)

- I. Kodeli, L. Petrizzi, P. Batistoni:

Transport, Sensitivity and Uncertainty Analysis of FNG 14 MeV

Neutron Bulk Shield Experiment, ICRS 9 conference, Tsukuba,

Ibaraki, Japan (October 17-22, 1999)

- K. Furuta, Y. Oka, S. Kondo, SUSD: A Computer Code for

Cross-Section Sensitivity and Uncertainty Analysis Including

Secondary Neutron Energy and Angular Distributions, Draft of

UTNL-R0185 (English Translation, August 1986)

- W. A. Rhoades, R. L. Childs, TORT-DORT:

Two- and Three-Dimensional Discrete Ordinates Transport,

Version 2.7.3, RSIC-CCC, ORNL RSIC, Oak Ridge, TN (1993)

- R. E. Alcouffe et al.:

DANTSYS 3.0 - A Diffusion-Accelerated, Neutral-Particle

Transport Code System, LA-12969-M, LANL (1995)

- W. W. Engle:

A User's Manual for ANISN, A One-Dimensional Discrete Ordinates

Transport Code with Anisotropic Scattering, K-1693, Union Carbide

Corporation (1967)

- R. E. MacFarlane, D. W. Muir:

The NJOY Nuclear Data Processing System, Manual LA-12740-M (1994).

- I. Kodeli, E. Sartori:

ZZ-VITAMIN-J/COVA - Covariance Data Library, OECD/NEA-DB,

NEA 1264/03 package (1990).

- I. Kodeli:

Multidimensional Deterministic Nuclear Data Sensitivity

and Uncertainty Code System: Method and Application

Nucl. Sci. and Eng. 138, pp45-66 (2001)

NEA-1628/03, included references:

- I. Kodeli:SUSD3D - 2008

A Multi-Dimensional, Discrete Ordinates based Cross Section Sensitivity and

Uncertainty Code (January 2008)

- B. Zefran:

Upgrading the Sensitivity/Uncertainty Analysis code system SUSD-3D concerning

memory and data management and language level (from Fortran-77 to Fortran-95)

(January 2005)

[ top ]

11. HARDWARE REQUIREMENTS

Core requirements depend on problem complexity. Up to eight tape units are used in addition to the standard input and output devices. The size of the flux moment files, produced by the TORT code for the above mentioned VENUS-3 benchmark analysis was 234 MB (two files required). For comparison, using the angular flux approach the corresponding file size would be 1407 MB.

Core requirements depend on problem complexity. Up to eight tape units are used in addition to the standard input and output devices. The size of the flux moment files, produced by the TORT code for the above mentioned VENUS-3 benchmark analysis was 234 MB (two files required). For comparison, using the angular flux approach the corresponding file size would be 1407 MB.

[ top ]

[ top ]

[ top ]

[ top ]

NEA-1628/03

SUSD3D.INF Information filesusd3d-doc.pdf SUSD3D documentation and input instructions

fortran95.pdf report on FORTAN77 to FORTRAN95 upgrading

susd3d\susd3d.f90 SUSD3D source code

susd3d\compile.sh shell script to compile SUSD3D on LINUX/UNIX

sunjoy\Makefile procedure to compile NJOY-SUSD3D (SUNJOY) on LINUX

sunjoy\compile-unx.sh

sunjoy\sunjoy_linux.f NJOY-SUSD3D (SUNJOY) main source code (for LINUX)

sunjoy\sunjoy_sun.f NJOY-SUSD3D (SUNJOY) main source code (for UNIX)

sunjoy\groups.f GROUPS source code

sunjoy\errorr34.f ERR34 source code

sunjoy\seadr.f SEADR source code

sunjoy\dummy.f dummy subroutines for SUNJOY

sunjoy\sunjoy_vms.for NJOY-SUSD3D (SUNJOY) main source code (for VAX)

Test1\ data for sample problem 1

Test2\ data for sample problem 2

[ top ]

- D. Depletion, Fuel Management, Cost Analysis, and Power Plant Economics
- O. Experimental Data Processing

Keywords: angular distribution, covariance matrices, cross sections, data uncertainties, neutron transport theory, radiation detectors, response functions, sensitivity analysis.