NEA-0438 BRIGITTE-KA.
BRIGITTE-KA, ENDF/B to KEDAK Data Conversion with Resonance Cross-Sections Tables Generator
NAME OR DESIGNATION OF PROGRAM, COMPUTER, NATURE OF PHYSICAL PROBLEM SOLVED, METHOD OF SOLUTION, RESTRICTIONS ON THE COMPLEXITY OF THE PROBLEM, TYPICAL RUNNING TIME, UNUSUAL FEATURES OF THE PROGRAM, RELATED AND AUXILIARY PROGRAMS, STATUS, REFERENCES, MACHINE REQUIREMENTS, LANGUAGE, OPERATING SYSTEM OR MONITOR UNDER WHICH PROGRAM IS EXECUTED, ANY OTHER PROGRAMMING OR OPERATING INFORMATION OR RESTRICTIONS, NAME AND ESTABLISHMENT OF AUTHOR, 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 |
|---|---|---|---|
| BRIGITTE-KA | NEA-0438/01 | Tested | 01-MAR-1976 |
Machines used:
| Package ID | Orig. computer | Test computer |
|---|---|---|
| NEA-0438/01 | IBM 370 series | IBM 370 series |
[ top ]
[ top ]
4. METHOD OF SOLUTION
The retrieval and processing codes of ENDF (4) have been used, but some have been modified. Point-wise cross sections are calculated from resonance parameters. In the resolved resonance region all resonances are taken into account for each energy point. In order to guarantee linear interpolation with an error less than eps in the resolved resonance region, an energy mesh constructed by using the UNICORN code (6) is refined by adding points, if a cross section value calculated from the resonance parameters differs appreciably from the value calculated by interpolation. The various ENDF interpolation rules are reduced to the linear-linear rule used by KEDAK. Pointwise cross sections are calculated from the given parameters (e.g. the angular distributions). Some data of ENDF/B MF=5 (energy distributions of secondary neutrons) are also converted.
The retrieval and processing codes of ENDF (4) have been used, but some have been modified. Point-wise cross sections are calculated from resonance parameters. In the resolved resonance region all resonances are taken into account for each energy point. In order to guarantee linear interpolation with an error less than eps in the resolved resonance region, an energy mesh constructed by using the UNICORN code (6) is refined by adding points, if a cross section value calculated from the resonance parameters differs appreciably from the value calculated by interpolation. The various ENDF interpolation rules are reduced to the linear-linear rule used by KEDAK. Pointwise cross sections are calculated from the given parameters (e.g. the angular distributions). Some data of ENDF/B MF=5 (energy distributions of secondary neutrons) are also converted.
[ top ]
5. RESTRICTIONS ON THE COMPLEXITY OF THE PROBLEM
Because variable dimensioning is used for nearly all arrays, there are only few restrictions. These are the following:
- One (natural) element may have up to 10 isotopes.
- Five different L-states (L=0,1,2,3,4) are allowed in the resolved Breit-Wigner resonance parameter set.
- Three different L-states and 5 different J-states for each L-state are allowed in the unresolved Breit-Wigner resonance parameter set.
- One hundred points are allowed as primary energy grid for energy distributions of secondary neutrons.
Because variable dimensioning is used for nearly all arrays, there are only few restrictions. These are the following:
- One (natural) element may have up to 10 isotopes.
- Five different L-states (L=0,1,2,3,4) are allowed in the resolved Breit-Wigner resonance parameter set.
- Three different L-states and 5 different J-states for each L-state are allowed in the unresolved Breit-Wigner resonance parameter set.
- One hundred points are allowed as primary energy grid for energy distributions of secondary neutrons.
[ top ]
6. TYPICAL RUNNING TIME
All times and storage requirements will be given for an IBM 370/168. (a) To convert chlorine which has no resonance parameters, takes about 30 sec CPU time and 10 to 20 minutes elapsed time in a 480k bytes main storage.
(b) To convert tantalum-181 which has 76 single level Breit-Wigner resonances, takes about 630 sec CPU time, 45 minutes elapsed time and 480k bytes main storage.
(c) To convert plutonium-240 which has 201 resolved multilevel Breit-Wigner resonances, takes between six and ten hours CPU time. This value is an estimated extrapolation of a half hour run. The main storage needed in this half hour computation has been 1200k bytes, and the elapsed time was about 90 minutes.
Converting Pu-240 using the single-level Breit-Wigner formalism took about 474 minutes CPU time, 1200k bytes main storage and about 113 minutes elapsed time.
Most of the computing time is spent in calculating the point-wise cross sections from the resonance parameters.
More than half of the materials on ENDF/B-4 can be converted by using 480k bytes or less of main storage, the rest by using up to 1500k bytes, which is the upper limit for normal jobs on the used installation.
All times and storage requirements will be given for an IBM 370/168. (a) To convert chlorine which has no resonance parameters, takes about 30 sec CPU time and 10 to 20 minutes elapsed time in a 480k bytes main storage.
(b) To convert tantalum-181 which has 76 single level Breit-Wigner resonances, takes about 630 sec CPU time, 45 minutes elapsed time and 480k bytes main storage.
(c) To convert plutonium-240 which has 201 resolved multilevel Breit-Wigner resonances, takes between six and ten hours CPU time. This value is an estimated extrapolation of a half hour run. The main storage needed in this half hour computation has been 1200k bytes, and the elapsed time was about 90 minutes.
Converting Pu-240 using the single-level Breit-Wigner formalism took about 474 minutes CPU time, 1200k bytes main storage and about 113 minutes elapsed time.
Most of the computing time is spent in calculating the point-wise cross sections from the resonance parameters.
More than half of the materials on ENDF/B-4 can be converted by using 480k bytes or less of main storage, the rest by using up to 1500k bytes, which is the upper limit for normal jobs on the used installation.
[ top ]
7. UNUSUAL FEATURES OF THE PROGRAM
This is the only program for the given problem. There are some differences of the new version of the program compared to the earlier versions. Variable dimensioning for almost all arrays with problem dependent lengths has been implemented. Many improvements and extensions, some of which are listed in other sections of this abstract, have been made. In the KEDAK library generated by this program, each material starts on a new record. The last word of the first record points to the first unused record in the library.
This is the only program for the given problem. There are some differences of the new version of the program compared to the earlier versions. Variable dimensioning for almost all arrays with problem dependent lengths has been implemented. Many improvements and extensions, some of which are listed in other sections of this abstract, have been made. In the KEDAK library generated by this program, each material starts on a new record. The last word of the first record points to the first unused record in the library.
[ top ]
8. RELATED AND AUXILIARY PROGRAMS
BRIGITTE-KA uses some routines which are all members of local libraries (private or public). These routines are:
- DEFI (7) and DINF (7) to use the IBM-FORTRAN IV DEFINE FILE statement in an input directed more dynamic way (ASSEMBLER).
- BUFREG (8) to estimate, when starting the program, the space needed for buffers from the given data control block specifications (ASSEMBLER).
- XTAREA (9) and FREESP (10) to make use of the main storage region used for data in a dynamical way. (ASSEMBLER) - CONVX (11) to convert EBCDIC numbers into internal values (ASSEMBLER).
- A8FORM (12) to print text in a large format (FORTRAN).
- DATUM (13) and ZEIT (14) to give date, day-time and CPU time (ASSEMBLER and FORTRAN).
If it is not possible to use the ASSEMBLER routines distributed with the package, one may either use own routines or replace them by FORTRAN routines with the loss of some or all the functions performed. E.g. XTAREA can be replaced by a routine which reserves an array of fixed length at compilation time.
BRIGITTE-KA uses some routines which are all members of local libraries (private or public). These routines are:
- DEFI (7) and DINF (7) to use the IBM-FORTRAN IV DEFINE FILE statement in an input directed more dynamic way (ASSEMBLER).
- BUFREG (8) to estimate, when starting the program, the space needed for buffers from the given data control block specifications (ASSEMBLER).
- XTAREA (9) and FREESP (10) to make use of the main storage region used for data in a dynamical way. (ASSEMBLER) - CONVX (11) to convert EBCDIC numbers into internal values (ASSEMBLER).
- A8FORM (12) to print text in a large format (FORTRAN).
- DATUM (13) and ZEIT (14) to give date, day-time and CPU time (ASSEMBLER and FORTRAN).
If it is not possible to use the ASSEMBLER routines distributed with the package, one may either use own routines or replace them by FORTRAN routines with the loss of some or all the functions performed. E.g. XTAREA can be replaced by a routine which reserves an array of fixed length at compilation time.
[ top ]
10. REFERENCES
- J.C. Schepers:
BRIGITTE. A Computer Program for Translating Neutron Cross Section Data from the ENDF/B Evaluated Nuclear Data File to the KEDAK Format
C.E.N. Mol. Work document presented at the Meeting on Conversion
of Formats and Related Problems,
Bologna (June 7-9, 1972).
- E. Stein, J.C. Schepers and P. Vandeplas:
BRIGITTE. A Computer Program for Translating Evaluated Nuclear Data from the ENDF Representation into the KEDAK Representation
KFK Report (in preparation).
- Data Formats and procedures for the ENDF Neutron Cross Section
Library
BNL 50274 (T-601). ENDF 102 vol.1 (October 1970).
- Description of the ENDF/B Processing Codes and Retrieval Subroutines
BNL 50300. ENDF 110 Revised (June 1971).
- B. Krieg:
Handling and Service Programs for the Karlsruhe Nuclear Data File KEDAK, Part I - Management and Retrieval Programs
KFK 1725 (June 1973).
- G. Arnecke and H. Bachmann:
DINF und DEFI. Programme zur Dynamisierung des DEFINE FILE Statements in IBM-FORTRAN IV
Internal Report.
- Stein:
BUFREG. Personal Communication
- W. Hoebel:
XTAREA/REXTAR. Eine Subroutine zur Vollstaendigen Dimensionierung von IBM/360-FORTRAN-Programmen
Internal report.
- Ch. Hinze:
FREESP(I). Eine Subroutine zur Bestimmung des noch freien
Kernspeichers fuer FORTRAN-benutzer an der IBM/360-65
Internal Report.
- K. Gogg:
CONVX. FORTRAN-Unterprogramm fuer die IBM/360 zur Umwandlung von in maschineninterner Darstellung vorliegenden Fest- und Gleitkommazahlen in alphanumerische Darstellung
Internal Report.
- O. Eggenberger:
A8FORM. Ausdruck von Texten in Grossformat
Internal Report.
- Ch. Hinze:
DATUM(DDAT,DZEIT). Datums- und Uhrzeitroutine fuer FORTRAN-Benutzer an der Rechenanlage IBM/360-65
Internal Report.
- Ch. Hinze:
ZEIT(Z). Zeitkontrollroutine fuer FORTRAN-Benutzer an der Rechenanlage IBM/360-65
Internal Report.
- IBM system/360 and system/370, FORTRAN IV language
file no. S360-25. Re-order no. GC28-6515-9. (March 31, 1973).
- J.C. Schepers:
BRIGITTE. A Computer Program for Translating Neutron Cross Section Data from the ENDF/B Evaluated Nuclear Data File to the KEDAK Format
C.E.N. Mol. Work document presented at the Meeting on Conversion
of Formats and Related Problems,
Bologna (June 7-9, 1972).
- E. Stein, J.C. Schepers and P. Vandeplas:
BRIGITTE. A Computer Program for Translating Evaluated Nuclear Data from the ENDF Representation into the KEDAK Representation
KFK Report (in preparation).
- Data Formats and procedures for the ENDF Neutron Cross Section
Library
BNL 50274 (T-601). ENDF 102 vol.1 (October 1970).
- Description of the ENDF/B Processing Codes and Retrieval Subroutines
BNL 50300. ENDF 110 Revised (June 1971).
- B. Krieg:
Handling and Service Programs for the Karlsruhe Nuclear Data File KEDAK, Part I - Management and Retrieval Programs
KFK 1725 (June 1973).
- G. Arnecke and H. Bachmann:
DINF und DEFI. Programme zur Dynamisierung des DEFINE FILE Statements in IBM-FORTRAN IV
Internal Report.
- Stein:
BUFREG. Personal Communication
- W. Hoebel:
XTAREA/REXTAR. Eine Subroutine zur Vollstaendigen Dimensionierung von IBM/360-FORTRAN-Programmen
Internal report.
- Ch. Hinze:
FREESP(I). Eine Subroutine zur Bestimmung des noch freien
Kernspeichers fuer FORTRAN-benutzer an der IBM/360-65
Internal Report.
- K. Gogg:
CONVX. FORTRAN-Unterprogramm fuer die IBM/360 zur Umwandlung von in maschineninterner Darstellung vorliegenden Fest- und Gleitkommazahlen in alphanumerische Darstellung
Internal Report.
- O. Eggenberger:
A8FORM. Ausdruck von Texten in Grossformat
Internal Report.
- Ch. Hinze:
DATUM(DDAT,DZEIT). Datums- und Uhrzeitroutine fuer FORTRAN-Benutzer an der Rechenanlage IBM/360-65
Internal Report.
- Ch. Hinze:
ZEIT(Z). Zeitkontrollroutine fuer FORTRAN-Benutzer an der Rechenanlage IBM/360-65
Internal Report.
- IBM system/360 and system/370, FORTRAN IV language
file no. S360-25. Re-order no. GC28-6515-9. (March 31, 1973).
NEA-0438/01, included references:
- I. Langner and R. Meyer:The KEDAK Program Compendium Part VII, Calcul - Calculation of
Composed Cross Sections
KFK 2387/VII (November 1977).
- E. Stein:
The KEDAK Program Compendium Part VI, Mechanized Transfer of
Nuclear Data from ENDF/B to KEDAK and vice versa
KfK 2387/VI (March 1978).
[ top ]
11. MACHINE REQUIREMENTS
BRIGITTE needs about 156k bytes of main storage for the program and the I/O buffers. The data main storage depends on the material to be translated. For materials with few energy points, a data main storage of 144k bytes is sufficient. The total main storage is then 144k bytes. But there are materials which require a total main storage of 1500k bytes.
The machine must have at least one direct accessible device (e.g. disk) on which several data sets can reside (including the KEDAK output data set).For the input ENDF/B data, which are normally read from tape, a tape unit is needed.
If the direct access I/O is simulated by software (see point 14) only sequential devices are required.
See also item 6 above.
BRIGITTE needs about 156k bytes of main storage for the program and the I/O buffers. The data main storage depends on the material to be translated. For materials with few energy points, a data main storage of 144k bytes is sufficient. The total main storage is then 144k bytes. But there are materials which require a total main storage of 1500k bytes.
The machine must have at least one direct accessible device (e.g. disk) on which several data sets can reside (including the KEDAK output data set).For the input ENDF/B data, which are normally read from tape, a tape unit is needed.
If the direct access I/O is simulated by software (see point 14) only sequential devices are required.
See also item 6 above.
[ top ]
[ top ]
[ top ]
[ top ]
NEA-0438/01
| File name | File description | Records |
|---|---|---|
| NEA0438_01.001 | INFORMATION | 6 |
| NEA0438_01.002 | SOURCE PROGRAM (F4) EBCDIC | 6618 |
| NEA0438_01.003 | AUXILIARY PROGRAM (FORTRAN) EBCDIC | 143 |
| NEA0438_01.004 | AUXILIARY PROGRAM (ASSEMBLER) EBCDIC | 1647 |
| NEA0438_01.005 | OVERLAY CARDS (USED AT CPL) EBCDIC | 80 |
| NEA0438_01.006 | OVERLAY CARDS (ORIGINAL) EBCDIC | 80 |
| NEA0438_01.007 | SAMPLE PROBLEM INPUT DATA EBCDIC | 4 |
| NEA0438_01.008 | JCL EBCDIC | 11 |
| NEA0438_01.009 | SAMPLE PROBLEM PRINTED OUTPUT | 7932 |
Keywords: ENDF/B, KEDAK, cross sections, data processing.