RSIC CODE PACKAGE PSR-534

 

1.  NAME AND TITLE

PUFF‑IV, Version 6.1.0: Code System to Generate Multigroup Covariance Matrices from ENDF/B-VI Uncertain­ty Files.

 

2.   CONTRIBUTOR

Oak Ridge National Laboratory, Oak Ridge, Tennessee.

 

3.   CODING LANGUAGE AND COMPUTER

Fortran 90; Personal Computers and Workstations (P00534MNYCP01).

 

4.   NATURE OF PROBLEM SOLVED

The PUFF‑IV code system processes ENDF/B-VI formatted nuclear cross section covariance data into multigroup covariance matrices. PUFF‑IV is the newest release in this series of codes used to process ENDF uncertainty information and to generate the desired multi-group correlation matrix for the evaluation of interest. PUFF‑IV enhances support for resonance parameter covariance formats described in the ENDF standard and now handles almost all resonance parameter covariance information in the resolved region, with the exception of the long range covariance sub-subsections. It is written in Fortran 90 and allows for a more modular design, thus facilitating future upgrades. This version supersedes the previous release PUFF IV, Version 1.0.2, and includes several corrections and enhancements.

l      Various bug fixes concerning calculation of derivatives and group averaging of derivatives

l      Improved processing of lrf=7 covariance matrices

l      Support for BLAS to decrease computation time in the resolved resonance range

l      Support for user defined flux functions

l      Module covcomp has additional functionality for comparing and adding coverx formatted files

l      Module coverr allows to convert COVERX files to NJOY covariances tapes

 

PUFF-IV is normally used in conjunction with an AMPX master library containing group averaged cross section data. Two utility modules are included in this package to facilitate the data interface. The module SMILER allows one to use NJOY generated GENDF files containing group averaged cross section data in conjunction with PUFF-IV.  The module COVCOMP allows one to compare two files written in COVERX format.

 

5.   METHOD OF SOLUTION

Cross section and flux values on a "super energy grid," consisting of the union of the required energy group structure and the energy data points in the ENDF/B-V file, are interpolated from the input cross sections and fluxes.  Covariance matrices are calculated for this grid and then collapsed to the required group structure.

 

6.   RESTRICTIONS OR LIMITATIONS

PUFF-IV cannot process covariance information for energy and angular distributions of secondary particles. PUFF-IV does not process covariance information in Files 34 and 35; nor does it process covariance information in File 40. These new formats will be addressed in a future version of PUFF.

 

7.   TYPICAL RUNNING TIME

Running times vary.

 

8.   COMPUTER HARDWARE REQUIREMENTS

PUFF‑IV has been tested on workstations and personal computers under Unix, Linux and Windows+Cygwin operating systems.

 

9.   COMPUTER SOFTWARE REQUIREMENTS

Fortran 90 and C compilers are required. The GNU Autoconf procedure is used to generate Makefiles for creating executables. PUFF‑IV has been tested on the following systems. Older versions of g95 may not work.

Personal Computer under WindowsXP w/Cygwin and g95 - Stable Version 0.91 or up

AMD Opteron running RedHat Enterprise Linux 4 with G95 - Stable Version 0.91 or up

AMD Opteron running RedHat Enterprise Linux 4 with Intel compiler.

AMD Opteron running RedHat Enterprise Linux 4 with Portland Group, Inc.

IBM AIX Powerpc with XLF90 compiler

Alpha Compaq Unix Tru64 V5.1A with HP Fortran V5.5A-3548-48D88

 

10.       REFERENCE

D. Wiarda and M.E. Dunn, “PUFF-IV: A Code for Processing ENDF Uncertainty Data into Multigroup Covariance Matrices,” ORNL/TM-2006/147/R1 (October 2006, Revised 2008).

 

11. CONTENTS OF CODE PACKAGE

The package is transmitted on a CD that includes the referenced document and a GNU zipped tar file which contains the source code, sample problem input and output, and cross sections for running sample problems.

 

12. DATE OF ABSTRACT

August 2006, revised November 2008.

 

KEYWORDS:      COVARIANCE DATA PROCESSING; ENDF/B; MULTIGROUP CROSS SECTION PROCESSING; NEUTRON CROSS SECTION PROCESS­ING