Abstract
A set of library files can be considered to be a validated application library when the following requirements are met: the library processing has been verified; the library files have been tested on internationally accepted benchmarks (validation); every aspect in their creation has been documented The verification process ensures that no problems were encountered during processing (due to faulty data, bad formatting, . . . ) and that corrective actions are taken or a warning is issued if a problem was found. It is important to note that the verification process starts with the selection of the NJOY processing path and all relevant parameters, even before the first library fie is produced. The third requirement (document everything) is closely linked to this part of the verification process. Because the library is produced for use with MCNP(X), it must be validated for use with MCNP(X) to ensure that it gives accurate and reliable results. This can be done by testing the library against internationally accepted benchmarks. And again, this part must be well documented. Documenting every aspect in the creation of the library is important for future reference. A data preprocessing code (in our case NJOY) should never be considered as
More>>
Citation Formats
Haeck, W, and Verboomen, B.
Verification and validation of a multi-temperature JEFF 3.1 library for MCNP(X) - JEF/DOC-1099.
Belgium: N. p.,
2005.
Web.
Haeck, W, & Verboomen, B.
Verification and validation of a multi-temperature JEFF 3.1 library for MCNP(X) - JEF/DOC-1099.
Belgium.
Haeck, W, and Verboomen, B.
2005.
"Verification and validation of a multi-temperature JEFF 3.1 library for MCNP(X) - JEF/DOC-1099."
Belgium.
@misc{etde_20902513,
title = {Verification and validation of a multi-temperature JEFF 3.1 library for MCNP(X) - JEF/DOC-1099}
author = {Haeck, W, and Verboomen, B}
abstractNote = {A set of library files can be considered to be a validated application library when the following requirements are met: the library processing has been verified; the library files have been tested on internationally accepted benchmarks (validation); every aspect in their creation has been documented The verification process ensures that no problems were encountered during processing (due to faulty data, bad formatting, . . . ) and that corrective actions are taken or a warning is issued if a problem was found. It is important to note that the verification process starts with the selection of the NJOY processing path and all relevant parameters, even before the first library fie is produced. The third requirement (document everything) is closely linked to this part of the verification process. Because the library is produced for use with MCNP(X), it must be validated for use with MCNP(X) to ensure that it gives accurate and reliable results. This can be done by testing the library against internationally accepted benchmarks. And again, this part must be well documented. Documenting every aspect in the creation of the library is important for future reference. A data preprocessing code (in our case NJOY) should never be considered as a black box. With the documentation, a user should be capable of producing exactly the same library and understand every aspect of its creation. Before even starting with the production of a library, we should examine what has been done in the past in the field of library preparation and verification. This is why we started by looking into the documentation of older libraries to see what NJOY processing path they had chosen, what values they used for important parameters such as the reconstruction tolerance, . . . Other important information is the type of verification tests that were performed and what types of benchmarks were used to validate the library. And during this process the most important question is why. We should learn from all that experience and decide upon our own approach. At this point we can already draw a st important conclusion: documentation is of key importance. This has resulted in the creation of ALEPH-LIB (a multi-temperature library for standard use by MCNP(X)) and ALEPH-DLG (Data Library Generator). The temperatures included in the ALEPH-LIB library are 300, 600, 900, 1200, 1500 and 1800 K. Library files were produced using the JEF 2.2, JEFF 3.0, JEFF 3.1, JENDL 3.3 and ENDF/B-VI.8 nuclear data evaluations. This will be extended with ENDF/B-VII when it becomes available. ALEPH-DLG is an auxiliary computer code to ALEPH, the Monte Carlo burn-up interface code under development at SCK CEN in collaboration with Ghent university. This code automates the entire process of generating library files with NJOY and takes care of the st requirement of a validated application library: verify the processing. It produces tailor made NJOY input files using data from the original ENDF file (such as initial temperature, the fact if the nuclide is fissile or if it has unresolved resonances, . . . ) When the library files have been generated, ALEPH-DLG will also process the output from NJOY. It will extract all messages and warnings from NJOY and print out a short explanation of the message in question, it will test the unresolved resonance probability tables (to see if there are any negative cross sections, . . . ), . . . If ALEPH-DLG finds anything out of the ordinary, it will warn the user or perform corrective actions. In what follows we will briefly explain the verification process as it is implemented into ALEPH-DLG and give a summary of the results of this verification on the JEFF 3.1 files included in ALEPH-LIB. We will also present some results of the validation effort that we are performing on ALEPH-LIB using the Lawrence Livermore pulsed sphere experiments. These pulsed spheres can be used to validate nuclear data at high energy (above 2 MeV), mainly threshold reactions, inelastic scattering, . . . For ALEPH-LIB we are also performing a series of criticality benchmarks, but these will not be reported here. The validation is therefore only partial: the pulsed spheres tell us nothing about the data below 2 MeV.}
place = {Belgium}
year = {2005}
month = {Nov}
}
title = {Verification and validation of a multi-temperature JEFF 3.1 library for MCNP(X) - JEF/DOC-1099}
author = {Haeck, W, and Verboomen, B}
abstractNote = {A set of library files can be considered to be a validated application library when the following requirements are met: the library processing has been verified; the library files have been tested on internationally accepted benchmarks (validation); every aspect in their creation has been documented The verification process ensures that no problems were encountered during processing (due to faulty data, bad formatting, . . . ) and that corrective actions are taken or a warning is issued if a problem was found. It is important to note that the verification process starts with the selection of the NJOY processing path and all relevant parameters, even before the first library fie is produced. The third requirement (document everything) is closely linked to this part of the verification process. Because the library is produced for use with MCNP(X), it must be validated for use with MCNP(X) to ensure that it gives accurate and reliable results. This can be done by testing the library against internationally accepted benchmarks. And again, this part must be well documented. Documenting every aspect in the creation of the library is important for future reference. A data preprocessing code (in our case NJOY) should never be considered as a black box. With the documentation, a user should be capable of producing exactly the same library and understand every aspect of its creation. Before even starting with the production of a library, we should examine what has been done in the past in the field of library preparation and verification. This is why we started by looking into the documentation of older libraries to see what NJOY processing path they had chosen, what values they used for important parameters such as the reconstruction tolerance, . . . Other important information is the type of verification tests that were performed and what types of benchmarks were used to validate the library. And during this process the most important question is why. We should learn from all that experience and decide upon our own approach. At this point we can already draw a st important conclusion: documentation is of key importance. This has resulted in the creation of ALEPH-LIB (a multi-temperature library for standard use by MCNP(X)) and ALEPH-DLG (Data Library Generator). The temperatures included in the ALEPH-LIB library are 300, 600, 900, 1200, 1500 and 1800 K. Library files were produced using the JEF 2.2, JEFF 3.0, JEFF 3.1, JENDL 3.3 and ENDF/B-VI.8 nuclear data evaluations. This will be extended with ENDF/B-VII when it becomes available. ALEPH-DLG is an auxiliary computer code to ALEPH, the Monte Carlo burn-up interface code under development at SCK CEN in collaboration with Ghent university. This code automates the entire process of generating library files with NJOY and takes care of the st requirement of a validated application library: verify the processing. It produces tailor made NJOY input files using data from the original ENDF file (such as initial temperature, the fact if the nuclide is fissile or if it has unresolved resonances, . . . ) When the library files have been generated, ALEPH-DLG will also process the output from NJOY. It will extract all messages and warnings from NJOY and print out a short explanation of the message in question, it will test the unresolved resonance probability tables (to see if there are any negative cross sections, . . . ), . . . If ALEPH-DLG finds anything out of the ordinary, it will warn the user or perform corrective actions. In what follows we will briefly explain the verification process as it is implemented into ALEPH-DLG and give a summary of the results of this verification on the JEFF 3.1 files included in ALEPH-LIB. We will also present some results of the validation effort that we are performing on ALEPH-LIB using the Lawrence Livermore pulsed sphere experiments. These pulsed spheres can be used to validate nuclear data at high energy (above 2 MeV), mainly threshold reactions, inelastic scattering, . . . For ALEPH-LIB we are also performing a series of criticality benchmarks, but these will not be reported here. The validation is therefore only partial: the pulsed spheres tell us nothing about the data below 2 MeV.}
place = {Belgium}
year = {2005}
month = {Nov}
}