The Rigorous Two Step (R2S) Method¶

The Rigorous Two Step (R2S) method is used to determine the photon fluxes that result from neutron activation of materials for a given geometry. The first step of this method is neutron transport over some geometry in order to calculate flux distributions in space and energy. Using this neutron flux distribution and the isotopic composition of each part of the geometry, material activation calculations can be done in order to calculate a distribution of photon source strengths in space and energy. The second step of R2S is doing a photon transport calculation on these decay photons, in order to determine a photon flux distribution in space and energy. The “two step” part of the name specifies that neutron and photon transport steps are done separately, as opposed to the Direct One Step (D1S) method which performs both transport calculations simultaneously.

University of Wisconsin Rigorous Two Step Activation Work Flow (R2S-ACT)¶

R2S-ACT is composed of a series of scripts written in the Python programming language to facilitate R2S calculations in complex 3D geometries. The end product of R2S-ACT is a single Cartesian MOAB mesh file with each voxel tagged with:

1. neutron fluxes and relative errors for every neutron energy group, total neutron flux (neutrons/cm^2/s)
2. material fractions and relative errors for every material in the geometry
3. photon source strengths for every photon energy group (photons/cm^3/s)
4. total photon source strength over the entire geometry (photons/s)
5. photon flux and/or dose for every photon energy group and relative errors, total flux and/or dose (photons/cm^3/s, any dose units)

The backbone of R2S-ACT is the physics codes used for transport and activation. The Direct Accelerated Geometry Monte Carlo (DAG-MC) version of MCNP5 (Los Alamos National Laboratory), known as DAG-MCNP5 (UW-Madison) is used for both neutron and photon transport. This allows for geometry and materials information to be specified using CAD software, namely CubIt (Sandia National Laboratory). For the photon transport step, a custom compiled version of DAG-MCNP5, with a custom source.F90, is used. Analytic and Laplacian Adaptive Radioactivity Analysis (ALARA), developed at UW-Madison, is used for material activation.

Compiling DAG-MCNP5 with R2S-ACT source.F90¶

In MCNP5, fixed source information is typically specified with the SDEF card. However, the SDEF card is limited to 1000 distributions. Building DAG-MCNP5 against R2S-ACT source.F90 allow for an unlimited number of distributions. This is necessary because in the R2S-ACT work flow, every mesh point is a particle source and generally high resolution spacial and energy information is sought after. To use the R2S-ACT source.F90 version, no source cards (SDEF, SI, SP, etc.) appears in the input file. Instead, this version reads in a file (gammas) that contain all of the source information. This file is automatically generated by the R2S work flow. Technically this version is only necessary for the photon transport step (and not the neutron transport step), however if the MCNP input includes a source definition, the problem run like normal. Thus the same executable can be used for bother regular, and R2S-ACT photon transport calculations. Instructions for compiling DAG-MCNP5 with R2S-ACT source.F90 appear below.

1. Navigate to the trunk/Source/src folder of DAG-MCNP5.
2. Delete the file source.F90
3. Create a soft link (named source.F90) to the file source_gamma_refactor.F90, found in the folder r2s-act/mcnp_source/.
4. Build DAG-MCNP5 in the usual fashion.

Executive Summary¶

1. Create geometry, and prepare it for DAGMCNP and/or unstructured mesh.
2. Run neutron transport problem.
3. Setup the problem with r2s_setup.py, and make any changes needed to r2s.cfg and the ALARA snippet file.
4. Create ALARA materials library file.
5. Run r2s_step1.py.
6. Run ALARA.
7. Run r2s_step2.py.
8. Set up photon transport (e.g. tallies) and run modified MCNP/DAGMCNP with mesh-sampling source.F90.
9. Examine results; Use read_meshtal.py to tag photon mesh tallies to the mesh for visualization.

Full Summary¶

1. Create geometry. Using CubIt, create the geometry and specify materials by adding volumes to materials groups. Group names should be formatted like “mat_X_rhoY” where X is the material number and Y is either mass density (negative value) or atom density (positive volume). Instructions for doing this can be found in the DAG-MCNP5 user manual <http://svalinn.github.io/DAGMC/doc/usersguide/index.html>. Be sure to remember to imprint and merge all of the geometry. Once completed, export the geometry as a .sat file (Standard ACIS Text format) and when prompted specify an ACIS version of 1900 and the “export attributes” option. Alternatively, if the geometry already exists in the form of an MCNP input file, MCNP2CAD can be used to convert the geometry information from the MCNP input file into a .sat file. If the geometry contains small features, users may need to specify a smaller tolerance for merging surfaces (using the -t flag).

Once the .sat file exists, it can be converted to an .h5m file (binary format MOAB mesh file) using dagmc_preproc or mbconvert. This is not necessary, but it prevents DAG-MCNP5 from having to process the .sat file every time it is run. Using a .h5m file also allows for the use of of a DAG-MCNP5 version that is not built against CubIt. In either case, either DAG-MCNNP or dagmc_preproc/mbconvert creates a faceted representation of the geometry. Users can specify the maximum distance between the points in the geometry and the faceted representation on the geometry. This is known as the faceting tolerance. In dagmc_preproc, this is specified with the -f flag. In DAG-MCNP5 this can be specified on the command line by using ftol=faceting_tolerace (e.g. ftol=1E-4). In addtion, dagmc_preproc can also be used to specify a length tolerance using the -l flag. The length tolerance is the maximum length of a facet edge.

2. Create DAG-MCNP5 input file and run neutron transport calculation. Other than the geometry cards, the rest of the DAG-MCNP5 input file should be identical to that of a native MCNP input file. Make sure the material numbers in the input file match the numbers of the material groups in CubIt.

For neutron activation, the MCNP/DAG-MCNP5 input file must contain a type 4 FMESH tally over the geometry of interest. Both structured (native MCNP) and unstructured mesh tallies are supported.

If a native mesh tally is used, the tally results appear in a MCNP meshtal output file. Otherwise, with an unstructured mesh tally, results are tagged directly to the mesh.

The output from the type 4 mesh tally will have units of neutrons/cm^2/source particle. This output needs to be converted to flux, by multiplying by the total neutron source strength (referred to as the neutron normalization factor) which has units of source particles/second. The recommended way of doing this is to use and FM tally multiplier card to specify the neutron normalization factor on the FMESH4 tally, which will result in a meshtal file with fluxes in the correct units. If this is not done, normalization can be done when fluxes are tagged to mesh using the read_meshtal.py script.

3. Run r2s_setup.py. This script creates two configuration files called r2s.cfg and alara_snippet in whatever folder the script is run from. It is best to run all subsequent scripts out of this folder.

4. Modify r2s.cfg to suit the problem. The configuration file allows the user to specify important parameters and also the names of files used by and created by r2s_step1.py, ALARA, and r2s_step2.py. The file r2s.cfg is printed with default file names and parameters and also some instructive annotations. Typically there is no reason to change the file names aside from personal preference. In order to better understand the parameters specified, users should consult the “Step 1 Parameters” and “Step 2 Parameters” portion of this file. Of particular note, if an unstructured mesh is being used in the problem, the structuredmesh entry in r2s.cfg should be set to False.

5. Create ALARA materials library. The script mats2ALARA.py can be used to convert MCNP materials definitions to ALARA materials definitions. However, generally speaking activation definitions should be much more detailed than transport definitions, as minor impurities can dominate activations. ALARA material libraries rely on isotope libraries. A script to write both MCNP and ALARA definitions for R2S style problems is currently in development.

6. Modify ALARA snippet. The ALARA snippet file contains all of the information needed to run ALARA, apart from the geometry and materials information. It is appended to the ALARA geometry/materials file produced by r2s_step1.py to create the full ALARA input file. Most of the entries in the default snipppet produced by r2s_setup.py need not be changed. However, every problem will have a different irradiation schedule so special attention should be taken to change this from the default irradiation schedule. The keywords for ALARA input files are documented in the ALARA Manual<https://trac.cae.wisc.edu/trac/svalinn/wiki/ALARAInputFileSyntax>. The ALARA snippet file also specifies the isotope library, material library, and data libraries. These libaries, or links to them must be present in the folder that ALARA is run out of. The default activation and photon source libraries are both specified as “FENDL2” in the ALARA snippet. However, soft links to these files must be named FENDL2.bin and FENDL2.gam for the activation and photon source libraries, respectively.

7. Run r2s_step1.py. This script is a wrapper script that reads the meshtal, geometry, MCNP neutron input file, and alara_snippet file specified in r2s.cfg and outputs a structured mesh file tagged with neutron fluxes with relative errors, material fractions with relative errors, and a complete ALARA input file.

8. Run ALARA. ALARA reads in the geometry, material, and irradiation schedule information from the ALARA input file and outputs a file containing photon source strengths (phtn_src) for every voxel and energy group. In addition, ALARA can calculate isotopic inventories, decay heat, and more (see ALARA user manual), which is printed to standard output by default. This output can be piped to an output file. Assuming the ALARA snippet file was written correctly and appended the command for this step will be /path/to/ALARA/ alara_geom > output_file. Currently, ALARA does not print out phtn_src entries for entries of material “void.” This causes indexing problems in Step 2. To get around this, use SED or some text editor to replace “void” with “pseudo_void” (or something similar). Then make an entry in the ALARA material library for “pseudo_void” and assign the density to be equal to zero. This produces output that is mathematically correct.

9. Run r2s_step2.py. Like r2s_step1.py, this script is a wrapper for several other scripts. This script takes an ALARA phtn_src file, tags the information onto the structured mesh file from Step 1 and creates a gammas file used to specify the photon source distribution for the gamma transport step. It also modifies the MCNP neutron input file to create an MCNP photon input file. However this file may need additional user attention before running DAG-MCNP5, especially if the photon tally region is different from the neutron tally region. Flux to dose conversion factors should be added if dose results are desired.

11. Run DAG-MCNP5, complied with R2S-ACT source.F90. The custom compiled version of DAG-MCNP5 reads the gammas file (must be present within the same folder), and output as a meshtal file with photon fluxes and/or doses.

10. Run read_meshtal.py. Run this script with the -m flag in order tag photon fluxes and/or doses onto the mesh with the rest of the information on it. This script is run by r2s_step1.py, so more information about this script can be found in the “Scripts run by r2s_step1.py” section of this manual.

12. Visualize Results. To visualize results, stuctured mesh .h5m files must be first converted to .vtk viles. This can be done using the MOAB mbconvert tool (syntax: mbconvert <mesh_file.h5m> <mesh_file.vtk>). The best way of visualizing the results on the resulting .vtk file is using VisIT. Fluxes/doses are best viewed as “pseudo color” or “volume” plots. The geometry can be superimposed onto these plots. To do this, save the geometry as a .stl file in CubIt. Then open this file in VisIt and visualize it as a “mesh” plot. It is often useful to visualize results during intermediate steps of the work flow. For example it may be useful to visualize the neutron flux distribution and errors prior to continuing with the work flow.

r2s_setup.py¶

r2s_step1.py¶

Step 1 Parameters¶

These settings can be modified in a problem’s r2s.cfg file.

r2s_step2.py¶

Step 2 Parameters¶

These settings can be modified in a problem’s r2s.cfg file.

mats2ALARA.py¶

get_mesh_values.py¶

wwinp_to_h5m.py¶

h5m_to_wwinp.py¶

magic.py¶

tag_ebins.py¶

r2s_step2setup.py¶

Fin.