XSPROC: The Material and Cross Section Processing Module for SCALE

M. L. Williams, L. M. Petrie, R. A. Lefebvre, K. T. Clarno, J. P. Lefebvre, U. Merturyek, D. Wiarda, and B. T. Rearden

ABSTRACT

The modern material and cross section processing module of SCALE (XSProc) was developed for the 6.2 release to prepare data for continuous-energy and multigroup calculations. XSProc expands material input from Standard Composition Library definitions into atom number densities and, for multigroup calculations, performs cross section resonance self-shielding, energy group collapse, and spatial homogenization. XSProc implements capabilities for problem-dependent temperature interpolation, calculation of Dancoff factors, resonance self-shielding using Bondarenko factors with full-range intermediate resonance treatment, as well as use of continuous energy resonance self-shielding in the resolved resonance region. XSProc integrates and enhances the capabilities previously implemented independently in BONAMI, CENTRM, PMC, WORKER, ICE, and XSDRNPM, along with some additional capabilities that were provided by MIPLIB and SCALELIB. The use of the modern XSProc sequence instead of the legacy codes of previous versions of SCALE generally results in the preparation of cross sections in less time, with substantial speedups for more I/O bound problems. Additionally, the memory requirements of XSProc are improved by generating only the data needed for a particular calculation instead of generating a general-purpose library that contains substantial amounts of data that are not needed for a particular calculation.

ACKNOWLEDGMENTS

XSProc has evolved from the concept of a Material Information Processor library (MIPLIB) that used alphanumeric material specifications, which was initially proposed and developed by R. M. Westfall. J. R. Knight and J. A. Bucholz expanded and refined MIPLIB in early SCALE releases. Through SCALE 6.1, many enhancements were made by S. Goluoglu, D. F. Hollenbach, N. F. Landers, J. A. Bucholz, C. F. Weber, and C. M. Hopper, with L. M. Petrie taking the lead responsibility. With the SCALE modernization initiative beginning in SCALE 6.2, MIPLIB is no longer part of the XSProc analysis, but the original concepts and input formatting were preserved in the new implementation. The authors wish to thank Dan Ilas for helping convert the original MIPLIB documentation and Sheila Walker for editing and formatting this document. Special thanks to Don Mueller for his detailed review and checking of the document.

Introduction

Self-shielding of multigroup cross sections is required in SCALE sequences for criticality safety, reactor physics, radiation shielding, and sensitivity analysis. In all previous versions of SCALE, resonance self-shielding calculations were done by executing a series of stand-alone executable codes, each dedicated to a specific aspect of the self-shielding operations. Each sequence had its own unique internal coding to launch the executable codes. Multigroup (MG) and continuous-energy (CE) cross sections and other data were passed between the individual executable codes by external I/O, which could require a substantial amount of clock time. In the modern version of SCALE, all self-shielding operations are consolidated into a single driver module named XSProc, and the stand-alone executable codes have been transformed into callable “computational modules” [rearden_modernization_2015]. The functions of XSProc are to (a) read input data, (b) generate in-memory data structures (objects) containing problem-definition information (compositions, cell geometries, computation options), as well as self-shielding information (MG and CE cross sections and fluxes), and (c) execute appropriate computational modules for the requested self-shielding option. Calculated results produced by one module may be stored in the internal data objects and passed to other modules through application program interfaces (APIs). At the completion of XSProc the self-shielded MG cross sections on the data objects can be passed along to transport solvers for continued execution of the control sequence or can be written to an external AMPX library file.

In the future, XSProc will be extended to parallel computations in which self-shielding calculations are done simultaneously for multiple types of unit cells. At the present time, however, XSProc is limited to serial computations; but even in serial mode it typically requires less time than older versions of SCALE to process shielded cross sections, and significant speedups have been observed for heavily I/O bound problems. Integrating the self-shielding capabilities into a single module has a number of additional beneﬁts as well, including maintainability, extensibility, and the ability to easily replace an entire computational module with a future implementation containing new features. Additionally, the size of the problem-dependent MG library generated by XSProc may be greatly reduced compared to previous versions of SCALE because macroscopic cross sections are stored rather than a general-purpose library of microscopic data.

Techniques

XSProc integrates and enhances the capabilities previously implemented independently in BONAMI, CENTRM, PMC, WORKER, ICE, and XSDRNPM, as well as other capabilities formerly provided by MIPLIB and SCALELIB. It provides capabilities for problem-dependent temperature interpolation of both CE and MG nuclear data, calculation of Dancoff factors, and resonance self-shielding of MG cross sections using several available options. XSProc produces shielded microscopic data for each nuclide or macroscopic data for each material. Additionally, a flux-weighting spectrum can be applied to collapse cross sections to a coarser group structure and/or to integrate over volumes for homogenized cross sections. The flux-weighting spectrum can be input by the user or calculated using one-dimensional (1-D) coupled neutron/gamma transport model. These operations are performed by the sequences CSAS-MG, CSAS1, CSASI, and T-XSEC described in XSProc input data.

Overview of XSProc procedures

XSProc reads the COMPOSITION and CELL DATA blocks of the SCALE input, which are described in the following sections. After reading the user input data, XSProc loads the specified MG library to be self-shielded and, depending on the selected self-shielding method, additional CE data files for nuclides appearing in the problem specification. Finally XSProc performs MG self-shielding calculations for all compositions by calling APIs to computational modules such as BONAMI (BONdarenko AMPX Interpolator), CRAWDAD (Code to Read And Write DAta for Discretized solution), CENTRM (Continuous ENergy TRansport Module), PMC (Produce Multigroup Cross sections), CHOPS (Compute HOmogenized Pointwise Stuff), CAJUN (CE AJAX UNiter), WAX (Working AJAX), XSDRNPM (XSection Development for Reactor Nucleonics with Petrie Modifications), and/or MIXMACRO to provide a problem-dependent cross section library. Many computational modules have been modernized compared to earlier executable codes distributed in previous versions of SCALE.

Like earlier versions of SCALE, XSProc provides several options for self-shielding an input MG library [Wil11]. The first, based on the Bondarenko method [IlichB64], uses the computational module BONAMI. BONAMI is always used to compute self-shielded cross sections for all energy groups. If parm=bonami is specified, the shielded cross sections provided by BONAMI are the final values output from XSProc. However the Bondarenko method has several limitations, especially in the resolved resonance range. Therefore XSProc provides another self-shielding method, with several computation options, which often produces more accurate MG data in the resolved resonance and thermal energy ranges. If parm=centrm or parm=2region is specified on the sequence line, XSProc calls APIs for the modules CRAWDAD, CENTRM, and PMC to compute CE flux spectra for processing problem-specific, self-shielded cross sections “on the fly [WA95]. CENTRM performs MG transport calculations in the fast and lower energy ranges, coupled to pointwise (PW) transport calculations that use CE cross sections in the resonance range. PMC uses the PW flux spectra from CENTRM to compute MG values, which replace the previous values obtained from BONAMI over the specified range of the CE calculation. The original BONAMI shielded cross sections are retained for all other groups.

The CENTRM/PMC approach is the default for criticality and lattice physics calculations, while the BONAMI-only method is default for radiation shielding calculations. The end results of an XSProc calculation are self-shielded macroscopic and/or microscopic MG cross sections stored in memory for subsequent transport calculations; or alternatively a shielded MG AMPX library can be written to an external file and saved for future use.

Standard composition material processing

A primary function of the XSProc module is to expand user input in the COMPOSITION block into nuclear number densities (nuclei/b-cm) for every nuclide in each defined mixture. Mixtures can be specified through the direct use of materials presented in the Standard Composition Library, which includes individual nuclides, elements with natural abundances, numerous compounds, alloys and mixtures found in engineering practice, as well several variations of fissile solutions. Additionally, users may define their own materials as atom percent or weight percent combinations. Nuclear masses and theoretical densities are provided in the Standard Composition Library, and methods are available to determine equilibrium states for fissile solutions. Input options for composition data are described in Standard composition specification data with several examples provided in Appendix A.

Unit cells for MG resonance self-shielding

XSProc utilizes a unit cell description to provide information for resonance self-shielding calculations of the input mixtures. As many unit cells as needed to describe the problem may be specified; however, each mixture (other than 0 for a void mixture) can appear only in one unit cell in the CELLDATA block. If a nuclide appears in more than one mixture, multiple sets of self-shielded cross sections are calculated for the nuclide—one for each mixture in each unit cell. Four types of cells are available for self-shielding calculations: INFHOMMEDIUM, LATTICECELL, MULTIREGION, and DOUBLEHET. The default calculation type is CENTRM/PMC for CSAS (see CSAS5: Control Module For Enhanced Criticality Safety Analysis Sequences With KENO V.a), TRITON, (see 3-0) and TSUNAMI (see 6-0) sequences and BONAMI for MAVRIC. All materials not specified in a unit cell are treated as infinite homogeneous media and shielded with BONAMI only, unless the mixture contains a fissionable nuclide, in which case an infinite medium CENTRM/PMC model is used. Note that previous versions of SCALE used infinite medium CENTRM/PMC calculations for all unassigned mixtures. The default type of self-shielding calculation can be overridden, as described in XSProc input data. The following is a brief description of the types of unit cells that can be input in CELLDATA and the computation procedures used.

INFHOMMEDIUM (infinite homogeneous medium) Treatment

The INFHOMMEDIUM treatment is best suited for large masses of materials where the size of each material is large compared with the average mean-free path of the material or where the fraction of the material that is a mean-free path from the surface of the material is very small. When INFHOMMEDIUM cell is specified, the material in the unit cell is treated as an infinite homogeneous lump. Systems composed of small fuel lumps or resonance nuclides sandwiched between moderating regions should not be treated as infinite homogeneous media. In these cases a MULTIREGION or LATTICECELL geometry should be used.

LATTICECELL Treatment

The LATTICECELL model is appropriate for arrays of resonance absorber mixtures—with or without clad—arranged in a square or a triangular pitch configuration within a moderator. Annular fuel (e.g., with an internal moderator in the center) can also be addressed. Input data for the LATTICECELL treatment are described in Unit cell specification for LATTICECELL problems. Self-shielded cross sections are generated for each material zone in a unit cell of the lattice. If a nuclide appears in more than one zone, self-shielded cross sections are produced for each zone where the nuclide is present. Limitations of the LATTICECELL treatment are listed below.

1. The cell description is limited to unit cells for arrays of spherical, plate (slab), or cylindrical fuel bodies. In the case of cylindrical pins in a square-pitch lattice, the default (parm=centrm) self-shielding calculation uses the CENTRM method of characteristics (MoC) option to represent the 2D rectangular unit cell with reflected boundary conditions. By default, self-shielding for all other arrays uses a CENTRM 1D SN calculation for the unit cell (spherical and cylindrical geometries use Wigner-Seitz cells). If parm=bonami is specified, heterogeneous self-shielding effects are treated by equivalence theory [Wil11] The computation option parm=2region, described in XSProc data checking and resonance processing options, can also be used for self-shielding lattice cells.

2. Only predefined choices of cell configurations are available. The available options are described in detail in Unit cell specification for LATTICECELL problems.

3. The basic treatment for LATTICECELL assumes an infinite, uniform array of unit cells. This assumption is a good approximation for interior fuel regions within a large, uniform array. The approximation becomes less rigorous for fuel regions on the periphery of the array or adjacent to a nonuniformity (e.g., control rod, water hole, etc.) in the lattice. For some cases it may be desirable to address this issue by specifying a different lattice cell for this type of fuel pin and using a modified procedure to define an effective unit cell, as described below.

**** LATTICECELL treatment for nonuniform arrays*.

Nonuniform lattice effects may be treated in CENTRM calculation by specifying the keyword DAN2PITCH=dancoff in the optional CENTRM DATA (see Optional CENTRM DATA parameter data). In this approach, the SCALE standalone code MCDancoff must be run prior to the self-shielding calculation in order to compute Dancoff factors for the fuel regions of interest in the nonuniform lattice configuration. MCDancoff performs a simplified one-group Monte Carlo calculation to compute Dancoff factors for complex geometries (see 7-8). The Dancoff value for the fuel region of interest is assigned to the DAN2PITCH keyword in the input for the corresponding cell. Using an iterative procedure, CENTRM computes the pitch of a uniform lattice that has the same Dancoff value as the nonuniform lattice.

MULTIREGION Treatment

The MULTIREGION treatment is appropriate for 1-D geometric regions where the geometry effects may be important, but the limited number of zones and boundary conditions in the LATTICECELL treatment are not applicable. The MULTIREGION unit cell allows more flexibility in the placement of the mixtures but requires all regions of the cell to have the same geometric shape (i.e., slab, cylinder, sphere, buckled slab, or buckled cylinder). Lattice arrangements can be approximated by specifying a non-vacuum boundary condition on the outer boundary. See Unit cell specification for MULTIREGION cells for more details. Limitations of the MULTIREGION cell treatment are listed below.

1. A MULTIREGION cell is limited to a 1-D approximation of the system being represented. An exact 1D model can be defined for the following multizone geometries with vacuum boundary conditions: spheres, infinitely long cylinders, and slabs; and for an infinite array of slabs with reflected or periodic boundaries.

2. The shape of the outer boundary of the MULTIREGION cell is the same as the shape of the inner regions. Cells with curved outer surfaces cannot be stacked physically to create arrays; however, arrays can be approximated by a Wigner-Sietz cell with a white outer boundary condition, where the outer radius is defined to preserve the area of the true rectangular or hexagonal unit cell.

3. Boundary conditions available in a MULTIREGION problem include vacuum (eliminated at the boundary), reflected (reflected about the normal to the surface at the point of impact), periodic (a particle exiting the surface effectively enters an identical cell having the same orientation and continues traveling in the same direction), and white (isotropic return about the point of impact). Reflected and periodic boundary conditions on a slab can represent real physical situations but are not valid on a curved outer surface. A single, non-interacting cell has a vacuum outer boundary condition. If the cell outer boundary condition is not a vacuum boundary, the unit cell approximates some type of array.

4. When using the CENTRM/PMC self-shielding method, the MULTIREGION cell model must include fissionable material. This can be accomplished by adding a trace amount of a fissionable material to one or more mixtures, or by modeling a region of homogenized fuel and water, or by adding a thin (e.g., 1e-6 cm-thick) layer containing at least a trace of a fissionable nuclide on the periphery of the problem.

DOUBLEHET Treatment

DOUBLEHET cells use a specialized CENTRM/PMC calculational approach to treat resonance self-shielding in “doubly heterogeneous” systems. The fuel for these systems typically consists of small, heterogeneous, spherical fuel particles (grains) embedded in a moderator matrix to form the fuel compact. The fuel-grain/matrix compact constitutes the first level of heterogeneity. Cylindrical(rod). spherical (pebble), or slab fuel elements composed of the compact material are arranged in a moderating medium to form a regular or irregular lattice, producing the second level of heterogeneity. The fuel elements are also referred to as “macro cells.” Advanced reactor fuel designs that use TRISO (tri-material, isotopic) or fully ceramic microencapsulated (FCM) fuel require the DOUBLEHET treatment to account for both levels of heterogeneities in the self-shielding calculations. Simply ignoring the double-heterogeneity by volume-weighting the fuel grains and matrix material into a homogenized compact mixture can result in a large reactivity bias.

In the DOUBLEHET cell input, keywords and the geometry description for grains are similar to those of the MULTIREGION treatment, while keywords and the geometry for the fuel element (macro-cell) are similar to those of the LATTICECELL treatment. The following rules apply to the DOUBLEHET cell treatment and must be followed. Violation of any rules may cause a fatal error.

1. As many grain types as needed may be specified for each unique fuel element. Note that grain type is different from the number of grains of a certain type. For example, a fuel element that contains both UO2 and PuO2 grains has two grain types. The same fuel element may contain 10000 UO2 grains and 5000 PuO2 grains. In this case, the number of grains of type UO2 is 10000, and the number of grains of type PuO2 is 5000.

2. As many fuel elements as needed may be specified, each requiring its own DOUBLEHET cell. This may be the case for systems with many fuel elements at different fuel enrichments, burnable poisons, etc. Each fuel element may have one or more grain types.

3. Since the grains are homogenized into a new mixture to be used in the fuel element (macro-cell) cell calculation, a unique fuel mixture number must be entered. XSProc creates a new material with the new mixture number designated by the keyword fuelmix=, containing all the nuclides that are homogenized. The user must assign the new mixture number in the transport solver geometry (e.g., KENO) input unless a cell-weighted mixture is created.

4. The type of lattice or array configuration for the fuel-element may be spheres on a triangular pitch (SPHTRIANGP), spheres on a square pitch (SPHSQUAREP), annular spheres on a triangular pitch (ASPHTRIANGP), annular spheres on a square pitch (ASPHSQUAREP), cylindrical rods on a triangular pitch (TRIANGPITCH), cylindrical rods on a square pitch (SQUAREPITCH),annular cylinderical rods on a triangular pitch (ATRIANGPITCH), annular cylindrical rods on a square pitch (ASQUAREPITCH), a symmetric slab (SYMMSLABCELL), or an asymmetric slab (ASYMSLABCELL).

5. If there is only one grain type for a fuel element, the user must enter either the pitch, the aggregate number of particles in the element, or the volume fraction for the grains. The code needs the pitch and will directly use it if entered. If pitch is not given, then the volume fraction (if given) is used to calculate the pitch. If neither the pitch nor the volume fraction is given, then the number of particles is used to calculate the pitch and the volume fraction. The user should only enter one of these items.

If the fuel matrix contains more than one grain type, all types are homogenized into a single mixture for the compact. As for the one grain type case, the pitch is needed for the spherical cell calculations. However, the pitch by itself is not sufficient to perform the homogenization. Since each grain’s volume is known (grain dimensions must always be entered), entering the number of particles for each grain type essentially provides the total volume of each grain type and therefore enables the calculation of the volume fraction and the pitch. Likewise, entering the volume fraction for each grain type essentially provides the total volume of each grain type and therefore enables the calculation of the number of particles and the pitch. Therefore, one of these two quantities must be entered for multiple grain types. In these cases, since pitch is not given, the available matrix material is distributed around the grains of each grain type proportional to the grain volume and is used to calculate the corresponding pitch. Over-specification is allowed as long as the values are not inconsistent to greater than 0.01%.

1. For cylindrical rods and for slabs, fuel height must also be specified. For slabs the slab width must also be specified.

2. The CENTRM calculation option must be Sn.

Cell weighting of MG cross sections

Cell-weighted self-shielded cross sections are created when CELLMIX= is specified in a LATTICECELL or MULTIREGION cell input. In this case, after finishing the self-shielding calculations for all mixtures in the cell, XSProc calls the computational module XSDRNPM, which solves the 1-D MG transport equation to obtain k and space-dependent MG fluxes for the cell. The resultant fluxes are used to compute MG flux disadvantage factors for processing cell-weighted cross sections of all nuclides in the cell. When the cell-weighted cross sections are used with homogenized number densities of the cell nuclides, the reaction rates of the homogenized mixture preserve the spatially averaged reactions rates of the heterogeneous configuration. The user must input a new mixture ID to identify the homogenized mixture associated with the cell-weighted cross sections. This homogenized mixture should not be used in the heterogeneous geometry data for other transport codes such as KENO, NEWT, etc. Instead, the cell-homogenized mixture that is created should be used at the location of the original cell. Also, cell weighted homogenized cross sections should not be used in MG sensitivity data calculations performed using the TSUNAMI sequences.

XSPROC Input Data Guide

XSProc input data are entered in free form, allowing alphanumeric data, floating-point data, and integer data to be entered in an unstructured manner. Up to 252 characters per line are allowed. Data can usually start or end in any column. Each data entry must be followed by one or more blanks to terminate the data entry. For numeric data, either a comma or a blank can be used to terminate each data entry. Integers may be entered for floating values. For example, 10 will be interpreted as 10.0 if a floating point value is required. Imbedded blanks are not allowed within a data entry unless an E precedes a single blank as in an unsigned exponent in a floating-point number. For example, 1.0E 4 would be correctly interpreted as 1.0 × 104. A number with a negative exponent must include an “E”. For example 1.0-4 cannot be used for 1.0E-4.

The word “END” is a special data item. An END may have a name or label associated with it. The name or label associated with an END is separated from the END by a single blank and is a maximum of 12 characters long. At least two blanks or a new line MUST follow every labeled and unlabeled END. WARNING: It is the user’s responsibility to ensure compliance with this restriction. Failure to observe this restriction can result in the use of incorrect or incomplete data without the benefit of warning or error messages.

Multiple entries of the same data value can be achieved by specifying the number of times the data value is to be entered, followed by either R, *, or $, followed by the data value to be repeated. Imbedded blanks are not allowed between the number of repeats and the repeat flag. For example, 5R12, 5*12, 5$12, or 5R 12, etc., will enter five successive 12’s in the input data. Multiple zeros can be specified as nZ where n is the number of zeroes to be entered.

XSProc data checking and resonance processing options

To check the XSProc input data, run CSAS-MG and specify PARM=CHECK or PARM=CHK after the sequence specification as shown below.

=CSAS-MG PARM=CHK

In this case the actual XSProc cross section processing calculations are not performed. The input data are checked, the problem description is printed, appropriate error and warning messages are printed, and a table of additional data is printed.

Resonance processing will automatically be performed by the default method for the sequence selected. The default methods are CENTRM/PMC for CSAS, TRITON, and TSUNAMI sequences and BONAMI for the MAVRIC sequences. Alternatively, a resonance processing procedure may be chosen by entering PARM=option, where option CENTRM selects the recommended CENTRM/PMC transport method for each cell type, option 2REGION selects the CENTRM/PMC two-region calculation, and option BONAMI applies full range Bondarenko factors to all energy groups without utilizing CENTRM/PMC. For example, to run CSAS1X sequence using only BONAMI for self-shielding, rather than the default CENTRM/PMC method, enter the computational sequence specification shown below.

=CSAS1X PARM=BONAMI

Multiple PARM options are specified by enclosing parameters in parenthesis, such as

=CSAS1X PARM=(CHK, BONAMI)

XSProc resonance self-shielding options are summarized below.

PARM=BONAMI.

This is the fastest MG processing method. It performs resonance self-shielding for all energy groups using the Bondarenko method. BONAMI computes the appropriate background cross section of a given unit cell and then interpolates the corresponding shielding factor from Bondarenko factors on the MG library. Dancoff factors needed to evaluate the background cross section for lattices are computed internally, but these can be overridden by input values in the MORE DATA block. More details on this method are given in the BONAMI section of the manual.

PARM=CENTRM.

This executes the CENTRM/PMC modules to process shielded MG cross sections using CE flux spectra calculated with the recommended type of CE transport solver for the designated type of cell. The CENTRM-recommended CE transport solvers are (a) infinite homogeneous medium calculation for INFHOMMEDIUM cells; (b) 2D MoC transport calculation for a LATTICECELL consisting of cylindrical fuel pins in a square lattice; and (c) 1-D discrete Sn transport for all other LATTICECELLs and for all MULTIREGION cells. The recommended type of transport solver can be overridden for individual cells, as well as for selected energy ranges, by using the CENTRM DATA block described in Optional CENTRM DATA parameter data.

PARM=2REGION.

The CENTRM two-region (2R) option computes the PW flux using a simplified collision probability method for an absorber (e.g., fuel) region surrounded by an external moderator region which has an asymptotic energy spectrum. To account for the heterogeneous effects of a lattice, a correction known as the Dancoff factor is applied to the escape probabilities in the 2R calculation (see the CENTRM chapter of the SCALE manual). These Dancoff factors are calculated internally by XSProc for a uniform array of mixtures in slab, spherical, or cylindrical geometries. These mixture-dependent Dancoff factors can be modified by user input using the DAN parameters contained in the MORE DATA block, as defined in Optional MORE DATA parameter data.

Note on CENTRM/PMC self-shielding options:

The energy range of the CENTRM flux calculation is subdivided into three sections: fast, PW, and low energy. PMC only computes self-shielded cross sections for groups within the PW range defined by parameters demax and demin, which, respectively, define the upper and lower energies of the CENTRM PW flux calculation. Problem-dependent cross sections for groups in the fast and low energy ranges are obtained with the more approximate BONAMI method. Default values for parameters demax and demin are defined appropriately for self-shielding of important resonance materials in thermal reactor systems. The PW self-shielding range can be extended or decreased for individual cells by modifying these parameters using CENTRM DATA.

XSProc input data

The types of input data required for XSProc are given in Table 34, and individual entries are explained in the text following the table. The title, cross section library name (either CE or MG), and standard composition specification data (READ COMP input block) are required for all sequences that use XSProc. The name of the cross section library is used to determine if the transport solver is executed using CE or MG data (e.g., CE or MG KENO calculations). The unit cell descriptions (READ CELL input block) are only used for MG self-shielding calculations. If the specified sequence executes in CE mode, the cell data input can be omitted, or it will be skipped if present. If the cell data information is omitted for MG calculations, all mixtures are self-shielded using the infinite medium approximation.

There are seven standard SCALE sequences that run just XSProc, and produce a MG cross section library or libraries.

=XSPROC produces three libraries with an optional fourth library.

• sysin.microLib is a self-shielded library of the individual nuclides in the problem for use in a later transport calculation,

• sysin.macroLib is a self-shielded library of the mixture cross sections in the problem for use in a later transport calculation,

• sysin.smallMicroLib is a self-shielded library of specific reaction rate cross sections and the elastic and total inelastic scattering transfer matrices for later use in calculating reaction rates and sensitivity values, and

• sysin.xsdrnWeightedLib is an optional library produced if the input specifies having XSDRN do a weighting calculation. This can be a cell weighted and/or a group collapse calculation. The library can be either individual nuclides or mixtures, depending on input.

=CSAS-MG produces an ft04f001 library that is equivalent to the sysin.microLib. With appropriate input it can also produce an ft03f001 which is equivalent to sysin.xsdrnWeightedLib above.

=CSASI or =CSASIX produce an ft04f001 library that is equivalent to sysin.microLib, and an ft02f001 library that is equivalent to sysin.macroLib. CSASIX will run an XSDRN on the first cell without any MOREDATA input. With appropriate input they both can produce an ft03f001 that is the equivalent of sysin.xsdrnWeightedLib.

=CSAS1 or =CSAS1X produce an ft04f001 library that is the equivalent of sysin.microLib. Both sequences will run an XSDRN on the first cell. With appropriate input, they both can produce an ft03f001 that is the equivalent of sysin.xsdrnWeightLib.

=T-XSEC produces an ft04f001 library that is equivalent to sysin.macroLib. and an ft44f001 library that is equivalent to sysin.microLib.

The reactions (MT numbers) written to each library are listed in the SequenceNeutronMT.txt file located in the etc directory installed with SCALE.