SEACAS

The Sandia National Laboratories (SNL) Engineering Analysis Code Access System (SEACAS) contains the Exodus and IOSS finite element input/output libraries and supporting meshing, preprocessing, postprocessing, graphics, database translation applications, and support libraries. It is used by analysts at SNL and elsewhere and runs on laptops to large HPC systems.

WIKI: FAQ and Extensions

The SEACAS WIKI contains some FAQ-like answers and also a listing of proposed Exodus extensions.

Recent Changes Documents changes to the SEACAS

applications and libraries that may not yet be covered in the documentation.

Using SEACAS on Parallel Computers

Out-of-date, but some information is still useful. Instructions to run the SEACAS/ACCESS system on Parallel computers. Currently specific to Sandia National Laboratories systems.

Download

SEACAS source code and build instructions can be accessed at <https://github.com/sandialabs/seacas>.

Exodus Library

Exodus is a model developed to store and retrieve data for finite element analyses. It is used for preprocessing (problem definition), postprocessing (results visualization), as well as code to code data transfer. An Exodus data file is a random access, machine independent, binary file that is written and read via C, C++, or Fortran library routines which comprise the Application Programming Interface. Exodus uses NetCDF Library as the on-disk storage format.

The new version includes support for names, nodeset variables, sideset variables, named attributes, coordinate frames, concatenated element block definition, optional multiple named node and element maps, assemblies, blobs, and other cleanups.

Element Types

A list of all element types supported by Exodus (and IOSS Library) is provided in Exodus Element Types

Full Model Topology

The API has also been modified to store the full model topology nodes->edges->faces->elements including blocks and sets of all entities. The API extensions are documented in Chapter 4 of SAND2007-0525, A data storage model for novel partial differential equation discretizations.

64 Bit Integers

The API was modified in 2012 to permit the storage of more than 2.1 Billion nodes and/or elements. The changes are documented in 64-bit integers

Large File

Documentation of the modifications needed to use the “large-file” modifications which permit storage of models with more than ~30 million elements is found in Large Model

Polyhedral elements

The documentation for this is in Polyhedral Element Storage

Long Names

The API now supports names longer (or shorter) than 32 characters. The names include variable names and entity names. The changes are documented in Long Names.

Nemesis

All nemesis functions have been added to the exodus API. In most cases, the name of the exodus function is the same as the original nemesis function except that the leading ne_ has been replaced with ex_. There are a few exceptions. See Nemesis API Mapping for a complete list of functions.

Extensions

Potential and proposed Exodus extensions and their status.

Exodus Python Modules

Two python modules are available for accessing exodus databases:

IOSS Library

The IO Subsystem (IOSS) is a high-level database access API that has been designed to give a format-agnostic interface to I/O capabilities with multiple format-specic backends to output the data to a particular format. Currently supported input and output formats are Exodus, CGNS, HeartBeat, History, Generated, Pamgen, TextMesh, ADIOS2, FAODEL, Catalyst, Null.

It has been designed to support multiple database formats simultaneously. It is possible to have the finite element model read from an Exodus database; two results files being written to an Exodus file with a third results file being written to a CGNS file; and the restart file being written to yet another Exodus file. Each of these output databases can have a different schedule for when to write and what data is to be written.

  • The IOSystem.pdf is a medium- to low-level view of the IO system (IOSS) targeted at developers who will be adding or modifying the database IO portion of the system. It should give enough detail that a new database type could be added by reading this document and looking at an existing database class. It is also helpful to have the doxygen-generated documentation for the .

  • Ioss-exodus-mapping.pdf describes how an Exodus model is mapped into the IOSS representation.

  • Properties can be used to control the behavior of the IOSS library.

SUPES Library

The SUPES library is a collection of subprograms which perform frequently used non-numerical services for the engineering applications programmer using FORTRAN-77. The three functional categories of SUPES are: (1) input command parsing, (2) dynamic memory management, and (3) system dependent utilities. The subprograms in categories one and two are written in standard FORTRAN-77, while the subprograms in category three are written to provide a standardized FORTRAN interface to several system dependent features.

Nemesis Library

Note

All of the functionality of the Nemesis library is available in the Exodus library. The Nemesis library is only maintained for backward compatability with older applications. If you need this capability, please use the Exodus library instead.

The Nemesis library is an enhancement to the Exodus finite element database model used to store and retrieve data for unstructured parallel finite element analyses. Nemesis adds data structures which facilitate the partitioning of a scalar (standard serial) Exodus file onto parallel disk systems found on many parallel computers. Since the Nemesis application programming interface (API) can be used to append information to an existing Exodus database, any existing software that reads Exodus files can be used on files which contain Nemesis information. The Nemesis information is written and read via C or C++ callable functions which compromise the Nemesis API.

  • All nemesis functions have been added to the exodus API. In most cases, the name of the exodus function is the same as the original nemesis function except that the leading ne_ has been replaced with ex_. There are a few exceptions. See Nemesis API Mapping for a complete list of functions.

  • Fortran to C Function Mapping

Chaco Library

Note

CHACO is an older library which is not being enhanced. If you are writing a new code that needs graph partitioning, you should use the Zoltan2 library instead.

Graph partitioning is a fundamental problem in many scientific contexts. This document describes the capabilities and operation of Chaco 2.0, a software package designed to partition graphs. Chaco 2.0 allows for recursive application of several methods for finding small edge separators in weighted graphs These methods include inertial, spectral, Kernighan-Lin, and multilevel methods in addition to several simpler strategies Each of these approaches can be used to partition the graph into two, four, or eight pieces at each level of recursion In addition, the Kernighan-Lin method can be used to improve partitions generated by any of the other algorithms. Brief descriptions of these methods are provided along with references to relevant literature. Chaco 2.0 can also be used to address various graph sequencing problems, and this capability is briefly described. The user interface input/output formats and appropriate settings for a variety of code parameters are discussed in detail and some suggestions on algorithm selection are offered.

Note

The short *assignment argument to the interface function has been changed to int *assignment to permit decompositions with more than 32,768 processors.

There are also now a CHACO_VERSION_MAJOR, CHACO_VERSION_MINOR, and CHACO_VERSION_PATCH defines in chaco.h.

Algebra

Algebra allows the user to manipulate data from a finite element analysis before it is plotted. The finite element output data is in the form of variable values (e.g., stress, strain, and velocity components) in an Exodus database. The ALGEBRA program evaluates user-supplied functions of the data and writes the results to an output Exodus database which can be read by plot programs.

Aprepro

Aprepro is an algebraic preprocessor that reads a file containing both general text and algebraic, string, or conditional expressions. It interprets the expressions and outputs them to the output file along with the general text. Aprepro contains several mathematical functions, string functions, and flow control constructs. In addition, functions are included that implement a units conversion system, querying of an exodus database, and 2-dimensional arrays.

Blot

Blot is a graphics program for post-processing of finite element analyses output in the Exodus database format. BLOT produces mesh plots with various representations of the analysis output variables. The major mesh plot capabilities are deformed mesh plots, line contours, filled (painted) contours, vector plots of two/three variables (e.g., velocity vectors), and symbol plots of scalar variables (e.g., discrete cracks). Pathlines of analysis variables can also be drawn on the mesh. BLOT’s features include element selection by material, element birth and death, multiple views for combining several displays on each plot, symmetry mirroring, and node and element numbering. BLOT can also produce X-Y curve plots of the analysis variables. BLOT generates time-versus-variable plots or variable-versus-variable plots. It also generates distance-versus-variable plots at selected time steps where the distance is the accumulated distance between pairs of nodes or element centers.

CGNS_Decomp

An application which will give information on how the parallel CGNS structured mesh decomposition will perform. This application runs in serial, but performs the same calculations as the parallel CGNS structured mesh decomposition routines. Various metrics about the decomposition will be output to allow comparisons of various options and to judge the quality of the decomposition. The input database must be a structured mesh in a CGNS format file.

Conjoin

Conjoin joins two or more Exodus databases into a single database. The input databases should represent the same model geometry with similar variables. The output database will contain the model geometry and all of the non-temporally-overlapping results data. If two databases have overlapping timestep ranges, the timesteps from the later database will be used. For example, if the first database contains time data from 0 to 5 seconds, and the second database contains time data from 4 to 10 seconds; the output database will contain time data from 0 to 4 seconds from the first database and time data from 4 to 10 seconds from the second database. If two nodes have the same global id and are also collocated, then they are combined to a single node in the output. Similarly, elements with the same global id and the same nodal connectivity are combined into a single element in the output file. The output database will contain the union of the meta and bulk data entities (i.e., nodes, elements, element blocks, sidesets, and nodesets) from each input database. The existence of an entity at a particular timestep is indicated via a status variable. Replaces conex

CPUP

CPUP combines multiple CGNS structured mesh databases produced by a parallel application into a single CGNS database. Similar to EPU.

Decomp

A script which calls nem_slice and nem_spread to decompose an Exodus database for use in a parallel application which uses a file-per-rank method for the input mesh database. EPU can be used to recombine decomposed parallel files into a single database.

EJoin

EJoin is used to join two or more Exodus databases into a single Exodus database. The input databases must have disjoint meta and bulk data. That is: element blocks are not combined in the output model. Each element block in each input file will produce an element block in the output file. Similarly for nodesets and sidesets. Each node in each input file will produce a node in the output file unless one of the node matching options (-match node ids or -match node coordinates) is specified. Each element in each input file will produce an element in the output file. Elements are never combined even if all of the nodes on two elements are combined, the output file would have two elements with identical connectivity which is usually not desired. If any of the input databases have timesteps, then the timestep values and counts must match on all databases with timesteps.

EPU

EPU combines multiple Exodus databases produced by a parallel application into a single Exodus database. Replaces nem_join. One of the typical processes for performing parallel analyses with Exodus databases is to decompose the finite element model into multiple pieces such that each processor can read and write its own portion of the finite element model and results data. For example, if a parallel analysis is to be run on the mesh file mesh.g using 8 processors, then mesh.g will be decomposed into 8 pieces or submeshes: mesh.g.8.0, mesh.g.8.1, …, mesh.g.8.7. Each submesh will contain a subset of the nodes and elements of the entire mesh and some communication data indicating which nodes and elements are on the boundary of this submesh and the submesh of one or more other processors.

The analysis code is then executed in parallel and each processor reads its portion of the mesh from its respective submesh; when it outputs results and/or restart data, it creates a new file containing its portion of the submesh and the results that are calculated on that submesh. An “N” processor run will create “N” separate files for each results and/or restart “dataset” that it creates.

The analyst may want to visualize or postprocess the data in the submeshes as a single mesh, so each submesh needs to be joined together to create a single “global” file containing all of the data.1

This joining together of parallel submeshes is the purpose of EPU. It will read the data from each submesh and map it into the correct location in the “global” file; discarding duplicate data as required.

Exo_Format

Exo_format gives information on the internal information of an exodus file including the integer and floating point size, the name length, and the underlying variant of the netCDF file.

The command exo_format --config will report on the third-party library versions that the Exodus library was configured with.

Exo2Mat

Exo2Mat See Mat2Exo documentation.

Exodiff

Exodiff compares the results data from two Exodus databases. The databases should represent the same model, that is, the Ex odus meta data should be identical as should be the Exodus portion of the bulk data. The only differences should be in the values of the transient bulk data. Exodiff’s main purpose is to detect and report these differences. Exodiff will compare global, nodal, element, nodeset, and sideset transient variables at each selected timestep; it will also compare element attribute variables on each element block containing attributes.

Exomatlab

ExoMatLab outputs selected global data to a text matlab file.

Exotxt

Exotxt converts an exodus file into a text file which can be edited or used as input to other processing codes that need a text format. Can be converted back to exodus using TxtExo. (The netCDF utilities ncdump/ncgen can also be used to convert an exodus files to/from text.)

Explore

Explore is a program that examines the input to a finite element analysis or the output from an analysis in the Exodus database format. EXPLORE allows the user to examine any value in the database. The display can be directed to the user’s terminal or to a print file.

Fastq

FASTQ is an interactive two-dimensional finite element mesh generation program, It is designed to provide a powerful and efficient tool to both reduce the time required of an analyst to generate a mesh, and to improve the capacity to generate good meshes in arbitrary geometries. It is based on a mapping technique and employs a set of “higher-order” primitives which have been developed for automatic meshing of commonly encountered shapes (i.e. the triangle, semi-circle, etc.) and conditions (i.e. mesh transitioning from coarse to fine mesh size. ) FASTQ has been designed to allow user flexibility and control. The user interface is built on a layered . command level structure. Multiple utilities rue provided for input, manipulation, and display of the geometric information, as well as for direct control, adjustment, and display of the generated mesh. Enhanced boundary flagging has been incorporated and multiple element types and output formats are supported.

Memos documenting features (such as paving) not discussed in the sand report are available in FASTQ-memo.

Gen3D

Gen3D is a three-dimensional mesh generation program. The three-dimensional mesh is generated by mapping a two-dimensional mesh into three-dimensions according to one of four types of transformations: translating, rotating, mapping onto a spherical surface, and mapping onto a cylindrical surface. The generated three-dimensional mesh can then be reoriented by offsetting, reflecting about an axis, and revolving about an axis. GEN3D can be used to mesh geometries that are axisymmetric or planar, but, due to three-dimensional loading or boundary conditions, require a three-dimensional finite element mesh and analysis. More importantly, it can be used to mesh complex three-dimensional geometries composed of several sections when the sections can be defined in terms of transformations of two-dimensional geometries.

Additional commands not documented in the main report are available in gen3d-updates.

GenShell

GenShell is a three-dimensional shell mesh generation program. The three-dimensional shell mesh is generated by mapping a two-dimensional quadrilateral mesh into three dimensions according to one of several types of transformations: translation, mapping onto a spherical, ellipsoidal, or cylindrical surface, and mapping onto a user-defined spline surface. The generated three-dimensional mesh can then be reoriented by offsetting, reflecting about an axis, revolving about an axis, and scaling the coordinates. GENSHELL can be used to mesh complex three-dimensional geometries composed of several sections when the sections can be defined in terms of transformations of two-dimensional geometries.

GJoin

GJoin is a two- or three-dimensional mesh combination program. GJOIN combines two or more meshes written in the Exodus mesh database format into a single Exodus mesh. Selected nodes in the two meshes that are closer than a specified distance can be combined The geometry of the mesh databases can be modified by scaling, offsetting, revolving, and mirroring. The combined meshes can be further modified by deleting, renaming, or combining material blocks, sideset identifications, or nodeset identifications. EJoin is a newer application which has some of the capabilities of GJoin.

Grepos

Grepos is a mesh utility program that repositions or modifies the configuration of a two-dimensional or three-dimensional mesh. GREPOS can be used to change the orientation and size of a two-dimensional or three-dimensional mesh; change the material block, nodeset, and sideset IDs; or “explode” the mesh to facilitate viewing of the various parts of the model.

IO_Info

Utility application which reads a database and provides a summary of the information in the database. See io_info --help for a list of the capabilities of io_modify. If you enter io_info --configuration it will output a summary of the third-party libraries that were used in the build of this installation of SEACAS.

IO_Modify

IO_Modify is an application which can be used to query add assembly and entity attribute information to an existing Exodus database. See io_modify --help for more information. There is an internal HELP command which will provide additional information on the capabilities.

IO_Shell

Utility application which reads a database, possibly applies a modification, and then writes that database. See io_shell --help for a list of the capabilities of io_shell. It can read and write datbases in the Exodus and CGNS format. There is also a generated mesh option. To see the generated mesh options, try: io_shell --in_type generated 1x1x1+help output.g

Mapvar

Note

MAPVAR is buggy and cannot correctly handle the mapping of nodal variables in a mesh with multiple element blocks and multiple timesteps. It should work OK if there is only a single timestep; however, there may be an issue with nodes shared between multiple element blocks. Element variable interpolation should be OK in all cases.

Mapvar is designed to transfer solution results from one finite element mesh to another. MAPVAR draws heavily from the structure and coding of MERLIN II, but it employs a new finite element data base, Exodus, and offers enhanced speed and new capabilities not available in MERLIN II. In keeping with the MERLIN II documentation, the computational algorithms used in MAPVAR are described. User instructions are presented. Example problems are included to demonstrate the operation of the code and the effects of various input options.

Mapvar-kd

Note

MAPVAR is buggy and cannot correctly handle the mapping of nodal variables in a mesh with multiple element blocks and multiple timesteps. It should work OK if there is only a single timestep; however, there may be an issue with nodes shared between multiple element blocks. Element variable interpolation should be OK in all cases.

Mapvar-kd is almost exactly the same as mapvar except that it uses a KD algorithm for the internal search. It is much faster than mapvar in certain situations, and should never be slower.

Mat2Exo

Mat2Exo is a program which translates mesh data from Matlab mat-file format to Exodus format. This tool is the inverse of the commonly used tool Exo2Mat which translates Exodus data to the Matlab mat-file format. These tools provide a means for preprocessing an Exodus model file or post-processing an Exodus results file using Matlab.

Nas2Exo

Convert Nastran format to Exodus database.

nem_join

Note

nem_join is no longer maintained. Use EPU instead.

nem_slice

nem_slice reads in a FEM description of the geometry of a problem from an Exodus file and generates either a nodal or elemental graph of the problem, calls Chaco to load balance the graph, and outputs a Nemesis Library load-balance file which can be read by

nem_spread

Nem_spread reads it’s input command file (default name nem_spread.inp), takes the named Exodus and spreads out the geometry (and optionally results) contained in that file out to a parallel disk system. The decomposition is taken from a scalar Nemesis load balance file generated by the companion utility nem_slice. Here is an example nem_spread input file.

Numbers

Numbers is a program which reads and stores data from a finite element model described in the Exodus database format. Within this program are several utility routines which calculate information about the finite element model. It is limited to models with hexahedral elements only.

Skinner

Create an Exodus mesh consisting of the skin or surface of the input Exodus database.

Slice

Slice decomposes an Exodus Database into multiple databases for use in a parallel application which uses a file-per-rank method for the input mesh database. Experimental.

Struc_To_Unstruc

An application which will read a structured mesh in CGNS format and convert it to an unstructured mesh which can be output in either CGNS or Exodus format.

TxtExo

Txtexo converts a text file written by ExoTxt back to an exodus file. (The netCDF utilities ncdump/ncgen can also be used to convert an exodus files to/from text.)

Zellij

Zellij takes 1 or more “unit cell” template databases and tiles them into a single output database. The unit cells must have a structured boundary on the I-J faces; the K face can have an arbitrary mesh as can the interior of the unit cells. A lattice file is parsed to specify the unit cells and their arrangement into the output database. The output file can be written in a decomposed “file-per-rank” set of files for using with a parallel application. Zellij is optimized for large files.

NetCDF Library

(External, not developed or maintained as part of SEACAS, but is used by Exodus, so include here).

The netCDF I/O library stores and retrieves data in self-describing, machine-independent files. Each netCDF file can contain an unlimited number of multi-dimensional, named variables (with differing types that include integers, reals, characters, bytes, etc.), and each variable may be accompanied by ancillary data, such as units of measure or descriptive text. The interface includes a method for appending data to existing netCDF files in prescribed ways, functionality that is not unlike a (fixed length) record structure. However, the netCDF library also allows direct-access storage and retrieval of data by variable name and index and therefore is useful only for disk-resident (or memory-resident) files. NetCDF information is available from Unidata.

The ncgen and ncdump utilities can be used to convert an Exodus file from/to a text representation.