You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
263 lines
13 KiB
263 lines
13 KiB
2 years ago
|
\chapter{Conjoin}\label{ch:conjoin}
|
||
|
|
\section{Introduction}
|
||
|
|
A finite element analysis can often generate multiple results and/or
|
||
|
|
restart databases due to model bulk data changes or restarting the
|
||
|
|
analysis.
|
||
|
|
|
||
|
|
The bulk data changes are caused by element death, element creation,
|
||
|
|
and surface evolution. Because the \exo{} database requires a
|
||
|
|
constant model meta data and bulk data description, the analysis code
|
||
|
|
must close the current results and restart databases and create new
|
||
|
|
databases if the model meta data or bulk data change. At the end of
|
||
|
|
the analysis, there will be multiple database files all describing a
|
||
|
|
similar model with similar meta data, but with each database
|
||
|
|
containing a different set of bulk data descriptions of that meta
|
||
|
|
data.
|
||
|
|
|
||
|
|
Another cause of multiple \exo{} databases is when the analysis job is
|
||
|
|
restarted one or more times. This is typically the case for
|
||
|
|
long-running analyses whose total execution time exceeds the maximum
|
||
|
|
allowed queue time on a compute cluster. A restart database is
|
||
|
|
written at the end of each compute segment and then the job is
|
||
|
|
resubmitted to the compute cluster and is restarted one or more
|
||
|
|
times. At the end of the analysis, there will be at least one database
|
||
|
|
for each compute segment and it is often desirable to combine these
|
||
|
|
databases into a single database for visualization and post processing.
|
||
|
|
|
||
|
|
\conjoin{} joins two or more \exo{} 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 colocated, 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. For example,
|
||
|
|
if the first database is the mesh shown in Figure~\ref{fig:cjm1} at
|
||
|
|
time 0.0 seconds, and the second database is the mesh shown in
|
||
|
|
Figure~\ref{fig:cjm2} at time 1.0 seconds, then the combined mesh
|
||
|
|
would be the union of the nodes and elements in the two meshes as
|
||
|
|
shown in Figure~\ref{fig:cjall}.
|
||
|
|
\begin{figure}[htp]
|
||
|
|
\centering
|
||
|
|
\subfloat[First Database, time 0]{\label{fig:cjm1}\includegraphics[scale=0.4]{figures/cjm1.png}}
|
||
|
|
\subfloat[Second Database, time 1]{\label{fig:cjm2}\includegraphics[scale=0.4]{figures/cjm2.png}}
|
||
|
|
\subfloat[Combined Database, time 0, 1]{\label{fig:cjall}\includegraphics[scale=0.4]{figures/cjall.png}}
|
||
|
|
\caption{Example Input and Output Meshes for Conjoin}
|
||
|
|
\end{figure}
|
||
|
|
|
||
|
|
The values of the element status variable for each element at each
|
||
|
|
timestep are shown below. A status value of 0 indicates that the
|
||
|
|
element is active and a value of 1 indicates it is inactive. The nodes
|
||
|
|
have a similar status variable.
|
||
|
|
|
||
|
|
\begin{tabular}{l|ccccccc}
|
||
|
|
\hline
|
||
|
|
element & 1 & 2 & 3 & 4 & 5 & 6 & 7 \\
|
||
|
|
status at time 0 & 0 & 0 & 0 & 0 & 1 & 1 & 1 \\
|
||
|
|
status at time 1 & 1 & 1 & 1 & 0 & 0 & 0 & 0 \\
|
||
|
|
\hline
|
||
|
|
\end{tabular}
|
||
|
|
|
||
|
|
The output database will contain the variables that are defined on the
|
||
|
|
first mesh. If later databases do not have those variables, then the
|
||
|
|
data will be zero-filled for those timesteps. If later databases have
|
||
|
|
additional variables, they will be ignored.
|
||
|
|
|
||
|
|
\section{Invoking \conjoin}
|
||
|
|
The minimal command line for executing \conjoin{} is:
|
||
|
|
\begin{syntax}
|
||
|
|
conjoin \{list of files to join\}
|
||
|
|
\end{syntax}
|
||
|
|
This would create an output mesh named \file{conjoin-out.e} containing
|
||
|
|
the entities and data from the files listed on the command line. The
|
||
|
|
files must be listed in order of increasing time step values. A nodal
|
||
|
|
variable named \param{node\_status} and an element variable named
|
||
|
|
\param{elem\_status} will be added to the variables already existing
|
||
|
|
on the first file. The value of the status variables will be 0 to
|
||
|
|
indicate an active node or element and 1 to indicate an inactive node
|
||
|
|
or element.
|
||
|
|
|
||
|
|
Several options are available to modify the default behavior of \conjoin{}.
|
||
|
|
|
||
|
|
\subsection{Options}
|
||
|
|
\renewcommand\arraystretch{1.5}
|
||
|
|
\begin{longtable}{lp{4.0in}}
|
||
|
|
\param{-help} & Print this summary and exit \\
|
||
|
|
\param{-output <val>} & Name of the output file to create. The
|
||
|
|
default output filename is \file{conjoin-out.e}. If the specified
|
||
|
|
file already exists, it will be overwritten. \\
|
||
|
|
|
||
|
|
\param{-alive\_value <val>} & Value (1 or 0) which will be used to indicate that an
|
||
|
|
element is alive or active. The
|
||
|
|
default behavior is that 0 indicates
|
||
|
|
alive or active nodes and elements. \\
|
||
|
|
|
||
|
|
\param{-combine\_status\_variables <val>} & The conjoin elem\_status variable will be combined
|
||
|
|
with the specified status variable (val) existing on the mesh.
|
||
|
|
Both variables must have the same value (1 or 0) to
|
||
|
|
indicate that the element is alive or active.
|
||
|
|
If 1 is alive, then the combined variable is the minimum of the two values.
|
||
|
|
If 0 is alive, then the combined variable is the maximum of the two values.
|
||
|
|
Use the \param{alive\_value} option to set conjoin's alive value \\
|
||
|
|
|
||
|
|
\param{-element\_status\_variable <val>} & Element variable name to use as element
|
||
|
|
existence status variable; it must not exist on input files. If specified
|
||
|
|
as NONE, then it will not be written to the output file.
|
||
|
|
Default = elem\_status \\
|
||
|
|
|
||
|
|
\param{-nodal\_status\_variable <val>} & Nodal variable name to use as nodal status variable;
|
||
|
|
it must not exist on input files. If specified as NONE,
|
||
|
|
then it will not be written to the output file. Default = node\_status \\
|
||
|
|
|
||
|
|
\param{-omit\_nodesets} & Don't transfer nodesets to output file. \\
|
||
|
|
\param{-omit\_sidesets} & Don't transfer sidesets to output file. \\
|
||
|
|
\param{-gvar <val>} & Comma-separated list of global variables to be
|
||
|
|
joined or ALL or NONE. The default behavior is that all global
|
||
|
|
variables will be written to the output file.\\
|
||
|
|
\param{-evar <val>} & Comma-separated list of element variables to be joined or ALL or NONE.
|
||
|
|
Variables can be limited to certain blocks by appending a
|
||
|
|
colon followed by the block id. For example,
|
||
|
|
\param{-evar sigxx:10:20} would result in the variable
|
||
|
|
\param{sigxx} would be written for element blocks with
|
||
|
|
id 10 and 20. The default behavior is that all element
|
||
|
|
variables will be written to the output file. \\
|
||
|
|
\param{-nvar <val>} & Comma-separated list of nodal variables to be
|
||
|
|
joined or ALL or NONE. The default behavior is that all nodal
|
||
|
|
variables will be written to the output file. \\
|
||
|
|
\param{-nsetvar <val>} & Comma-separated list of nodeset variables to
|
||
|
|
be joined or ALL or NONE. The default behavior is that all nodeset
|
||
|
|
variables will be written to the output file.\\
|
||
|
|
\param{-ssetvar <val>} & Comma-separated list of sideset variables to
|
||
|
|
be joined or ALL or NONE. The default behavior is that all sideset
|
||
|
|
variables will be written to the output file.\\
|
||
|
|
\param{-debug <val>} & Debug level (values are or'd)\par
|
||
|
|
\begin{tabular}{l@{ = }l}
|
||
|
|
1 & timing information.\\
|
||
|
|
4 & Verbose Element block information.\\
|
||
|
|
8 & Check consistent nodal coordinates between parts.\\
|
||
|
|
16 & Verbose Sideset information.\\
|
||
|
|
32 & Verbose Nodeset information.\\
|
||
|
|
64 & put exodus library into verbose mode.\\
|
||
|
|
128 & Check consistent global field values between parts. \\
|
||
|
|
\end{tabular}\\
|
||
|
|
\param{-width <val>} & Width of output screen, default = 80 \\
|
||
|
|
\param{-version} & Print version and exit \\
|
||
|
|
\param{-copyright} & Show copyright and license data. \\
|
||
|
|
\end{longtable}
|
||
|
|
|
||
|
|
When \conjoin{} is invoked, it will parse the user-specified options from
|
||
|
|
the command line and it will also see if the environment variable
|
||
|
|
\param{CONJOIN\_OPTIONS} exists. If the variable exists, then it will be
|
||
|
|
parsed first followed by the parsing of the options on the command line.
|
||
|
|
|
||
|
|
\section{Example}
|
||
|
|
The output below shows the results of running the following command:
|
||
|
|
\begin{syntax}
|
||
|
|
conjoin -output adapt.e adaptBox.e*
|
||
|
|
\end{syntax}
|
||
|
|
|
||
|
|
The first section of the output shows the code version followed by some
|
||
|
|
information showing the map from part number to filename. In this case
|
||
|
|
there are 6 parts which will be joined.
|
||
|
|
|
||
|
|
\begin{verbatim}
|
||
|
|
conjoin
|
||
|
|
(A code for sequentially appending Exodus II databases. Supersedes conex and conex2.)
|
||
|
|
(Version: 1.2.1) Modified: 2011/01/12
|
||
|
|
Part 1: 'adaptBox.e'
|
||
|
|
Part 2: 'adaptBox.e-s0002'
|
||
|
|
Part 3: 'adaptBox.e-s0003'
|
||
|
|
Part 4: 'adaptBox.e-s0004'
|
||
|
|
Part 5: 'adaptBox.e-s0005'
|
||
|
|
Part 6: 'adaptBox.e-s0006'
|
||
|
|
\end{verbatim}
|
||
|
|
\sectionline
|
||
|
|
The next section of the output shows the output filename
|
||
|
|
(\file{adapt.e} in this example), and gives a summary of the output
|
||
|
|
mesh showing the entity counts and the variable summary.
|
||
|
|
\begin{verbatim}
|
||
|
|
Output: 'adapt.e'
|
||
|
|
IO Word size is 8 bytes.
|
||
|
|
Title: Default Sierra Title
|
||
|
|
|
||
|
|
Number of coordinates per node = 3
|
||
|
|
Number of nodes = 151
|
||
|
|
Number of elements = 84
|
||
|
|
Number of element blocks = 1
|
||
|
|
|
||
|
|
Number of nodal point sets = 0
|
||
|
|
Number of element side sets = 2
|
||
|
|
|
||
|
|
|
||
|
|
Reading and Writing element connectivity & attributes
|
||
|
|
|
||
|
|
Wrote coordinate names...
|
||
|
|
Wrote coordinate information...
|
||
|
|
Found 7 global variables.
|
||
|
|
ExternalEnergy InternalEnergy KineticEnergy Momentum_x
|
||
|
|
Momentum_y Momentum_z TIMESTEP
|
||
|
|
|
||
|
|
Found 7 nodal variables.
|
||
|
|
displ_x displ_y displ_z vel_x vel_y
|
||
|
|
vel_z node_status
|
||
|
|
|
||
|
|
Found 7 element variables.
|
||
|
|
stress_xx stress_yy stress_zz stress_xy stress_yz
|
||
|
|
stress_zx elem_status
|
||
|
|
\end{verbatim}
|
||
|
|
\sectionline
|
||
|
|
|
||
|
|
The last section shows the progress of transferring the transient data
|
||
|
|
to the output mesh. For each timestep, the output shows the part from
|
||
|
|
which the data is being read and which timestep within that part is
|
||
|
|
being used. It also show how many elements are active in that part
|
||
|
|
and an estimate of how long it will take to finish transferring the
|
||
|
|
data. Figures~\ref{fig:cjex1} to~\ref{fig:cjex4} show four views of
|
||
|
|
the output mesh.
|
||
|
|
|
||
|
|
\begin{verbatim}
|
||
|
|
Step 1/11, time 0.0000e+00 (Part 1/6, step 1) Active Elem: 4 [ 9%, Elapsed=0.0s, ETA=0.0s]
|
||
|
|
Step 2/11, time 5.0000e-01 (Part 1/6, step 2) Active Elem: 4 [ 18%, Elapsed=0.0s, ETA=0.0s]
|
||
|
|
Step 3/11, time 1.0000e+00 (Part 2/6, step 1) Active Elem: 11 [ 27%, Elapsed=0.0s, ETA=0.0s]
|
||
|
|
Step 4/11, time 1.2500e+00 (Part 3/6, step 1) Active Elem: 74 [ 36%, Elapsed=0.0s, ETA=0.0s]
|
||
|
|
Step 5/11, time 1.5000e+00 (Part 3/6, step 2) Active Elem: 74 [ 45%, Elapsed= <1s, ETA= <1s]
|
||
|
|
Step 6/11, time 1.7500e+00 (Part 4/6, step 1) Active Elem: 46 [ 55%, Elapsed= <1s, ETA= <1s]
|
||
|
|
Step 7/11, time 2.0000e+00 (Part 4/6, step 2) Active Elem: 46 [ 64%, Elapsed= <1s, ETA= <1s]
|
||
|
|
Step 8/11, time 2.5000e+00 (Part 4/6, step 3) Active Elem: 46 [ 73%, Elapsed= <1s, ETA= <1s]
|
||
|
|
Step 9/11, time 3.0000e+00 (Part 5/6, step 1) Active Elem: 18 [ 82%, Elapsed= <1s, ETA= <1s]
|
||
|
|
Step 10/11, time 3.5000e+00 (Part 6/6, step 1) Active Elem: 4 [ 91%, Elapsed= <1s, ETA= <1s]
|
||
|
|
Step 11/11, time 4.0000e+00 (Part 6/6, step 2) Active Elem: 4 [100%, Elapsed= <1s, ETA=0.0s]
|
||
|
|
******* END *******
|
||
|
|
\end{verbatim}
|
||
|
|
\begin{figure}[ht]
|
||
|
|
\centering
|
||
|
|
\subfloat[Time 0.0]{\label{fig:cjex1}\includegraphics[scale=0.25]{figures/cjex1s.png}}
|
||
|
|
\subfloat[Time 1.0]{\label{fig:cjex2}\includegraphics[scale=0.25]{figures/cjex2s.png}}\\
|
||
|
|
\subfloat[Time 1.5]{\label{fig:cjex3}\includegraphics[scale=0.25]{figures/cjex3s.png}}
|
||
|
|
\subfloat[Time 2.5]{\label{fig:cjex4}\includegraphics[scale=0.25]{figures/cjex4s.png}}
|
||
|
|
\caption{Example Output Mesh from Conjoin}
|
||
|
|
\end{figure}
|
||
|
|
|
||
|
|
\section{Related Codes}
|
||
|
|
\conjoin{} replaces the functionality of the
|
||
|
|
\code{conex}\cite{bib:conex} code. \conjoin{} provides a superset of
|
||
|
|
the \code{conex} functionality and is also much faster. \conjoin{}
|
||
|
|
also supports nodeset and sideset variables and the naming of element
|
||
|
|
blocks, nodesets, and sidesets which is not supported by
|
||
|
|
\code{conex}. The input mesh databases to \code{Conex} are required to
|
||
|
|
have the exact same bulk data which limits its use for adaptive
|
||
|
|
analyses and other analyses which have varying numbers of active and
|
||
|
|
inactive nodes and elements. \code{Conex} is no longer actively
|
||
|
|
supported.
|