Combining Data from Multiple Output Databases

You can filter the data that the utility includes in the combined output database to include results only from selected steps or frames, from selected output variables, or from a combination of these options. For example, a filter can enable you to include results data only from the last step and the last frame of the specified output databases, and the same filter can dictate that only Mises stress results are included in the combined output database. You can also establish multiple filters if you want to set up different filtering conditions for the first step than in the second step.

The abaqus odbcombine utility also provides two levels of filtering: output database–specific filters, which filter results from only a single output database; and default filters, which apply to the entire job. The output database–specific filters take precedence over the default filters, so Abaqus/CAE employs the settings in the default filters only when the default filter you define does not conflict with filters for one of the individual output databases.

The filtering syntax is flexible enough to allow you to specify multiple steps, frame, or output variable values. You can specify multiple step names in a comma-separated list, such as Step-1, Step-2, Step-4. For frames you can include ranges or individual values; for example, entering 1, 3, 5, 7:9 returns frames 1, 3, 5, 7, 8, and 9 to the combined output database.

You can also use the symbolic constants 'ALL', 'FIRST', and 'LAST' as shortcuts to specify the data you want to include. These options enable you to include results data from all steps or frames and data from all output variables rather than one or more selected variables.

One output database in every combine operation is designated as the primary output database. The utility first transfers all field output data, subject to filtering selections, from the primary output database to the combined output database. The utility then locates results data from matching steps and frames in the subsequent output databases and copies only those data into the combined output database. This strategy provides a more coherent structure for the combined results data.

The abaqus odbcombine utility uses data in configuration files to determine which output databases to combine, the file to designate as the primary output database, and the filtering options to enforce by default and for each output database. The configuration file must be in .xml format, and it can have three types of elements in the following order:

• The <DefaultFilters> element specifies one or more default filtering definitions. This section is optional, but you must include it if you want to set up default filtering for your combine operation. If this section is omitted, the default value used for Steps is 'LAST', for Frames is 'LAST', and for Vars is 'ALL' (that is, results of all variables from the last frame of the last step are included).

• The <PrimaryOdb> element specifies the location of the primary output database and, if desired, one or more filtering definitions for the data in that output database. This section is required.

• One <Odb> element is required for each additional output database that you want to include in the combine operation.

You can then specify default filters for output database–specific filters by embedding <Filter> elements within the <DefaultFilters> element or within one of the output database elements.

The following example illustrates the structure of the configuration file for the abaqus odbcombine utility.

<?xml version='1.0' encoding='UTF-8'?>
Your XML file declaration may differ from this one.
<OdbInput>    <DefaultFilters>
    The default filtering element is optional. If you include this element in the configuration file,
    you must include at least one <Filter> element within this section. Filter elements can
    use the Steps or Frames attributes to refer to symbolic constants or the StepName or
    FrameIndex attributes to refer to individual steps or frames, as shown in the following examples:
        <Filter Steps='step names or symbolic constants'
            Frames='frame numbers or symbolic constants'
            Vars='variable or symbolic constants' />
        <Filter StepName='full step name'
            FrameIndex='individual frame number'
            VariableName='variable' />      </DefaultFilters>
    <PrimaryOdb Name='path to primary output database'>
    Filtering elements for the primary output database are optional. If you want to filter
    the data from this output database, include a <Filter> element within this section
    for each filtering option you want to define.
    </PrimaryOdb>
    <Odb Name='path to output database'>
    Filtering elements for the output database are optional. If you want to filter
    the data from this output database, include a <Filter> element within this section
    for each filtering option you want to define.
    </Odb>
    Append an <Odb> element for each additional output database you want to include.
</OdbInput>

The following types of output data are not included when you combine output database files:

• History output.

• Surface data.

• Data from analytical rigid part instances.

• Local coordinate systems associated with field output data.

When you run a steady-state dynamic analysis, you can use the frequency split method to combine the output files for different frequency partitions. You invoke this method using the ssd_split and ssd_partition parameters on the Abaqus execution procedure. All types of output data supported for steady-state dynamic analyses are included when you combine output database files for multiple frequency partitions.