Abaqus Python：如何编辑或删除 odb 输出数据库中的数据

Question

So in the Abaqus Scripting Interface there is no direct way to delete data off an odb file or to duplicate an odb with some data removed from the database.

So how can we edit the odb to remove unwanted output?

Answer 1

One option to construct an edited odb is to use the "odbcombine" Abaqus plugin. I believe this plugin is an officially supported Abaqus product.

The plugin can be imported from the abaqus_plugins folder by

sys.path.insert('r'/abaqus_main/6.14-1/code/python2.7/lib/abaqus_plugins/odbCombine')
import odbCombineKernel
odbCombineKernel.combineOdbs(jobName='myjob',
                             configName='myconfig.xml', 
                             loadODB=1) 

For this command to work it requires an XML input file. The XML file must be configured with something like:

<?xml version="1.0" ?>
<OdbInput>
    <MasterOdb Name="oldjob.odb"/>
</OdbInput>

If you only specify a Master Odb, you don't need additional odb's to combine. Then the odbcombine tool works as an odb filter rather than a combiner.

Answer 2

Another way to create a more or less blank odb is to submit a new analysis to the Abaqus server using the "datacheck" option. However this odb will still contain steps from the input file, but no frames.

To get rid of the steps, you need to submit a modified input file without *step/*end step keywords. A simple function can do that, for example:

def _remove_inp_steps(path, newpath):
    """Remove steps from input file so that a blank odb can be generated

    Parameters
    ----------
    path : str
        Path of original input file 
    newpath : str
        Path of new input file to create with steps removed 
    """
    with open(path, 'r') as f:
        lines = f.readlines()

    key_start = '*step'
    key_end = '*end step'
    is_in_step = False

    # Go through file and remove all data contained within steps.
    line = '** DUMMY INPUT FILE USED TO CREATE A BLANK OUTPUT DATABASE FILE; RUN WITH DATACHECK\n'

    newlines = [line]

    for line in lines:
        line2 = line.lower()

        if line2.startswith(key_start):
            is_in_step = True
        elif line2.startswith(key_end):
            is_in_step = False
        else:
            if not is_in_step:
                newlines.append(line)

    with open(newpath, 'w') as f:
        f.writelines(newlines)
    return