How to "compact" the rst file to only include the last substep data?

[Rfaelv]

Hi,
For my static nonlinear analysis I would like to save only the last substep data into my rst file (to avoid large files). It seems that via pyMAPDL it's not possible to do this. Then, is there any way to convert a rst file with all the substep to a rst file with only the last one using pydpf-post?

[PProfizi]

Hello @Rfaelv

We can ask @germa89 to confirm it is not possible to only store the last substep in the RST with PyMAPDL.

Otherwise, sadly DPF is not yet able to write RST files, so it could not convert RST to RST with less data.
What we can do however is filter down the initial RST data and store it in hdf5 or some other format currently available for export.
We have several serialization and deserialization operators you can find here under the serialization category.

I'll investigate whether some other tools would allow to do this.

[germa89]

@PProfizi I am not sure .. pinging @mikerife in case he knows.

[mikerife]

Hi @Rfaelv @PProfizi @germa89
Yes MAPDL, and so PyMAPDL, can be set up to only write the last substep of each load step to the result file during a solution; this is actually the default behavior of MAPDL. Please see the help on the OUTRES command:
PyMAPDL doc page on OUTRES
MAPDL Public Doc page on OUTRES

Now if you have an result file that you want to delete result sets from, you can do this in the /AUX3 processor. Its command DELETE sets which result set to delete, then the command COMPRESS actuall deletes the data from the file. FILEAUX3 sets the result file to be operated on. Please see the public MAPDL help page here on the command grouping, and review all of the individual command pages. Then maybe try a test on a sample model to get the hang of the procedure.

Mike

[Rfaelv]

Hi, @mikerife!
Thanks for the response. But what I'd really like to know is whether it's possible to save only the data from the last load step. Perhaps I have mistaken the terms.

[mikerife]

Hi @Rfaelv yes that is possible. Please see the help on the OUTRES command I linked prior. You can do this two ways:

1 Set up an array of times to save the results. In this case the array just has one entry, the end time. Use that in the 'Freq' field of the OUTRES command.

2 For the load steps prior to the last load step issue OUTRES,ALL,NONE to not write any results at all. Then for the last load step reissue the OUTRES command to write the last substep of this last load step.

I am assuming here you mean to ask "how do I set up the analysis to save the results at the last time point prior to actually solving". If you really mean "I have a result file that has the results at all time points; how do can I delete all results at all times except at the end time" then we need to use that /AUX3 and COMPRESS method. So, are you asking about how to set up a model for solving or how to work with the result file post solving?

Mike