Batch mode output text file problem

[ashtari]

I have been running SeismoStruct in the batch mode and I wanted to extract the results of non-linear time history analysis from the application. However, the text file it generates contains irrelevant displacement values. I checked each model responses in the post-processor and they were all fine. It is just the batch mode. Can you help me with this problem?

[z.gronti]

Dear ashtari,

Indeed there is a bug in the batch mode of SeismoStruct v7. The bug has been corrected in the upcoming version SeismoStruct 2016 that will
be published in the next 1-2 months.

[manuel.jordan]

Hello

By using SeismoBatch I also did some unexpected experiences.

First of all I recognized that Seismobatch returned a table with only 0 entries when I defined some fibers to return stresses and strains. I deleted the fiber information I wanted to extract in the pre-processor and re-ran all analyses. Afterwards SeismoBatch returned the right values.

Second I recognized that SeismoBatch does not return the right parameter values for different element quantities:
V3(A) in SeismoBatch corresponds to V2(A) in SeismoStruct
V2(A) in SeismoBatch corresponds to Mt in SeismoStruct
Mt in SeismoBatch corresponds to V3(A) in SeismoStruct

It’s possible that there are still other mismatchs.

Regards

[z.gronti]

Dear manuel.jordan,


Indeed there are the reported bugs in the batch mode of SeismoStruct v7. The bugs have been corrected in the upcoming version SeismoStruct 2016 that will be released next Monday, on 4th of July 2016.

[Romain]

Hello,
I am aware of the problems with the batch mode in SStruct V7.
However given that i am using a large amount of models and that they are generate automatically from a Matlab script, i had to consider the SStruct V7 (at least at that time, the *spf in v.2016 were not easy to reproduce).
I notice that we can open the v7 files with SStruct 2016 however, it does not seem to work in the 2016 batch mode.
Currently i am performing the analysis in V7 one by one in a nonautomatic manner, which is really boring and time consuming.
Do you have any suggestion to bypass this problem, even considering any unconventional way...
Many thanks,
Romain

[z.gronti]

Dear Romain,

You may run analyses and extract results of projects created with SeismoStruct v7 within the SeismoBatch 2016. Could you please check if you have installed the latest release of the program? In case that you cannot run the analyses with the current release, could you please send us the files at z.gronti@seismosoft.com to check it?

[Romain]

Thank you Zoi.
I am happy to see that the analyses can be performed with the latest release.
However, if the spf was build with the format of version 7, a message pops up saying that the "control node has not been defined... and the target displacement will not be calculated". The problem is that i can only suppress this message from appearing by changing the settings in SStruct 2016, which will not influence the models defined with v7.
I wonder if it is possible to somehow define the "control node" when setting up the model in the txt file compatible with v7 (even if that model will not open anymore in that version).
Do you have any other suggestion?
Many thanks and best regards,
Romain

[seismosoft]

There is no way to define the control node of the target displacement calculation from v7 models, since this is a new feature in SeismoStruct 2016. The only way to change this is to open the model with SeismoStruct 2016 and specify the control node. However, if you suppress the messages, the model will run in SeismoBatch and you will be able to extract the output results (the target displacement will not be calculated though).
Note also that from SStruct 2016 onwards the appearance of a message in SeismoBatch does not stop the analyses, the messages are automatically suppressed after 2 min,

Seismosoft Support

[Romain]

Hi,
Thanks again for your suggestions.
However I don't think they solve my problem (maybe i wasn't clear enough in my previous post).
Both your options to suppress the messages seem to work fine if the *spf was built or has been previously opened in SStruct2016.
However, if you try to run the analyses from a *spf from Version7 directly in SeismoBatch (i.e., without previously open the spf in SStruct2016), then the analysis won't run (the message pops up and don't disappear even after 2 min).
Hence, if I want to run hundreds of analysis, do I really need to open all the *spf with SStruct2016 before run them with SeismoBatch?
Many thanks and best regards,
Romain

[z.gronti]

Dear Romain,

Could you please send us one of the *.spf files to check it?