Page 1 of 1

Consecutive Time History Analyses

Posted: 28 Sep 2010, 23:46
by ruthchek
Hello -

I am fairly new to the use of Seismostruct, and I'm trying to learn if it is possible to run consecutive time history records on my structure. I want to observe the change in response of the damaged structure. So far, I have only been able to accomplish this by inputting one ground motion record that is a compilation of several individual records.

However, this method has the limitation that my records are too long (too many total time steps). I know it's possible to change the time step of the input record, but if it's possible I would like to be able to run consecutive records using the "delay" option in the load curves.

Is this possible? If so, what can I do to resolve these two errors:
"Node Top is already loaded in the x direction. The last load in this direction is cancelled." (This error happens because I try to apply the second earthquake load curve to the same node that I applied the first record, in Applied Loading).

"The end of the time-history stages should be within the loading curves range"

Sorry for the long question. Thanks.


Re: Consecutive Time History Analyses

Posted: 30 Sep 2010, 10:36
by seismosoft
Hi,

Thanks for sending us the model in question, in a separate email, which allows us to now provide a more informed answer.

Indeed, it is not possible to apply two different records to the same node. You might want to explore the following possible workarounds:
- combine the two records into one, eliminating non-relevant data points (e.g. first 10 seconds or final 45 seconds of both records)
- define a second node with the same coordinates, that is then rigid-linked to the first, applying record 1 to the first node and record 2 to second one

We note also:
- by defining an analysis time-step (0.05 sec) that is larger than the record time-step (0.003906) you might lose important input information
- accelerograms should be applied at the base, rather than at the top of your structure

Seismosoft Support

Re: Consecutive Time History Analyses

Posted: 08 Dec 2015, 10:48
by Behnam
Hi
I'm trying to apply 4 earthquake records to my model in IDA analysis.apart from the point that was mentioned above regarding defining several nodes(in my case 4) with the same coordination ,can you please tell me what can i do about the time stages?since i get this error."Errors detected in Time-history Stages Module:The end of the time-history stages should be within the loading curves range"
the point is my records have different duration ranging from 10 to 23 seconds.
BR
Behnam

Re: Consecutive Time History Analyses

Posted: 14 Dec 2015, 11:37
by z.gronti
Dear Behnam,

The reported message informs you that the defined time-history stage is out of the loading curve range. As you informed us, you have inserted load curves of different duration. You should make all the loading curves of the same duration. That is to say, e.g. if the curve lasts 10sec (loading step 0.01sec) insert two more load steps one for 10.01sec with 0.0 acceleration and another one for 23 sec with 0.0 acceleration (with 23sec the largest duration of all the defined curves). In this way you will not get the reported message again and you will be able to run the analysis.

Re: Consecutive Time History Analyses

Posted: 28 Dec 2015, 20:57
by Behnam
Dear Zoi
Thank you.issue is solved.
the best solution in my case would be running 4 different files(as there are 4 records)with the same model.
by doing the analysis at the same time,the main advantage is that you make the most out of your CPU.