Good day
I'm a new user in seismostruct
my question is about the dynamic time history analysis; if we have multiple accelerograms ( more than one for example a 15 acceleregrams ) already matched by seismomatch software which gives us a mean spectrum, the question is how should I apply in seismostruct this accelerograms matched (in the two steps; time-history curve and applied loads)
should i use load or load multiple or create in seismostruct
seismomatch gives us à mean spectrum but seismostrcut in dynamic time history applies (works with) accelerograms
how I can apply this mean spectrum to seismostruct (in the two steps time-history curve and applied loads)
I thank you very much in advance,
multiple accelerograms in dynamic time history analysis
Re: multiple accelerograms in dynamic time history analysis
If you simply want to do a response spectrum analysis in SeismoStruct, you can manually define the mean spectrum desired.
If you want to do a response history analysis in SeismoStruct, you need to save the matched versions of the accelerograms. You can define the 15 time history curves in the SeismoStruct file in the "Time History Curves" tab.
You can then apply one of the accelerograms in the "Applied Loads" tab. Here you will need to apply the appropriate factor. For example, if the accelerograms are in units of g's and the SeismoStruct acceleration unit is m/s/s, the factor to apply would be 9.81.
You can have 15 different versions of the same model in 15 files. The only difference in the 15 files is the curve specified in the "Applied Loads" tab.
You can then use SeismoBatch to perform the 15 analyses and extract the desired results.
If you want to do a response history analysis in SeismoStruct, you need to save the matched versions of the accelerograms. You can define the 15 time history curves in the SeismoStruct file in the "Time History Curves" tab.
You can then apply one of the accelerograms in the "Applied Loads" tab. Here you will need to apply the appropriate factor. For example, if the accelerograms are in units of g's and the SeismoStruct acceleration unit is m/s/s, the factor to apply would be 9.81.
You can have 15 different versions of the same model in 15 files. The only difference in the 15 files is the curve specified in the "Applied Loads" tab.
You can then use SeismoBatch to perform the 15 analyses and extract the desired results.
Tim Huff
Re: multiple accelerograms in dynamic time history analysis
thank you, your answer is very clear, i have followed all these steps
In regards to the results (displacements, acceleration,.....etc), when i use seismobatch i get 15 files , will there be 15 results related to each one for this 15 acceleregrams so 15 displacements ,15 acceleration,15 velocity .....etc Or be one result under the influence of this15 acceleregrams ?
Because I saw in scientific articles that it has one result; 1 acceleregrams so 1 displacements ,1 acceleration,1 velocity .....etc under a multiple accelerograms, How is that ?
with dynamic time history analysis.
Thanks.
In regards to the results (displacements, acceleration,.....etc), when i use seismobatch i get 15 files , will there be 15 results related to each one for this 15 acceleregrams so 15 displacements ,15 acceleration,15 velocity .....etc Or be one result under the influence of this15 acceleregrams ?
Because I saw in scientific articles that it has one result; 1 acceleregrams so 1 displacements ,1 acceleration,1 velocity .....etc under a multiple accelerograms, How is that ?
with dynamic time history analysis.
Thanks.
Re: multiple accelerograms in dynamic time history analysis
Well, it is not clear what you hope to accomplish. This will be a decision you have to make, depending on the purpose of your study.
If you want to match a single accelerogram to a target spectrum, then I don't understand why you are using 15 accelerograms to begin with.
A commonly encountered goal might be to take the average response (geometric mean) of the 15 results to come up with a single value for displacement, acceleration, drift, etc. It is also commonly desired to obtain a measure of response variability by computing the standard deviation (typically log-based) of the 15 results. But, if you are spectrally matching, as opposed to amplitude scaling, then you have likely artificially reduced the variability. So that may not have much meaning for your case.
So, I hope my explanations help you decide what to do. Again, it depends on what you wish to accomplish. Best of luck in your work.
If you want to match a single accelerogram to a target spectrum, then I don't understand why you are using 15 accelerograms to begin with.
A commonly encountered goal might be to take the average response (geometric mean) of the 15 results to come up with a single value for displacement, acceleration, drift, etc. It is also commonly desired to obtain a measure of response variability by computing the standard deviation (typically log-based) of the 15 results. But, if you are spectrally matching, as opposed to amplitude scaling, then you have likely artificially reduced the variability. So that may not have much meaning for your case.
So, I hope my explanations help you decide what to do. Again, it depends on what you wish to accomplish. Best of luck in your work.
Tim Huff
Re: multiple accelerograms in dynamic time history analysis
ok, Thanks for the response