EQ analysis time step

grahamo · Post by **grahamo** » Thu Dec 18, 2008 3:41 pm

I want to run the complete earthquake record from the PEER database. I am using the same set up as in Example 6 (ReadSMD) but I have a question about the Tmaxanalysis variable. When I specify 10 seconds with a time increment of .01, does that mean that only the first 10 seconds of the EQ record are used? Or does it run the whole record but breaks up the record in 1000 steps?

That has been a discussion that some of us have been having and it appears that the literature doesn't fully answer our question.

Thanks.

silvia · Post by **silvia** » Thu Dec 18, 2008 7:05 pm

i believe that in this proc the number of steps and the total time are defined independently.
therefore, the analyais runs only the first 10 seconds of the eq.

hope this clarifies.

Post by **fmk** » Fri Dec 19, 2008 6:12 am

the load patterns and the analsyis are totally separate .. you apply ground motions through a load pattern .. the input for the motion just consists of a series of points (mag and time for that mag) .. the load in the analysis is TOTALLY different ... when the analysis is being performed it just tells the domain what the current time is .. the domain tells the load patterns to apply the load for that current time .. UniformExciatation and MultiSupport just go get mag for that current time from their input list of points .. if time falls outside those defined 0.0 is returned, other linear interpretation is used.

so in your case only the first 10 seconds of the earthquake is run.