VariableTransient clarification

For posts concerning the documentation, errors, ommissions, general comments, etc.

Moderators: silvia, selimgunay, Moderators

Post Reply
jgingery
Posts: 3
Joined: Tue Nov 19, 2013 10:32 pm
Location: University of California, San Diego

VariableTransient clarification

Post by jgingery » Tue Nov 19, 2013 10:49 pm

I'm using analysis VariableTransient for a case where the input dt=0.005. The recorded output time goes something like:

0.005
0.01
.
.
7.21
7.21333
7.21833
.
.

Is there a way for the output to return back to its original time step interval after the variable time step occurs. In other words, after the dt=0.00333 time step, can I return to the original interval of n*0.005, so that the next times are 7.215, 7.22,...?

For the time 7.21333 and thereafter, is Opensees linearly interpolating the input acceleration value between the values specified? If so I'm concerned that this could introduce some error/noise in the signal.

Thanks,
James Gingery
UCSD

fmk
Site Admin
Posts: 5866
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Re: VariableTransient clarification

Post by fmk » Wed Nov 27, 2013 10:38 am

OpenSees does linearly interpolate the time steps, which is why i don't use this option myself for eq records. i just use a regular analysis with time step as record or some multiple of.

as for you question. there is no way to jump back other than to issue the analyze command again. and you cannot monitor the workings of the analyze command while analyze is going on. the only alternative would be to mimic the VariableTrasinet calculations in Tcl code using "analyze 1 $dt" at every step, where you calculate the new $dt. there is a command, "testIter" which will return the num iterations in the last analysis to allow you to do this.

http://opensees.berkeley.edu/wiki/index.php/TestIter

Post Reply