Dear Seismosoft,
in order to simulate the frequency dependent compliance of soil-foundation systems in time-domain soil-structure interaction analysis, I need the definition of fully coupled stiffness, damping and mass matrices. Until now, I have used LPMs with diagonal stiffness, damping and mass matrices but now I'm trying to account for more complex frequency dependent foundation behaviours that I have approximated by identifying a hidden state variable model.
Is there a way to simply allow a user to define fully coupled stiffness, mass and damping matrices?
Thanks in advance for Your attention.
Fully coupled stiffness, damping and mass matrices
-
- Posts: 7
- Joined: 31 Jan 2014, 12:26
-
- Posts: 6
- Joined: 02 Oct 2010, 12:42
Re: Fully coupled stiffness, damping and mass matrices
Dear colleague,
From your question I gather that you are trying to use lumped parameter models (LPMs) to account for the frequency-dependence of foundation impedances in the linear elastic range. Unfortunately, the lumped elements in SeismoStruct (link, mass and dashpot elements) have an uncoupled response in each direction. You cannot define the fully coupled matrices that your model requires.
Nevertheless, you may sometimes use a set of eccentric elements, connected by rigid links, in order to simulate a coupled matrix. If you want more information on that front, please contact me directly.
On the other hand, a future version of SeismoStruct will have foundation macro-elements available (which, by the way, make use of coupled, although frequency-independent, stiffness matrices). These are more adequate for structures responding in the nonlinear range, where those LPMs may not be so appropriate.
Best regards,
António A. Correia
From your question I gather that you are trying to use lumped parameter models (LPMs) to account for the frequency-dependence of foundation impedances in the linear elastic range. Unfortunately, the lumped elements in SeismoStruct (link, mass and dashpot elements) have an uncoupled response in each direction. You cannot define the fully coupled matrices that your model requires.
Nevertheless, you may sometimes use a set of eccentric elements, connected by rigid links, in order to simulate a coupled matrix. If you want more information on that front, please contact me directly.
On the other hand, a future version of SeismoStruct will have foundation macro-elements available (which, by the way, make use of coupled, although frequency-independent, stiffness matrices). These are more adequate for structures responding in the nonlinear range, where those LPMs may not be so appropriate.
Best regards,
António A. Correia
-
- Posts: 7
- Joined: 31 Jan 2014, 12:26
Re: Fully coupled stiffness, damping and mass matrices
Thank you Antonio for your reply. I have already used the tecnique you suggest but unfortunately, for the approximation of dynamic soil-foundation behaviour characterised by two or more peaks of resonances, I didn't find it so versatile.
Thanks also for information about future release.
S. Carbonari
Thanks also for information about future release.
S. Carbonari