Modelling issue

Forum for OpenSees users to post questions, comments, etc. on the use of the OpenSees interpreter, OpenSees.exe

Moderators: silvia, selimgunay, Moderators

Post Reply
ahmetalperparker
Posts: 90
Joined: Wed Oct 26, 2005 6:31 am
Location: Istanbul Technical University

Modelling issue

Post by ahmetalperparker » Wed Jul 28, 2021 6:52 pm

How can I model this in OpenSees?
Is it possible to apply large mass method in Ansys to OpenSees? For base excitation, you define a large mass and fix the to be restrained joints with constraints to the large mass joint. Then only apply earthquake acceleration to the large mass (large mass times ground acceleration in the only large mass degree of freedom). Is it possible in OpenSees? This method is sometimes used in Ansys and it has theorthical value.
In normal, you apply minus whole mass matrix times base acceleration on the right hand side of dynamic equation.
In this one, you define a big mass and the right hand side is plus large mass times base excitation and all other degree of freedoms related to masses are zero.
For referance in example you can look (maybe not a good reference but it tells very good)
https://www.eng-tips.com/viewthread.cfm?qid=97815

mhscott
Posts: 868
Joined: Tue Jul 06, 2004 3:38 pm
Location: Corvallis, Oregon USA
Contact:

Re: Modelling issue

Post by mhscott » Thu Jul 29, 2021 6:57 am

Interesting. Could you provide some figures describing the approach? The link is good, but a visual representation of the large mass method would help a lot.

mhscott
Posts: 868
Joined: Tue Jul 06, 2004 3:38 pm
Location: Corvallis, Oregon USA
Contact:

Re: Modelling issue

Post by mhscott » Thu Jul 29, 2021 8:24 am

It kinda sounds like the soil column with dashpot approach to inputting a ground acceleration.

See Figure 4 of this paper (http://dx.doi.org/10.1061/(ASCE)BE.1943-5592.0001353) and the references therein. It's fairly common in earthquake SSI.

ahmetalperparker
Posts: 90
Joined: Wed Oct 26, 2005 6:31 am
Location: Istanbul Technical University

Re: Modelling issue

Post by ahmetalperparker » Thu Jul 29, 2021 9:34 am

Please have a look at this one. This is better. It compares two approaches. I did this with multiple base excitation method with sap by defining a large mass to that joint, and apply the acceleration with multiplying by stiffness.
https://link.springer.com/article/10.10 ... 013-0941-x
But a proper new feature could be better.
Why I say this?
This is (large mass) a better approach. We with Edward L. Wilson and Jim Kelly applied this method to base isolationed system. It is physically the correct method. Also, when in base isolation, the normal approach cannot dissipate earthquake energy. It does not damp out. But with this approach you model the system physically correctly. So you get correct results for base isolation. The difference is at least 20%. In large mass method, think it as a fault. The normal method is mathematically correct but physically wrong. In normal method you apply accelerations to joints, not basement.
Can you implement this new large mass method?

mhscott
Posts: 868
Joined: Tue Jul 06, 2004 3:38 pm
Location: Corvallis, Oregon USA
Contact:

Re: Modelling issue

Post by mhscott » Thu Jul 29, 2021 11:42 am

Thanks!

This looks like something you can do pretty easily in an input script and does not require implementation in the core OpenSees.

Post Reply