Applying displacement and absorbing boundary condition

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

Moderators: silvia, selimgunay, Moderators

Post Reply
sanazzz
Posts: 16
Joined: Sun Oct 23, 2016 6:32 pm

Applying displacement and absorbing boundary condition

Post by sanazzz »

Hi,

As far as I understood, when you apply displacement, ( SP or imposedgroundmotion) OpenSees extrapolate the displacement of the node after SP to be zero which causes the dashpots assigned to the note not to work. Is there anyway to resolve this problem?
skamalzare
Posts: 112
Joined: Thu Jun 27, 2013 11:45 am
Location: Seattle, WA

Re: Applying displacement and absorbing boundary condition

Post by skamalzare »

Sanazzz,

To define absorbing boundary conditions, you need to follow these steps:

1. Remove all the constraints from your base (i.e. both x and y).
2. Tie all the base nodes to one of them in both x and y directions (i.e. equalDOF Node1 Node2 1 2 - Doesn't matter which one. Take one as the master and tie the rest to it.)
2. Define two sets of Pseudo nodes at place of each node at the base.
3. Fix Pseudo node set 1 in x and y.
4. Fix Pseudo node set 2 in y.
5. Tie Pseudo node set 2 to base nodes in both x and y.
6. Make a viscous material with the dashpot properties, and put it into a zero-length element. The zero length element should contain pseudo nodes.
7. Apply the earthquake to the pseudo nodes which are free to move in x, in form of load-time history (i.e. rho_base*Vs_base*AreaofNode_base*Velocity_TimeHistory).

There is an example in the Wiki, you can look through it. However, I prefer the above approach. In the Wiki example all the loads will be applied to one node, whereas in this approach loads will be distributed to all the nodes.

Bests,
Soheil
---
PhD, EIT, Geotechnical Engineer
Condon-Johnson & Associates INC
sanazzz
Posts: 16
Joined: Sun Oct 23, 2016 6:32 pm

Re: Applying displacement and absorbing boundary condition

Post by sanazzz »

Thanks Soheil,

I had a look on this example and thanks for the steps. It is fine. My main question is that how to apply displacement instead of load in the last step?
skamalzare
Posts: 112
Joined: Thu Jun 27, 2013 11:45 am
Location: Seattle, WA

Re: Applying displacement and absorbing boundary condition

Post by skamalzare »

You can apply displacements to your model by using "Multi-Support Excitation" command. Just be careful, if you are applying the displacements to a fixed node, your displacements will be the average of zero and the input displacement (double check this as I am not sure).
---
PhD, EIT, Geotechnical Engineer
Condon-Johnson & Associates INC
sanazzz
Posts: 16
Joined: Sun Oct 23, 2016 6:32 pm

Re: Applying displacement and absorbing boundary condition

Post by sanazzz »

That's the main problem. The command that you mentioned work as SP which forces the node to have zero displacement after end of the excitation which I do not want. I want to apply dashpots to the nodes to capture reflected waves.
skamalzare
Posts: 112
Joined: Thu Jun 27, 2013 11:45 am
Location: Seattle, WA

Re: Applying displacement and absorbing boundary condition

Post by skamalzare »

Sanazzz,

1) Are your base nodes fixed? I expect them to be free to move.

2) My understanding is that the "Mulit-Support Excitation" command applies external displacements to your system. Accordingly, if your earthquake stops, there will be no external displacements added to the system. But the system should be able to move freely due to the waves that are inside it or any unbalanced displacements.

If your base nodes are not fixed, I would expect them continue moving (maybe just a little, depending on many factors) after the motion stops till all the waves are damped out of the system.

Soheil
---
PhD, EIT, Geotechnical Engineer
Condon-Johnson & Associates INC
sanazzz
Posts: 16
Joined: Sun Oct 23, 2016 6:32 pm

Re: Applying displacement and absorbing boundary condition

Post by sanazzz »

No, I do not fix them. But, the problem is the way MultiSupport command is defined. Look at this

"in OpenSees, the prescribed motion is applied using single-point constraints" !
this command forces the node to be zero after the end of excitation which is not desired.
Thanks for your replies.


Bests
Post Reply