Free Field & Lysmer Kuhlemeyer Boundary Condition

[vokumus]

Hi

In ASDEA'S Youtube Channel, there is a video called "Compared simulation of absorbing boundaries in soil-structure interaction problems". I wonder if there is a reference for the last method shown "Free Field & Lysmer Kuhlemeyer Boundary Condition". It will be nice if there is a paper explaning how to model this boundary condition.

Regards.

[STKO Team]

Do you mean a research paper that describes that method, or a tutorial that explains how to do it in STKO?

[vokumus]

Both of them actually but only research paper will do the job too. I have read some papers but none of them explains it well. I wonder if there is a source explains it clearly.

Regards.

[STKO Team]

We will put a detailed tutorial on absorbing boundaries in the next few days,
In the meantime a good reference for the theory would the this FLAC Manual.

At page 20, it describes that they call "quite boundaries", that is only Lysmer-Kuhlemeyer Dashpots, while at page 21 it describes the Lysmer-Kuhlemeyer Dashpots + Free-Field boundaries.

[vokumus]

Awesome document. Thank you.

[vokumus]

Hi again;

I created a model which is a STKO free field implementation of OpenSees official site response analysis. When I fix Y direction at the base, results are ok and well correlated with official example.

I have wanted to use "Force From Reaction" option of STKO, but after the gravity analysis, system equilibrium is lost and system starts to move upwards. I have selected "loadconst" option, hoping to keep gravity loads active during the dynamic analysis but it didn't work. Maybe culprit is a simple thing but I couldn't find it after a couple of hours work.

I have attached my model. It would be great if someone helps me.

Best regards.

[STKO Team]

Hi,

We checked your model and there are a couple of notes that you can try to fix it:

• Your "Force From Reaction" is a load boundary condition. When you added it to a loadPattern in the analysis steps, you used a linear timeSeries, meaning that your loads will increase with time. Use a constant timeSeries to keep those values fixed.
• If you remove a fixity from a DOF and replace it with the corresponding reaction, is probably because you want to apply a L-K absorbing boundary there, but you didn't apply any value at the normal component of the LK boundaries under the soil and FF base lines. In the definition of you LK conditions, put a value in the normal component (Y) as you did for the tangential component(X), depending on the P-Wave velocity.

Let us know what happens when you do these changes.

[vokumus]

Thanks for your detailed answer. I'll modify my model soon.

Best regards.

[vokumus]

Hi again;

I have revised my model but there was still some uplift. Then I realized that my gravity analysis is transient. Making it static solved the problem. My model is all good now.

Thanks for your help.

[STKO Team]

We are glad you solved your problem. Yes if you do it as a transient, you have to make sure the time step is large enough to make it quasi static. However self weight in OpenSees is always applied all at once (i.e. without timeSeries) so the gravity analysis is always a problem if you do it transient. Otherwise, if you would like to apply self weigth incrementally, instead of specifing the body forces in the element definition, you can use the FaceForce (2D) or the VolumeForce(3D) in STKO, that will be translated as equivalent nodal loads in OpenSEES, and they can be used with timeSeries.