absorbing boundary

[spater]

Thank you very much!
I have a question:
In this website:

https://opensees.berkeley.edu/wiki/inde ... e_Recorder

屏幕截图 2021-12-30 090023.png (28.61 KiB) Viewed 3194 times

and In this website:

https://opensees.berkeley.edu/wiki/inde ... on_Pattern

屏幕截图 2021-12-30 090519.png (54.3 KiB) Viewed 3194 times

Then my question is: if absorbingboundary is adopted, are the output velocity, acceleration, displacement and other results absolute or relative?
Looking forward to your reply, thank you!!!

[STKO Team]

Only the UniformExcitation generates relative outputs, because its keeps the base fixed, so your model will move relatively to the fixed base.

The MultiSupportExcitation instead generates absolute results, in fact you will see that the base moves as well.

When you use the absorbing boundary elements, you input a velocity field, and the absorbing element transforms it into a boundary traction (forces). So the output will be absolute as well.

[spater]

What if I want to know what the acceleration at the bottom is? Do I get the acceleration at the bottom directly by adding the velocity integral to the input or do I get the acceleration at the bottom from the calculation? Because I realized that the acceleration that I got by integrating the input velocity is not the same as the bottom acceleration that I got from the calculation.

[STKO Team]

The acceleration recorded at the bottom part of the soil (nodes shared between the soil and the bottom absorbing boundaries) should be exactly equal to the derivative of velocity with respect to time, not the integral...

Of course it is true if you set the material parameters of the absorbing layer properly, so that they match the material parameters of the soil

[spater]

I'm sorry, but acceleration should be given by taking a derivative with respect to velocity. My speed is only applied in the X direction. I extracted the acceleration in the X direction of a certain point at the bottom by STKO calculation, which is inconsistent with the acceleration obtained by taking the derivative of the input speed. In my model, the parameters of my sendboundary are exactly the same as those of my soil. May I ask how to solve this problem? Attached is my model. Could you please help me check what is the problem? Thank you very much.

[STKO Team]

I took the derivative of your velocity with respect to time to obtain the acceleration, and I put it as a background in your monitor at the base so that you can compare the numerical solution.

You will notice that at the beginning they match pretty well, while later on, you will see a slight difference. This happens because what you record at the bottom is both the input waves moving up and the waves reflected from the top surface going down.

It makes perfect sense.

onelayer_without-box.zip

(408.5 KiB) Downloaded 239 times

[spater]

Thank you for replying!!!
Because I only need to use the acceleration data of the bottom and the top when the seismic wave first travels from the bottom to the top to calculate the transfer function. So how do I extract data without considering seismic reflection?

[STKO Team]

Because I only need to use the acceleration data of the bottom and the top when the seismic wave first travels from the bottom to the top to calculate the transfer function. So how do I extract data without considering seismic reflection?

If you want to understand how a wave travels from bottom to top, you should do it with a simpler input. In such a way that the input is done before the wave gets reflected from the top free surface.
The problem with a long ground motion is that when the wave travels back to the bottom, the input is still creating new waves going in...

Try with a simple pulse or ricker wavelet as I did in our webinar.

Anyway, you don't need to record the acceleration at the bottom. If you only want the input you already have it (derivative of velocity)

[spater]

My research is focus on the wave propagate in the soil considering the SSI.

[STKO Team]

My research is focus on the wave propagate in the soil considering the SSI.

Ok!
As I said here:

Anyway, you don't need to record the acceleration at the bottom. If you only want the input you already have it (a derivative of velocity)

You can simply record the acceleration on top and compare it with the input acceleration obtained taking derivatives of your velocity time series (not the one recorded at the base)