Absolute Acceleration vs. Relative Acceleration

[amb_1]

Dear STKO team:
Thank you so much for all the help you have provided, I have a new problem.
When using uniform seismic excitation of the base, I want to get the absolute acceleration of the site (e.g.,recorder Node -file node.txt -timeSeries 1 -node 1 -dof 1 accel), however the post-processing provides the relative acceleration (e.g., recorder Node -file node.txt - node 1 -time -dof 1 accel). Can you demonstrate how to get absolute acceleration using STKO? Or can you provide the python code to extract the absolute acceleration of the nodes at different locations of the site?Select nodes at different locations of the site to get the absolute acceleration.
Best wishes
ambcha

[STKO Team]

You can do it using the Python API but you need to provide the timeSeries you want to add, which is not available in the post processor.
Another way is to use, instead of the uniform excitation, the ground motion.
It will directly give you absolute values.

Just remember to remove the Fix in the direction of the input motion before applying the ground motion

[amb_1]

Dear STKO Team:
Thanks for all the help you've provided.. Can you provide a template utilizing the Python API with the model I provided?
The second approach uses ground motion instead of uniform excitation, which is a good approach and the most accurate. However, this approach needs to use velocity for loading and needs to calculate parameters for the three directions of the nodes at the bottom of the base, which is too much of a hassle compared to uniform excitation.
Because I need to calculate 20 models, if I use record command to record the absolute acceleration at different positions, I need to manually find the parallel computation (ID) where each node is located separately, which is too time-consuming.
Can you provide a program for recording? Python API preprocessing code for obtaining records, e.g. extract node accelerations along a column (A) of depth. After generating the mesh and exporting the calculation file. When I enter the number of the node along the depth column, it automatically generates a record of the acceleration tcl file, which automatically matches to the parallel calculation (ID) where the node along the depth is located. This way, when we modify the calculation file, it calculates the absolute acceleration of the node.
We are always looking for your help to have a code that can find the parallel computation (ID) where the node is located in the computation file generated by STKO, it can save a lot of time.
Best Wishes
Amabcha

[STKO Team]

However, this approach needs to use velocity for loading and needs to calculate parameters for the three directions of the nodes at the bottom of the base, which is too much of a hassle compared to uniform excitation.
Because I need to calculate 20 models

What I mean is the ground motion command. You can simply input the acceleration and it will impose it at the target nodes (on those nodes you have to remove the previous fixity in the X direction).
If you provide a smaller example of your model (for faster calculation) we will show you how to do it.
Next we will look at the other questions

[amb_1]

Dear STKO Team:
Thank you for the help you provided, I re-provided a new model, can you demonstrate how to extract the absolute acceleration of the site along the depth change A?

[amb_1]

Dear STKO team:
I've re-supplied a new model with periodic boundaries, can you help me modify the model, if it's convenient, I want to extract the absolute acceleration?
Thank you for your kind offer of help
Kind regards
ambcha

[STKO Team]

It's easier if you use the groundMotion+multiSupport pattern. It will automatically give you absolute accelerations.
Here's your file with the above-mentioned approach.

Sp_model_abs.zip

(260.13 KiB) Downloaded 8 times

I did these changes:

• I split the fix at the base into 2 conditions: 1 for Uy,Uz, and another for Ux (you will need to remove the fix Ux for the ground motion to work)
• Create a groundMotion condition at the base face
• use a removeConstraint patter (to remove fix Ux) before adding the multiSupport pattern
• Add the multiSupport pattern with the previously-defined ground motion instead of the UniformExcitation

The results will be the same as your original file, but with absolute accelerations.

Note:

groundMotion.zip

(3.48 KiB) Downloaded 7 times

If you are using U-P elements you have to copy this Python file (unzip it) in
C:\Program Files\STKO\external_solvers\opensees\conditions\Ground_Motion, replacing the existing one.
There was a small bug when used with U-P elements. This change will be available for everyone in the next STKO release

[amb_1]

Thank you for your help, the solution you provided is perfect for this problem! In addition, I have a question about the linux version of STKO, do you have a specific time frame for the new linux version?

[STKO Team]

It's ready and under testing now.
We are trying to release it as a snap package.

[amb_1]

Dear STKO team:
Thanks to the help you provided, I have been able to extract the absolute acceleration. However, a new problem has been encountered that
1. groundMotion+multiSupport pattern is compared with uniform excitation pattern, but the displacement, internal force, etc. of the site and structure obtained are not consistent, is there any other solution? For example, is it possible to provide python code in post-processing to convert relative acceleration to absolute acceleration, or to extract absolute acceleration in a uniform excitation pattern?
2, Secondly, can you provide a case study of how timeSeries records are added in analysis_steps? I tried several scenarios, however all models are not working properly.
3、At present, I can only use groundMotion+multiSupport pattern and uniform excitation pattern to calculate the acceleration and other responses such as structural internal force. Can you provide a solution?
Thank you for your kind offer of help
Kind regards
ambcha