Displacement discrepancy

[jon4son56206]

Dear STKO-Team

Currently, I am conducting a Time History Analysis (THA) on a complex building using STKO.
To validate the results, I have attempted to compare the displacement output at the base of the structure with that obtained from SeismoSignal. In theory, these values should be identical. However, I have observed a discrepancy between the outputs of SeismoSignal and STKO, as illustrated in the provided figure.

Displacements.jpg (85.33 KiB) Viewed 2216 times

To address this discrepancy, I have found that multiplying the STKO displacement output by a factor of 2 aligns it with the results obtained from SeismoSignal. To further verify the accuracy, I also attempted the reverse approach by extracting acceleration from the displacement outputs of both SeismoSignal and STKO, and comparing them with the input earthquake record. Once again, I found that applying a specific factor to the STKO results yields a matching record.

Accelerations.jpg (162.99 KiB) Viewed 2216 times

I attached the model of one part of the building. Can you find a reason for this discrepancy?
Thank's in advance!

Jonas

[STKO Team]

It's because you are using the Penalty Constraint Handler.

The penalty acts as a set of stiff springs to impose the constraint (in an approximate fashion).
In your model you first fixed Ux at 0, then you applied a ground motion in Ux.
With the Penalty handler, you now have 2 stiff springs, one will impose Ux=0 and another Ux=imposedMotion.
Since those springs have the same stiffness, the result will be the average (0+imposedMotion)/2.

So you have 2 options:

• Use the transformation handler
• Use the penalty, but before applying the ground motion, remove the fixity in X

In this specific example, you don't need the Penalty handler, you can go with the first option.
If you really need the penalty, also pay attention to the penalty value (1.0e18 may be too high in your model)