Learning Forum

[AR]

Hello,

I need to obtain the amplification function or the frequency response function for a seismic event but I don't know how. Does anyone know how to do it?

In the image below you can see a graph that compares the acceleration in a specific node against the acceleration entered. Thus, the amplification would be obtained (dimensionless since it is m/s^2 divided by m/s^2). However, I don't know what kind of analysis I can perform to get these values. I think a Harmonic Response should be done. However, I get no logical results. Does anyone know how to do it?

 

Regards

[peteroznewman]

A Harmonic Response is the correct analysis to obtain the acceleration Amplification factor from the Frequency Response result.  It is usually faster to use a Modal analysis and drop a Harmonic Response analysis on the Solution cell of Modal to create a mode superposition harmonic response analysis.

A Frequency Response Function (FRF) is obtained from a sweep of frequencies of a harmonic (sinusoidal) input vibration. In the figure you show, it looks like one of the modes is near 25 Hz.

A seismic vibration input to the base of a structure is not a harmonic input so you don’t get an FRF output. You get what you show in your other thread.

[AR]

Hi Peter,

I understand what you are telling me. However, as I said before, the amplification shown is dimensionless since it divides one acceleration value by another.
As far as I know, you have to divide the values obtained at a point of the structure by the values obtained at the base.

Amplification factor = values of the point of interest / values of the reference point

In the cases in which this test is carried out with a vibrating table, this last value (reference point) is taken from an accelerometer located on the table where the structure is anchored. However, if I take as a reference point a node of the model that is anchored to the ground, the values of the acceleration of that node will be equal to 0 (or almost 0), making the result of the division not suitable.

Any ideas on how to do it properly in Ansys?

[peteroznewman]

A Harmonic Response analysis can be configured to have a Base Excitation. The means that all the Fixed Supports will have an acceleration value applied to them, overriding the Fixed Support. If the input acceleration is 1g, that will be the denominator for the FRF, and the acceleration at any other point will be divided by 1.

[AR]

Ok, thank you for your answer

The truth is that I tried what you mention about Base Excitation. However, the model would stay for hours at 20% with the Modal Expansion (it was tied to a previous modal analysis) and would not continue calculating.

Then I tried a much simpler and smaller model. With this model the Base Excitation did work. However, I had no actual graphs for that model with which to compare the FEM model to see if it was correct.