Modal analysis with tie-rods

[sergey.churilov]

Hello to all,

I am trying to investigate the influence of tie-rods on the modal properties of shear wall structure.

The comparison will be done for a structure with and without tie-rods.

The tie-rods shall have circular solid cross-section and shall connect two opposite walls. There shall be some tension force in it to simulate post-tensioning bolt with plate, similar to ancient masonry buildings with tie-rods.

The question is how to define properly tie-rods between two walls with some tension force within?

Sergey

[Artur.Kosakowski · Accepted solution]

Hi @sergey.churilov 

What about short bars with either shortening (dilatation) or temperature load applied?

Mind to use the right modal analysis type 

If I managed to answer your question(s) press the Accept as Solution button please. This will help other users to find solution(s) much faster. Thank you.

[sergey.churilov]

Hi Artur,

Thanks for your reply.

I did what you suggested before. In fact these bars are not short bars, but rather long, up to 6-7 m.

The dilatation option was found suitable for application of tension force.

However, there was not change in the predominant mode shapes nor frequencies of the example structure.

Only some change in the mode shape order and decrease in the local modes.

Therefore, I asked if there might be some other approach.

[Artur.Kosakowski]

Hi @sergey.churilov 

Have you 'converted' the case with self-weight and tension force in these bars intro modal analysis with the level of stresses being considered and compared it with the 'standard' one? IMHO the larger the tension forces are the more visible the difference in the eigenfrequencies should be.

If I managed to answer your question(s) press the Accept as Solution button please. This will help other users to find solution(s) much faster. Thank you.

[sergey.churilov]

First I had 2 load cased.

DL1 with self-weight only (static) and DL2 tension force applied through dilatation (static).

Only DL2 was converted to modal analysis taking into account static forces.

There were no changes in the eigenfrequencies.

Then I combined self-weight + tension force into one load case DL1 and then I converted it to modal analysis taking into account static forces.

Again there were no changes.

I am comparing these results to a case were only tie-rod is modeled without any tension force.

They are the same and we all know they shall not be equal.

Maybe I am doing something wrong here?

You can check the model attached to this post

[Artur.Kosakowski]

Hi @sergey.churilov 

It seems that the (standard) modal analysis (if you want to have them both) should be prior to the modal with the static forces.

If I managed to answer your question(s) press the Accept as Solution button please. This will help other users to find solution(s) much faster. Thank you.

[sergey.churilov]

Hi Artur,

It seems that there is no difference in this particular model.

The regular Modal analysis calculates the individual modes for the tie-rod + the modes for the whole model.

The Modal analysis recognizing static forces does the same and due to the influence of the tension force, the calculated modes are characteristic for the whole model.

Please compare mode#5 from Modal and mode#1 from DL1 (Modal analysis taking into account static forces). They are the same with respect to eigenfrequency, mass participation and eigenvectors (the shape).

Also, pairing of mode#9 from Modal with mode#2 from DL2 essentially gives the same outcome.

So, in my opinion, both analyses give the same results, while I was expecting more rigid behavior with a tie-rod with tension force.

[sergey.churilov]

Even a Modal analysis without tie-rod (only walls) gives the same values for the dominant mode shape. In this case it is mode#2 (f=2.19 Hz, T=0.46 sec, massUY=37.11%)

The first calculated mode, mode#1 has very low mass participation of 0.54%, so in this case is not considered dominant.

[Artur.Kosakowski]

Hi @sergey.churilov 

These are the results I'm seeing:

 

If I managed to answer your question(s) press the Accept as Solution button please. This will help other users to find solution(s) much faster. Thank you.

[sergey.churilov]

Indeed, but they are not the same modes.

In the first analysis mode#1 is local mode of the tie-rod, while in the second analysis mode#1 is global mode of the model. Check the mass participation also.

The problem is that with or without tension force, the global modes and their properties are the same in both analyses. Only the order of the calculated modes is different. In Modal analysis recognizing static forces, local modes are not present.

[Artur.Kosakowski]

Hi @sergey.churilov 

IMHO the results of the analyses are correct but your expectation that this rod makes a big change to the behavior  (eigenvalues) of the wall is too large. Compare  the results for a model when I applied tension to the walls directly.

If I managed to answer your question(s) press the Accept as Solution button please. This will help other users to find solution(s) much faster. Thank you.

[sergey.churilov]

Yes Artur, I think you are right.

Maybe I was expecting too great effect.

Thank you.