Strange Normal Stress

[lcasprini]

I'm making a time history of a simple column. The column is a infrmFB, and at the base is applied an accelerogram. The hysteretic curves are ok, but how is possible that the normal stress in the column change so much? What's the matter? I tried it in the settings, in the discretization of the column, but without result.

[huffte]

lcasprini,

Is the normal stress you speak of a combined axial load (P/A) plus flexure (Mc/I) normal stress? If so then it seems it would vary a great deal throughout the laterally applied acceleration history. Not only would it vary with time, but also with location on the cross section. Perhaps you might elaborate if the above is not the solution.

Best of luck, lcasprini.

[ruipinho]

Hi lcasprini,

If you carry out a search on "spurious axial" in the Help System/User Manual, you do find a couple of warnings related to the analysis of single piers.

Such unexpected axial load variations are related to (spurious) vertical vibration modes, which can be induced by: the elongation of the member (caused by the non-baricentrical neutral axis), the presence of inertia mass in the vertical direction, the "collateral effects" of equivalent viscous damping.

You may find additional discussion of this topic in:

Correia A.A., Almeida J.P., Pinho R. (2013) “Seismic energy dissipation in inelastic frames: understanding state-of-the-practice damping models,” Structural Engineering International, Issue 2013/2, pp. 148-158.


Best,

Rui

[lcasprini]

In the plastic phase of the column, the bending moment is coupled to the normal stress, then the hysteretic cycles are not acceptable??

Is there some trick to reduce the problem?

Thank You

[ruipinho]

Yes, if you follow some of the suggestions indicated in the paper I mentioned above (e.g. remove mass DOFs in the z-direction, introduce appropriate equivalent viscous damping, etc) you may indeed manage to eliminate such spurious models.

I note that, as you will have read in the Verification Report, SeismoStruct has in the past won a blind prediction contest involving the dynamic testing of a single bridge pier, hence the software is indeed able to correctly estimate the seismic response of such structural elements.

Best,

Rui

[lcasprini]

I can not find on the web a free version of the document that you mentioned, and then I have not read.

What do you mean practically to "remove mass DOFs in the z-direction" ??
What command should I enter?

Thanks

[ruipinho]

Hi lcasprini,

You should indeed be able to easily find the Verification Report of SeismoStruct, since it is installed with the program, and can be accessed in a myriad of ways:

- from the Help menu of SeismoStruct
- from SeismoStruct toolbar
- from the Start Menu of Windows ('All Programs' -> 'Seismosoft' -> 'SeismoStruct_v7')
- from Windows Explorer (C:\Program Files (x86)\SeismoSoft\SeismoStruct_v7\Manuals)
- from Seismosoft website (http://www.seismosoft.com/en/SStructDocumentation.aspx)


For what concerns the issue of defining mass DOFs, I suggest you to read the mass-related sections of the 'Element Classes' and 'Settings' chapters of SeismoStruct's User Manual or Help System (both of which are also installed with the program, and can be accessed in the very same way as the Verification Report).

Rui