Link Elements

[ddonato]

Hi, my name is Diogo and im a student from University of Madeira, Portugal, im doing my thesis and its about pushover analysis vs dynamic incremental analysis. Im modeling the structure with elastic beams and columns with plastic hinges. for that im using link elements and im doing the modeling like this:

http://desmond.imageshack.us/Himg51/sca ... res=medium

But instead im getting a plastic response in the pushover analysis im getting a elastic response. So i guess im doing it wrong and i hope if someone can help me saying whats wrong in the modeling.

Im sending a link with the modeling of a column if someone want to see whats wrong with it please.

http://www.4shared.com/rar/w7RPL9Uf/Test.html

[huffte]

Hi Diogo. What sort of link are you using? I would imagine that either bl_kin (bi-linear kinematic hardening) or plst(plastic) with very large translational stiffness and realistic rotational stiffness is what you are using?

It seem sthat one of those should work if loaded beyond yield. This is more like a concentrated plasticity approach isn't it? If that's what you are shooting for then your approach seems valid. You may need to just double-check link type, dimensions, units, . . . .

Check out the Help system topics "Material Inelasticity", "Bilinear Kinematic Curve", "Link Elements", and "Plastic curve" for more detailed discussions of the above.

I hope this is some help Diogo. Best of luck on your work.

[ddonato]

Im using bl_sym link elements. looks like bl_kin system.
ty for your help, ill try do use that ones.

[ddonato]

SeismoStruct Help tells a little about the inelasticity model etc. but dont tell how to model the structure, and i think im doing something wrong in the modelling of the link elements.
Because the structure is in elastic response only. So seems that link elements are doing nothing there...

[seismosoft]

Dear Diogo,

We looked quickly into your model, and can report the following:

- your plastic hinges definition seem peculiar, with the exact same relationship being used for all 6 degrees-of-freedom (DOF); are you sure about this? Traditionally, one would have "infinite" stiffness in the translational DOFs (and possibly also on the torsional one), and then two different nonlinear relationships for the two bending DOFs

- in any case, if you increase the target displacement of your analysis (or decrease the yielding rotation of your hinges), you can get yielding


At this point, we would advise you to carry out a detailed literature survy on plastic hinge modelling, and/or to use the plastic hinge element of SeismoStuct.

Seismosoft Support