Bilin on BeamWithHinges

[kostast88]

I'd like incorporate Bilin material on a section, so that I use it in a fiber element. As of this example, my understanding is that we define everything in terms of Moment-Rotation and then aggregate this to a section. Fiber sections can be applied elsewhere on the beam:

eg
{
uniaxialMaterial Bilin 1 29598.246 0.1 0.1 232.572 -232.572 0.89 0.89 0.89 0.89 1 1 1 1 0.036 0.036 0.155 0.155 0.5 0.5 0.4 0.4 1 1
uniaxialMaterial Elastic 2 [expr $Es*$As]
section Aggregator 3 2 P 1 Mz;
WSection 4 $matID_Elastic 0.33 0.16 0.0115 0.0075 8 1 1 4 #IPE330

}

...in this case for Bilin, n=1, numbers obtained by Bilin Interactive database.
And then I apply this to a BeamWithHinges, a beam in a frame (with elastic interior to be consistent with Lignos theory).The beam is split in two parts, to create a middle node for a brace:
{
#element beamWithHinges $id $i $j $seci $Lpi $secj $Lpj $Es $A_beam $I_beam $transfTag_B
element beamWithHinges 1 1 2 3 0.3 4 0.3 $Es $A_beam $I_beam $transfTag_B
element beamWithHinges 2 2 3 4 0.3 3 0.3 $Es $A_beam $I_beam $transfTag_B
}

As of my understanding, this code is supposed to work well. While it does for static analyses, in time history, it is non convergent, and the results don't reproduce the Bilin backbone.

Has anyone tried to incorporate this material on beamwithhinges, with a fiber section next to it? Furthermore, Moment rotation OR Moment curvature should be used in this element (=moment rotation / Lpl). Lpl was used as the depth of the sections.

Any help is welcome.

[fmk]

moment-curvature is what is needed .. there is no reason any unixaial material cannot be used in the way you are using it .. if the stress-strain history is not what you expect it is either that the parameters you defined are incorrect or that there is a bug in the material (a distinct possibility with Bilin) . have you tries a similar material Bilin02 for example or the MultiLinear material (i no that material provides no degredation, but does it follow the stress-strain history ath you assume and does the model converge with it)

[kostast88]

Thanks for the reply!

Yes, when I switch the material I get what I want. What I don't understand is that there is a critical Lpl value, which if I exceed, the results get really ugly. It looks like the J section material, yields first, before the I section material on the element, and follows the pattern of the J material, but at some incorporates degradation (Bilin) from the I section material. It's really confusing.

About the calibration of Bilin, I started with initial values, and because I was after experimental results which are in my posession, I knew what the target was, and calibrated properly. The final calibration doesn't correspond to the initial results, if these were in terms of curvature. They rather look random (?!). I have knowledge of the Scott and Fenves papers.

So, since we need moment curvature, I suppose the values in the opensees example with Bilin and Beamwithhinges are wrong?

[fmk]

as i hinted at (bilin has problems) .. i hesitate to ask, but which example?

[kostast88]

http://opensees.berkeley.edu/wiki/index ... ibuted.tcl

The sections are defined in terms of moment-rotation, unless there's something goind on with this Lpl...
Here's the material / section definition:
http://opensees.berkeley.edu/wiki/index ... KModel.tcl

[thanp]

Hello,

Could somebody confirm that the example discussed above (http://opensees.berkeley.edu/wiki/index ... ment_Frame) is wrong, or shed some light on the modelling of the beamwithhinges element that is employed? To my understanding moment-curvature should have been used, but it "kind of" works with moment-rotation because an extremely small value was defined for the plastic hinge length. Any insights?

Thank you