About the isotropic hardening rules of Steel01/02

[ishinabe]

Hi

I'm confused about the isotropic hardening rules of Steel01/02(arguments a1,a2,a3,a4)
The rules in Steel01/02 are different from the equation ( 3.8 ) of EERC 83-19?
The equation in EERC 83-19 is "sigma_st / sigma_y = a1*(epsilon_max / epsilon_y - a2)" that seems to correspond literally to
the instruction of OpenSees User Manual "$a1 isotropic hardening parameter, increase of compression yield envelope as proportion of yield strength after a plastic strain of $a2*($Fy/E0).".

But the source code "Steel01.cpp" seems to implement the different rule from the EERC 83-19.
I can't understand the meaning of this implementation in the function "determineTrialState" ---> " TshiftN = 1 + a1*pow((TmaxStrain-TminStrain)/(2.0*a2*epsy),0.8 ); "

How can I learn the isotropic hardening rules implemented in Steel01/02???

Thanks.

[vesna]

If you do not understand the source code why don't you make a simple example on which you can test and see if the rules given on the command manual page are in agreement with the material models.

[ishinabe]

Thank you for your reply, vesna.

I have carried out some simple tests and understood the implementation roughly.

But the implementation seems not to accord the suggestion in EERC 83-19.

Where does the exponent "0.8" come from?

Regards.

[bridger]

I think it is not necessary to focus on the factor just like 0.8 herein.

[linguan118]

I think it's better to find it out in the source code.

[ishinabe]

Hello, bridger, inguan118.

I found the description <Steel01 -- bilinear steel model with linear kinematic and "exponential" isotropic hardening>
in the page "OpenSees Modeling Capabilities" ( http://opensees.berkeley.edu/OpenSees/u ... lities.php )

The implementations of Steel01/02 seem to have no relation with EERC83-19 as for isotropic hardening.....?

[linguan118]

The description may not cover every detail of the model.
Maybe you need to check the source code. I think the code is not too long for Steel01.

[ishinabe]

I have checked the source code "Steel01.cpp".
Then I think the source code seems to implement Ludwik's hardening equation (not the equation proposed by Dr. Filippou et al. in EERC83-19).
But the strain hardening exponent "0.8" is incomprehensible.
I think the value 0.2-0.3 is be applied generally for steel material.

[cxc08]

ishinabe wrote:
> I have checked the source code "Steel01.cpp".
> Then I think the source code seems to implement Ludwik's hardening equation
> (not the equation proposed by Dr. Filippou et al. in EERC83-19).
> But the strain hardening exponent "0.8" is incomprehensible.
> I think the value 0.2-0.3 is be applied generally for steel material.

Hi ishinabe-san, I am trying to understand the isotropic hardening rules of steel01. I want to use this material to simulate BRB by adopting the hardening rule proposed by Prof. Kasai. But it seems quite different from Kasai's definition.
I don't know how to determine the parameters a1 and a3. I wonder if you have known which hardening rule is used for Steel01? Could you give some suggestions for my question? Cannot thank you more for your help.

PS: As you said the source code looks similar with Ludwik's hardening equation, which is "sigm=sigmy+K*epsp^n". epsp is plastic strain. However in the source code, it is "sigm=sigmy*(1-b)*(1+a1*((maxStrain-minStrain)/(2*yieldstrain))^0.". They are same so maybe that's why the strain hardening exponent is 0.8 but not 0.2-0.3.