Seismosoft | Earthquake Engineering Solutions

Dear Seismosoft Team,

I tried to model one-bay one-story reinforced concrete frame in SeismoStruct utilizing Inelastic Force-Based Frame Element (inftmFB) with 5 integration points with cross-section discretized into 250 fibers. Concrete was modeled using Mander et al. (con_ma) model and reinforcing steel using Menegotto-Pinto (stl_mp) model. Material properties obtained from the sample specimens, just before the experiment, were assigned to the material models. Cyclic pushover was carried out using Static time history option, and the results are compared in the following figure.

https://drive.google.com/open?id=1GXpbx ... iCARy9OhJP

My numerical model cannot capture the unloading stiffness. What could be the reason for the "flat" hysteresis curve obtained from the mode as opposed to the "thin" hysteresis from the experiment?

What could be done to improve the accuracy of the numerical prediction? Or it is due to the fact that the fiber-based models cannot capture cyclic flexural-shear interaction and degradation?

Thank you for your time.

There are several reasons why you cannot reproduce the pinching behaviour from the experimental results.
One reason could be confinement, please check that you are using the correct confinement factors (i.e. you have modelled correctly the shear reinforcement).
Note that the fact that the fiber-based models cannot capture cyclic flexural-shear interaction and degradation place a role in large shear walls where large shear deformations are expected to occur. (and in these cases the differences are not so pronounced as in your case).

Seismosoft Support

Dear Seismosoft Team,

Thank you for the prompt response.

I have re-verified that I have correctly modeled the shear reinforcements in the model but the problem still persists.

What could be the other reasons why I cannot reproduce the pinching behavior?

Best Regards,
Naresh

Could it be a large concrete tensile strength? Or a large strain hardening value for steel?
Seismosoft Support

Dear SeismoSoft Team,

I have tried to figure out the modeling issue by using alternative material models, compression only concrete, and different element types; but the problem still persists.

Thank you for your help.

We would suggest that you take a look at similar experimental tests as described in the Verification Manual of SeismoStruct. Note that all the relevant input files are included in the setup of the programs, hence you will be able to see the modelling followed in these tests.

Seismosoft Support