Frist crack formation

[AntoMarinelli]

Dear STKO team,

I'm wondering to reproduce experimental results. I have modelled my beam with a shell and the bars as a truss and I have used the DamageTC3D. After setting properly the value according to references' ones and, for those that missed, international codes as Model Code 90 it seems that the STKO curve doesn't have the frist crack at the same value of the experimental one. The load that causes the first crack should be around 57 kN (evaluated manually) while the stiffness change takes place at more or less 150kN as it is shown in the picture. How is it possibile?

I attached the model and the chart that compares the experimental curve to the stko one.
Thanks in advance for the help.
Regards
Antonella

[marafini.f]

Dear Antonella,
how did you calculate the 57 kN? Can you post your calculations?
Francesca

[AntoMarinelli]

Dear Francesca,
the value actually is 53 kN (yesterday I made a typing mistake). The fctm considered is 3.

Thank

Antonella

[marafini.f]

Hi Antonella,

the moment you calculated corresponds to the first exit from the linear range of your structure, which in numerical terms can be equal to the moment of the first occurrence of damage. If you show material damage in the postprocessor you will see that it occurs more or less around step 14. If you check the displacement at that step and the corresponding reaction force you obtain a value of load around the one you calculated. So you should not get that change in stiffness at the value you calculated, as the pushover curve cannot represent everything in it.

As for the difference in results between STKO and the experimental response I suggest you play with the fracture energy. Analytical formulations for the calculations do have a margin of error so try to decrease of a percentage and see the result. Your analysis is quite fast so you can make many attempts.

Francesca

[AntoMarinelli]

Dear Francesca,
according to your suggestion I played with the fracture energy. The new curve is shown in the figure. Moreover, even I can catch the change of stiffness at more or less the same point, after that the software is not able to develop the non linear behaviour properly. What are the parameters on which I can make calibration?

Thanks in advance
Antonella

[STKO Team]

May I ask you how many rebars (and their diameters) you have in your experimental test?

[AntoMarinelli]

In this beam there are 8 phi 14 bars, with fy=506MPa e E=210 GPa. They are modelled as an equivalent bar (truss fem element) in the centroid of the layout bars.