Cracked/Uncracked stiffness for Pushover

[apostolisandreou]

Hello,

seismobuild having by default uncracked stiffness in the advance setting dialog. For Pushover analysis what the correct way? Using cracked or uncracked stiffness?

Thanks

[huffte]

Uncracked stiffness should be fine it seems to me. During the pushover analysis, the reduction in stiffness as fibers reach limiting values would account for cracking in concrete, yielding of reinforcement, fracture of steel, etc.

It would be an interesting exercise to run t both ways - once with uncracked and once with cracked properties - to see any differences in the end result.

And this also depends on the element type you are using. I typically use infrmFB elements in pushover. My comments above are primarily informed on that basis.

[apostolisandreou]

I have already run with both ways.. the scenario with cracked stiffness is much more unfavorable.. the differences in the results are huge.. thats why i have to know if running with uncracked stiffness is correct or not.. and why seismobuild have by default the uncracked stiffness :/

[z.gronti]

Dear apostolisandreou,

Uncracked stiffness is used in Pushover analysis. Please refer to Section 4.4.1.4 of KANEPE for example, which states that cracked sections are to be used in elastic types of analysis.

[apostolisandreou]

Hello,

thanks for your answer.

I want to mention the 7.2.3 (2) of KAN.EPE. where this formula is proposed for the active stiffness in inelastic analysis. Seismobuild have this option to use MY/θy and the results are way different if you use this formula in comparison with uncracked stiffness. Using uncracked stiffness is way much favorable and i wander if this is safe.

Thanks

[seismosoft]

Section 7.2.3 of KANEPE refers to the elastic methods of analysis. In the Nonlinear Static Procedures the uncracked stiffness should be employed, in the execution of both eigenvalue and pushover analysis. Specifically, in pushover the analysis starts with the uncracked/undamaged members and the plastification of the members is done gradually as we push laterally the model.

Seismosoft Support

[fsapostolou]

According to KAN.EPE., all members have a quasi-linear M–θ diagram, so the initial stiffness up to the yield point is constant and corresponds to the effective stiffness (not the uncracked one), as stated in paragraph 7.2.3. Of course, this stiffness changes after yielding. However, I don't understand why the software uses the uncracked stiffness in step 0. Could it be that the program uses a different diagram — not linear up to the yield point — and calculates the stiffness at each step in more detail? Is it possible to see these calculations in the report?

Thanks in advance.

[seismosoft]

You may use the displacement-based plastic hinge model for a linear M-theta diagram. (Main Window>Analysis Parameters>Advanced Settings>Advanced Building Modelling)
The default setting for the frame elements is with fiber-modelling.
Seismosoft Support