Failure in pushover

02-Getting started with the modelling
Post Reply
zhwang
Posts: 39
Joined: 23 Jul 2014, 20:09

Failure in pushover

Post by zhwang »

Dear Seismosoft support,

I run pushover analysis on a multistorey bare steel frame, but it is unable to apply the entire phase load. I think this is not caused by convergence difficulties because I have changed that several times but it still does not work. I guess the analysis stopped because the failure of structure. I have already check the yielding of beams and columns which all happened before 1311mm, but I can not find why the program stopped at 1311.19727mm. I have also check the deformation and force of link element, but still do not get the answer. Could anyone suggest me how I can find which part of the frame fail in my pushover analysis?

The spf. file is attached: https://www.dropbox.com/s/8xralu44pla5t ... e.spf?dl=0

Thanks very much.
User avatar
z.gronti
Posts: 824
Joined: 16 Oct 2013, 08:14

Re: Failure in pushover

Post by z.gronti »

Dear zhwang,

You may apply performance criteria to check your model. For more information about the performance criteria you may read the User Manual and Help System of SeismoStruct. The following video might also help you
https://www.youtube.com/watch?v=0ZCDciuq4Gg
https://www.facebook.com/22301629122752 ... =1&theater
Zoi Gronti
Seismosoft Srl.
fakharifar_mostafa
Posts: 8
Joined: 12 Apr 2009, 17:30

Re: Failure in pushover

Post by fakharifar_mostafa »

I quickly went through your model. As you thought, your pushover analysis stopped at 1311.19727mm because of failure of your frame. If you just want to get your model running and push your frame up to 2 meter (REALLY HIGH!), just simply go to your material properties tab and change all your steel material fracture/buckling strain from 0.06 to 0.1. That solves the problem!

However, I want to pinpoint another issue, rather than just converging your model. Your pushover curve up to 1311 mm corresponds to 15.4% lateral drift. (considering your frame height to be 8500 mm). Such drift level is extremely high to obtain, and most likely unattainable for majority of structures, even those designed according to late seismic design codes. Local buckling and global instability already kicks in and you should see its effect reflected in the capacity curve (pushover curve). Your model uses a bilinear steel model with positive strain hardening, and no surprise your pushover curve is exhibiting a monotonically increasing post peak behavior, which again needs further consideration. I don't know your intention from this model, however wanted to raise this point in addition to converging your model. In SStruct fracture buckling strain is something important and may be adjusted with some caution. With this fracture/buckling strain you can even simulate where bond loss between reinforcing bars govern. Hope this helps and good luck.
Ana Santos
Posts: 24
Joined: 20 Apr 2015, 11:47

Re: Failure in pushover

Post by Ana Santos »

Hi everyone,

I am doing a response control pushover with a target displacement of 2m ( the design displacement is 0,78m but the EC8-1 stated that the pushover curve should be done to displacements between 0 and 150% of the design displacement).... the problem is when the structure reachs the displacement 0,788m the analysis stopped and says "unable to aplly the entire load phase"

What it's means?
huffte
Posts: 978
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: Failure in pushover

Post by huffte »

It could well be that the structure is not robust enough to withstand more than the 0,788 meter displacement. I wonder if you could do some simplified hand calculations to get a ball-park estimate of the real displacement capacity to see if the result makes sense?
Tim Huff
Post Reply

Return to “02-Getting started with the modelling”