Dynamic time history-results

[catherine_gb]

Hi everybody:)I think I have a problem with my results and I want to ask a couple of question because this is the first time I use Seismo struct and I have some difficulties with understanding how it works.I am doing Time history analysis for a 234m high chimney.I apply an accelerogram and then I increase it by the multiply curve factor.I have a probelm with the results- I have crush of concrete at the bottom section but only in 2 (out of 3 integration sections) so the program does not recognize this as a failure and goes on with the analysis and then the analysis stops.I tried to decrease th time step to 0,001 and 0,0005s but the same happens.I have the message fbd_ite. I tried to fix all option for the convergence and iterations.I use element loop convergence tolerance 0,0001 element loop maximum iteration 1000,maximum number of itration 2500 and displacement tolerance =0,001 rotation tolerance 0,001 to avoid convergence difficulties ...

[huffte]

Hi catherine. If you have not already done so, I would recommend a forum search on "fbd_ite". You'll get a few hits to which SeismoSoft has responded and some of these may prove helpful.

There is also information available in SeismoStruct Help for the same search, but it sounds as though you may have already made that search?

Best of luck catherine.

[catherine_gb]

Thank you very much huffte:) well I tried to find the answer in the forum but for now...nothing helped me:?

[seismosoft]

Catherine,

What do you mean exactly with "I have crush of concrete at the bottom section but only in 2 (out of 3 integration sections) so the program does not recognize this as a failure and goes on with the analysis and then the analysis stops"?

You, as the user, should set the failure criteria you believe to be adequate and take the decision on when the structure fails, not?

Seismosoft Suppport

[catherine_gb]

well.. I use FB elements and I select 3 integration sections for each element. I have defined the strains and the stresses in the nonlinear material models. Then I define in the performance criteria menu the limiting strains for fracture,yielding,crush in the concrete cover,crush in the concrete core and peak stress in the concrete.

And when the analysis is run as a result I can see that I have crush in 2 out of the 3 integration sections at the bottom element, The results show crush in sec(0) and the crush in sec(2) but no crush in sec (1). So the program does not accept this as a failure of the structure and continues running the analysis.

Then after some time it stops with a message that it cannot apply the next load step and either shows sudden fracture in some element or nothing -just stops.I tried to vary with the convergence properties- number of loops ,number of iterations etc...decrease the time step from 0,01 till 0,001 -but it is the same.

I was wondering since the structure is a chimney which is a SDOF system if i can accept this as a failure although the analysis continue after this happens. And I was wondering after the ultimate strains are reached why the analysis continues. For instance at a certain moment I have strains in the concrete increasing to -0,095 after(-0,0035). How this is possible.Is this connected with the theoretical background of the program:?

I want to thank you in advance

[seismosoft]

Dear Kate,

As discussed in previous posts in this Forum (e.g. see http://www.seismosoft.com/forum/topic.asp?TOPIC_ID=772), the Performance Criteria module in SeismoStruct serves only the purpose of aiding a user to monitor the response of the structure. It is not in anyway related to the analysis algorithm. Given that it is creating some confusion, we suggest you to simply disable all Performance Criteria that you have set; just cancel all entries in that module.

For what concerns the strains in the concrete, these can indeed reach very large values, to which, however, will correspond very small stresses (unless your confinement factor is very high).

Once you will have finished running the analysis (or when the analysis stops) without any Performance Criteria set, to avoid confusion, you should then extract all the results of such analysis (time-histories of strains, rotations, member shear forces, drifts, etc) and compare the values agains the limits that you wish to check (taken from seismic design codes/guidelines or from the literature) in order to establish the time/moment when a given limite state (serviceability, life safety, collapse, etc) has occurred.

You may also perhaps consult more literature on nonlinear finite element modelling (including some of the publications suggested in the Help System) for further guidance.

Seismosoft Support