I have built a column model to do the Pushover analysis and the result (P-Δ curve) is good. However, I have a big question about the stop of that analysis. Specific speaking, the Pushover analysis can be done successfully and the final displacement is the right one which I design in the model. However, as we known, in the real experiment, the column structure will be destroyed already when the displacement approaches to a number and the experiment is over. So, I do not know how to find out when the structure will be destroyed and how much displacement I should design in my model can be approximate with the real experiment. In a word, the questions are, how does the OpenSees simulate the stop of the Pushover experiment and how to get the real displacement at the destroying point?
Thank you for your reading and I hope someone can help me to solve this problem.
The question about the stop of a Pushover analysis
Moderators: silvia, selimgunay, Moderators
Re: The question about the stop of a Pushover analysis
you have a numerical model .. OpenSees will either finish the Pushover analysis or fail if it reaches a point where
no solution can be obtained for your model .. how well it captures or fails to capture the failure in the experiment
is entirely dependent on how well you capture the failure mechanism in your model.
no solution can be obtained for your model .. how well it captures or fails to capture the failure in the experiment
is entirely dependent on how well you capture the failure mechanism in your model.
Re: The question about the stop of a Pushover analysis
Dear fmk:
Firstly, thank you very much for your answer. In my work, I have a real model for a experiment and the real experiment results (P-Δ curve). After the experiment, I built a numerical model in OpenSees, and then, I have faced the problem I told you above. In the experiment, the model is destroyed when the top displacement approaches 55 mm. However, when I use OpenSees model to simulate it, I cannot get the destroying point near the 55 mm, you know, the Pushover analysis can always be finished to the displacement which I designed, such as 80 mm, 100 mm, even 150mm,all of those results are wrong. So I begin to have the problem: what should I do in the numerical model so that the analysis can be stop near the real 55 mm instead of going on to the wrong displacements?
To that question, I have read some papers, but in almost all of them, I do not find the answer, they all just give the final result and the order source is almost the same as mine. So, I really do not what to do with it.
You said that it depend on how well I capture the failure mechanism in my model. Do you mean there are some other orders in OpenSees to control the failure mechanism? If I write that order to modify the failure mechanism in my model, will the problem be solved? If it will, and then, what those orders are?
Finally, thank you for your attention and your help again, and wish you a nice day.
Firstly, thank you very much for your answer. In my work, I have a real model for a experiment and the real experiment results (P-Δ curve). After the experiment, I built a numerical model in OpenSees, and then, I have faced the problem I told you above. In the experiment, the model is destroyed when the top displacement approaches 55 mm. However, when I use OpenSees model to simulate it, I cannot get the destroying point near the 55 mm, you know, the Pushover analysis can always be finished to the displacement which I designed, such as 80 mm, 100 mm, even 150mm,all of those results are wrong. So I begin to have the problem: what should I do in the numerical model so that the analysis can be stop near the real 55 mm instead of going on to the wrong displacements?
To that question, I have read some papers, but in almost all of them, I do not find the answer, they all just give the final result and the order source is almost the same as mine. So, I really do not what to do with it.
You said that it depend on how well I capture the failure mechanism in my model. Do you mean there are some other orders in OpenSees to control the failure mechanism? If I write that order to modify the failure mechanism in my model, will the problem be solved? If it will, and then, what those orders are?
Finally, thank you for your attention and your help again, and wish you a nice day.
Re: The question about the stop of a Pushover analysis
how close are the load displacement relationships between simulation and experiment up to failure point?