Modeling a cantilver RC column

[speedy]

Hi seismosoft,
I want to create a RC cantilver column which was published in an experimental paper. I create the model based on the paper exactly, but my results did not match to experimental results. My hysteresis results were bigger and the hysteresis was to fat than the experimental hysteresis. Because of this, I decided to add a link element at the base of the column. I used a calibrated rotational spring for M2 in this link. For other DOF (F1,F2,F3,M1,M3), I set 1E+010 (KN.m).
The new results were much better and slimmer that the previous results, but it still didn't match to experimental. The different were very clear and I don't know how could solve this problem. Could you please give me an idea? the type of the analysis was "static time history".

[seismosoft]

Adding arbitrarily a link element at the base of the column is probably not the way try to simulate the experimental behaviour. You need to identify the reason for the difference between the test and the analytical results. For instance, some things that might be different are (i) confinement and consequently the shape of the concrete hysteretic curve (ii) shear deformations that are not accounted for in the analysis (iii) Properties of the materials used (iv) loading, both the lateral but also the gravity loads (e.g. are the axial loads the same - have you measured the support forces?)
In any case, this is a technical issue that is probably better discussed with your supervisor, rather than the support forum of the program you chose to use.
Regards,
Seismosoft Support

[speedy]

Thank you for your response. I will try to use your comments to solve my problems in the model. I would be say that I had similar problems (fat hysteresis in analysis than experimental) in a RC beam. I think that some models like these, need to considere a specific property to show a slim hysteresis. But I don't know what is that property.

[speedy]

The confibement factor was automaticaly calculated by the sesismistruct. You think that I should change the calculated value?

[seismosoft]

The confinement factor is automatically calculated according to appropriate expressions. It is up to you to calculate it by yourself, however note to consider the correct reinforcement (both transverse and longitudinal)

Seismosoft Support

[speedy]

Thank you.
I see your verification report chapter4, blind-01 and bridge-02.
They give me several idea for modeling RC cantilver column, but I have questions. I would be thank if you answer them:
1- two methods used in these examples. For blind-01, infrmFB used, but in bridge-02, infrmDB. As it told in the manual text of the seismostruct, infrmDB used for short members. Do you think that I should be use infrmDB (my RC column has 1.2 m in hieght and 900x450 mm cross section).
2- In both examples, a mass element used in top of the column. I really confused that why they didn't use a vertical permanent liad instead. All of the above weights which were carried by the column in some FE models in several papers which I read, used the vertical load, not any lumed mass. Could you please give me more detailes for why use this method? Do you suggest to me that I use lumed mass instead vertical load?
3- In the example bridge-02, a elfrm element used in bottom. I couldn't calculate the stiffeness which were used for defining this element in your model. Also, I read the reference, but there is no details about these values. I would be happy if you help me to calculate these values. My experimental model has been placed in concrete foundation and I want to use an elfrm element for that, but I do not know how should calculate its stiffeness.
4- Do you think that I should be consider the method of blind-01 without any elfrm, or bridge-02 with elfrm in the base?
Thanks for your attention

[seismosoft]

Regarding the selection of the frame elements, refer to the Users Manual and the help of the program. You will find all the details you need to understand for the selection of the frame element types for your model. One cannot explain these details in a forum post. If you have specific questions on the different elements classes, we will happy to answer to them.
Further, the masses applied on the models can be transformed to vertical loads (and vice versa). Refer to the Mass & Gravity tab of the Project Settings (and the relevant help) for further details.
Regarding your modelling questions, as we explained in one of our previous posts, these are issues that are better discussed with your supervisor. We cannot provide advise on the modelling methods that you would want to apply, since we are unaware of what we are trying to simulate. The different examples in the verification report are provided as a tool to help users, however we cannot really provide meaningful help to users who have not read and learnt the difference between the force-based and the displacement-based classes for instance.
Please try to be more specific, and try to post after you have read the corresponding help files provided with the program. Most of the questions you make are general modelling questions that cannot be addressed inside this forum, which is a forum for the use of Seismosoft's programs, rather than a general forum on FE or earthquake engineering.

Seismosoft Support

[speedy]

Thank you. But I would like to know that why the seismosoft used two methods for these two similar examples in verification reports. Thanks again for your comments.

[seismosoft]

They are equivalent ways to model frame elements with different element types. The different examples in the verification report are provided for display purposes, and to show that they two frame modelling types are equally accurate.

Seismosoft Support