diferent results for RC frame

[Violet.de]

Hi seismosoft.
As I said before in a previous topic, I want to run a dynamic time history on a RC frame which is a 3D-structures with one frame and cantilver slabs. My results did not match with experimental results. I also ran an Eigen value analysis and saw that the first period is 4 seconds!! But in experimental test, the first period was only 0.7 seconds.
I also check my loads and found that total weights of my loads on whole of structure in seismostruct is 120 tonn which is exactly the weight of the experimental test.
The time history curve also check. It is a curve from your verification report and I only copy it to my model and apply 9.81 m/s2 as load factor.
Could you please give me some suggession to match my model? I use infrFB element.

[huffte]

Hi Violet. Your strategy of running an eigenvalue analysis was a good one. The 4 second natural period shows that there is either too much mass or too little stiffness in your model. Double-check the settings to see if you are generating additional mass from loads. Check the rigid end joint offsets in the element connectivity. I think with a bit of effort you will be able to find something obvious. If not, feel free to send the model to me and I will look at it as I find the time. Best of luck, Violet.

[Violet.de]

Thank you. I sent my model to support@seismosoft.com

[Violet.de]

Dear Seismosoft,
My problem was solved. Have you ever check the cantilever slabs which are specified by building modeler?
I sent my model to your team. Please have a look at the slabs and specially rigid offset of the element. all of these values calculated by building modeler and seismostruct automatically and I spent a weak to find my problem and found that the problem was from the software calculations not for my inputs. I set zero values for rigid offsets, Edit the nodes in Z-directions (because they also had problem after converted the building modeler to seismostruct files). Infact, if you check my model and see the values of Z-direction for nodes, you can see the problem because the first story was 2.45 and other was 2.7 (floor to floor) in building modeler, but in seismostruct files, all of nodes need to edit!!
However, my problem was solved by these editions. I really confused why there is this bad bug in seismostruct file! The slab and modeling the slab had a lot of effects on my results. Only with edit it, the results is correct. In a new post, I say this problem again.
Thanks to Tim Huff, because of his attention to my model and his help to me for attention to rigid-offsets. Regards

[seismosoft]

Hello Violet,
It was quite surprising to read your post, since you are the first that reports similar bugs on the Building Modeller. The Building Modeller has been used by thousands of users in the past without problems, hence we feel that (without excluding the possibility of a bug) the chances are that the problem lies your modelling. For instance the issue regarding the z-coordinates that you reported as a 'bug' is in fact due to the size of the beam, i.e. the axes of the beam should be in the middle of the beams depth (at height 2.70-0.50/2=2.45m), rather than at the upper side of the floor (2.70m). The cantilever slabs are modelled only as additional loads on the adjacent beams, since they do not have any rigid diaphragm effect. Note also that if you set 'zero values for rigid offsets' you simply delete the rigid offsets, which is not correct.
Since we received from you only the SeismoStruct project (*.spf), can you please send us both the Building Modeller project (*.bmf) and the SeismoStruct project (*.spf) so that we check things?
Finally, one piece of advice. When stating on a web forum "I really confused why there is this bad bug in seismostruct file! ", you should be really sure about this bug. It is a really strong comment. Please read the manual and you will realise that there is no bug, but rather an incomplete understanding of what the program is actually doing.
Seismosoft Support

[Violet.de]

Dear seismosoft,
Sorry for my comment. I should be carefull about my post.
My model is 3D form of your verification report, chapter 4, RC-01 A.
I create this simple model by building modeler without any problem and extracte the seismostruct file. I ran this model exactly based on acceleration of your example, but it was not match. I sent to you this extracted model.
After, I changed two issues:
1- with replacing all of rigid ends with zero values ( they calculated by seismobuilding)
2- with changing all of zero values. In this case, all of nodes in each level has same Z (in extracted seismostruct file, they were different).
After running Dynamic time history, my redults were matched very well. So, if I did not want to verify a model, I did not think that my results may be have any problems.
If you run that acceleration to my model and check it with your experimental values, you see that it does not match. I thing the slop of input diaphragms to seismostruct, should be zero. Also, rigid end values are zero in your verification example, but in extracted seismostruct file, it was not zero.
Please attention that I only want to help that this software which is really is my favorite, will be better in the future.
Thanks alot for your attention and for good forum

[Violet.de]

In line 7, "Z" value is correct... not "zero values".

[seismosoft]

Note that the verification report files are included in the installation. You can open the original file and check the differences with your file.
Seismosoft Support