Seismosoft | Earthquake Engineering Solutions

Hello
I am examining how to model a shear wall against progressive Collapse. The wall has two boundary elements.
Since I should remove half the wall and one of the boundary element columns on the ground floor to simulate a progressive Collapse,
I split the wall into two parts and model the boundary elements of the column separately. The element is InfrmFB.
the column and the wall are connected on each floor with rigid Links.The purpose of the study is to examine the vertical displacement at the top of removed column under the vertical analysis.
If the wall section is divided into two parts, remove one part on the ground floor
Good results are obtained in comparison with finite element results. If it is divided into four parts, the two parts should be removed on the ground floor
results will be very different. What do you think is the best modeling method?

A typical infrmFB element has 4-5 integration sections. Using 4 such along the length of one member means 20 integration sections for the member, which is a lot.
Hence, probably you should divide it in 2 elements.
Further, note that subdividing the member to 2 or 4 elements can lead to different results, due to localisation issues, i.e. by specifying different locations and lengths of the inelastic regions, one somehow 'guides' the inelastic response of the member.
Regards,
Seismosoft Support

Thanks
1-I mean the division of the wall is not in height but in Wall width. My goal is to remove half of the wall in a single story in my model (similar to a opening).
So I can not use the rcrws section in the software.hence I split the wall web into two parts (in Wall width) and model the boundary elements of the column separately.
2- Using 2 or 4 wall divisions (in Wall width), gives different results!?Why?
3- Using the rcrws section or the rcrs section for the wall web With similar bars, gives different results!?? (in rcrws section: width of section edges= Very small amount). Why?
4- The result of rcrws section with two elements in width approaches the reference result( of DIANA).
regards.

Regarding the division in 2 or 4 elements, have you connected the wall segments with rigid links?
We cannot comment on the difference between the rcrws and rcrs results, since we do not know the details of your models. If the materials and the reinforcement are similar then the results should be similar too.
Regards,
Seismosoft Support

We checked your modesl, and there are obviously differences in the results, since there are significant differences between the two model (we managed to spot already differences in the reinforcing bars, and even in the nodes coordinates).
Note that, since the model is highly nonlinear and highly unstable – difficulties in convergence – very small differences in the input result in large differences in the results.
Seismosoft Support

Thank you very much.
By reducing the convegence norms and Modify the model, I got a more accurate result.
In the case of using 4 elements along the wall width, a good result was obtained.
Regards.