floor modelling

[danyisi]

Dear Seismostruct,
I have a problem with the eigenvalue analysis. I modeled a frame for experiments (made of aluminum 60x30x90cm high) in a span of three floors. I modeled a constrain to simulate the action of the diaphram plate of each floor and I put the masses concentrated at the four nodes of intersection between floor and pillars. By comparing the frequencies with those calculated with other computer codes, which give the same results, I got completely different results in Seismo ... How can I model the behavior of the aluminum plate? I also tried by placing the link constraints, or link element (together with diaphram) trying to manipulate the penalty function. What is the error? What is the solution?
Thanks

[seismosoft]

Are you sure the origin of the differences is the diaphragm modelling? In other words, have you modelled the same structure without slab on all different packages that you are using and compared the results?

Seismosoft Support

[danyisi]

Sorry, Seismo assistant, I was not very clear in my question. The real question becomes: how can I model in Seismo, a 3-story frame where the floor must have the same behavior of a shell

[seismosoft]

Dear Danyisi,

Whilst one may try to calibrate the penalty function multipliers (or the stiffness of equivalent horizontal braces) to try achieve an in-plane stiffness value similar to that of the slab, it is not necessarily granted that one will manage to reproduce in this manner the modelling with shell elements (especially if the shell flexural dofs have not been released).

In any case, and as mentioned before, the first inter-software comparisons should be carried out with a bare model not featuring any sort of slab modelling, so as to make sure that possible subsequent differences may be attributed to the slab modelling alone.

Seismosoft Support