fbd_inv error with pinned beams

[jackrussel]

Hi. i have modeled two 3d braced buildings. first one with moment releases for beams and braces (IDA analysis). for moment releases i used capability of element connectivity. when i run the model it stops in the beginning with fbd_inv error. the other model was the same with no pinned beams but with pinned braces. it works without problem. i read explanations about fbd_inv in manual but it didnt help me.

[seismosoft]

Hi jackrussel,

The 'fbd_inv' flag message indicates the presence of a numerical instability in some of your elements.

Without seeing your particular model it is difficult for us to judge what might be happening, but, from your description, it does appear as if a condition of lack of static equilibrium (due to the release of DOFs) has arisen.

We suggest you to go step by step, releasing one DOF at a time, one element at a time, until you are able to pin-point exactly what is leading to this situation (and, in this way, find also the corresponding solution).

Best regards,

Seismosoft Support

[jackrussel]

Hi seismosoft.
I am wondered by releasing DOFs in element conectivity. it does'nt work when i use it in IDA! so i have modeled my pinned beams with links and the models work truely. what's the problem with this user friend capability??

[seismosoft]

jackrussel, hi.

Again, without seeing the models in question, we cannot really help much. If you send them to us (support@seismosoft.com), we will try to pay a look to this issue, and hopefully understand if the problem is with the software, or with the models.

Seismosoft Support

[jackrussel]

Hi,
I sent my file to <support@seismosoft.com>. thanks for your attention.

[seismosoft]

Hi jackrussel,

We have looked into your model, and noticed that by removing the (seemingly unnecessary) M3 end-releases, the model becomes numerically stable.

We also changed the convergence parameters, to improve things further:
Displ. Conv. Crit. = 0.001
Rot. Conv. Crit. = 0.001
El. Loop Conv. Tol. = 1.00E-04
El. Loop Max. Iter. = 1000

Hope this helps,

Seismosoft Support

PS: if we are not mistaken, there was also a beam that erroneously featured a beam cross-section..

[jackrussel]

Hi seismosoft,
Thanks for this answer but the model stops again. about 5.32s

[Stelios_Antoniou]

Can it be that it is a structural failure i.e. the structure cannot sustain the significant earthquake loading?

[jackrussel]

Hi Stelios_Antoniou,
I have multiplied 9.81 to the accelerogram and the masses were those that I get form the linear analyze in etabs2000.

[seismosoft]

jackrussel,

There must be another error somewhere in your model, since we were able to run the analysis until the end.

We will send you our file.

Seismosoft Support