error in element shear forces

[gon87]

Hi,
I have recently ran a time-history analysis with a fiber element building model, and observed that the shear force generated at the base of the elements (nearest gauss section) does not match by far the reaction at the corresponding supports.

I am using displacement-based elements without stress recovery, might that be the cause?

[seismosoft]

Dear G.Carvalho,

As discussed in the Help System, this is due to the fact that element shear forces are provided with reference to the element's local axis (directions 2 or 3), whilst support reactions are defined in the global coordinates system (directions X, Y).

Seismosoft Support

[gon87]

but all vertical elements are oriented on X and Y...
V3 matches with either X or Y reaction depending on the orientation, but V2 does not.

Could you please check if V2 is alright?
I can send a simple model for you to check.
Please forgive me if I am wrong.
Thank you,

[ricardomonteiro]

Hi Gonçalo,
Don't forget that the local axes for a deformed element are defined with respect to its deformed shape at the node of interest. Couldn't the difference you find come from that fact?

[seismosoft]

Gonçalo,

Just to complement Ricardo's response, we write to confirm that the fact that an element's local axis rotate with the element's chord is indeed the reason why a (heavily deformed) base element shear forces differ from the element base node horizontal reactions.

You can check this yourself by artificially stiffening the element (or decreasing loads), so that it does not deform much, and noticing that, in such case, the element shear will be similar to the support reactions.

We note again that the above is explicitly mentioned in the Help System (see chapter on 'Member Action Effects').

Seismosoft Support

[gon87]

Thank you Ricardo and Support,
it makes sense now, because the elements I checked were stiffer in the direction of V3, and clearly that's why I found differences between the two directions.
In fact I wasn't really familiarized with the co-rotational formulation discussed in the Help System and thought in a wrong way, but now I understand.
Thanks again and the best regards,

[seismosoft]

Dear Gonçalo,

There was actually a bug with the 'stress recovery' feature of SeismoStruct, that was leading to wrong shear forces in the V3 direction.

In the forthcoming release of SeismoStruct, this will have been corrected.

Thanks for signalling this problem to us.

Seismosoft Support