topic Shell element connection to steel beam element in Robot Structural Analysis Forum

Shell element connection to steel beam element

jonas.ragelis — Wed, 07 Aug 2024 11:06:33 GMT

Hello,

Simply supported orthotropic shell panels are analyzed. One end is hinged supported by linear support, the other one is supported by HEB beam. In connection of panel to HEB linear release is used. Simply supported steel beam is hinge-connected.

I am facing a problem where a composite action occurs in connection between shell and HEB beam, resulting in smaller bending moment on beam element (comparing to hand-made calculations), as shell takes some bending moment itself.

Is there a way to prevent shell element from taking bending moment in direction of the beam?

It is also noticed that changing the bending stiffness of shell panel to ~0 (the one which is participating taking this tiny amount of bending from the beam) makes beam have the same bending moment as in hand calculations. But this change in stiffness matrix of shell element is not appropriate, as it ruins the results in further analysis (e.g. shapes of the modes in modal analysis).

Thank you.

Re: Shell element connection to steel beam element

CaioSaporito — Wed, 07 Aug 2024 17:42:24 GMT

Hi @jonas.ragelis.

Zip and share the model if you could.

Best regards,

Caio

Re: Shell element connection to steel beam element

jonas.ragelis — Thu, 08 Aug 2024 08:05:27 GMT

Hello @CaioSaporito,

Attaching the zipped model.

Thank you.

Re: Shell element connection to steel beam element

CaioSaporito — Thu, 08 Aug 2024 13:27:24 GMT

Hi @jonas.ragelis.

Check if it is what you want.

See attatched file

Best regards,

Caio

Re: Shell element connection to steel beam element

jonas.ragelis — Fri, 09 Aug 2024 10:21:01 GMT

Hello @CaioSaporito,

Really appreciate Your assistance.

I find out that your solution was very similar using "material orthotropy" to what I was trying to do with "user-defined" - basically it turns the bending stiffness of shell in one of the directions equal to 0.

I was worried about the forms of the modes in modal analysis using this technique, as the forms were not really what I was expecting (see. form of 2nd mode in 1 pic.) Anyways, it turns out that if in modal analysis parameters mass matrix is set to consistent, the forms of the modes seem to become somewhat more logical (see 2 pic.)

Pic. 1

Pic. 2