Modelling trapezoid plate in frame

[jo2ny]

Hello,

I have a 2D frame made out of 2 columns rectangular pipes. Top and bottom "beams" are 6mm thick plates.

Inside the frame i have a trapezoid shaped, cold formed, bent plate, 2mm thick.

On this model I will apply a horizontal linear force on the left column.

The bottom of the frame will be fixed.

What I want to highlight is how much does the trapezoid plate rise the in-plane rigidity of the frame.

I would like to take into account also the out of plane buckling of the plate.

So i have a model with the plate modeled as shell, and another model with the plate modeled as cladding or curtain wall.

The results however seem incorect for the model with shells. To me the frame containing shells seams much to ridgid compared to the one with claddings.

SO my question to you would be. Is this modeled correctly? 

Can I trust these results? The bar ratio of the columns changes from 7.87 (cladding model) to 0.25 (shell model); I don;t think that the trapezoid plate helps that much...

Here are the files, thank you for your time.

https://www.hightail.com/download/OGhkK2V0NEg5eFhFdzhUQw

[Rafal.Gaweda]

Please upload file again.Link is already dead.
Cladding in Robot has no stiffness at all

[jo2ny]

File uploaded again.

https://www.wetransfer.com/downloads/a6195735bbc1d8c22aa03c7c1838e2c620131218061814/977f82d15c614f55...

Thanks for your support.

[Rafal.Gaweda]

As I told you Curtain Wall has no stiffeness

So you have to use Shell