Orthotropic Stiffness Coefficients Panel Instability

[Anonymous]

Hi all,

Modelling a 3D building I'm trying to artifically reduce the stiffness of the shell elements that model the floor slabs to ensure minimal framing action and that the stability loads are transferred to the core. To do this when defining panel properties i'm selecting an orthotropic model and then in the geometrical parameters modifying the coefficients. In some cases I am editing the f (membrane) and others the k (flexural) stiffnesses. Having entered this data when i then click on display button to check the stiffness matrices the values are correct for the factors. However I'm unsure they're working correctly and when i try and check them in a test model i have instability problems which gives me inappropriate results.

I have 5 wall panels all the same support conditions and the same thickness but with modified material or panel properties. The first one is 250thk homogenous panel with the material as default concrete, the second is a homogenous panel with a different material i've created with half the E-value (Youngs Modulus) of the concrete, the third has a "reduction of the moment of inertia" of 0.5 * Ig, the fourth is an orthotropic panel with the coefficients of flexural (k) stiffness all set as 0.5, and the fifth is an orthotropic panel with the coefficietns of membrane (f) stiffness all set as 0.5. I am applying test loads both in-plane (nodal force at the top) and out-of-plane (uniform pressure) but when i run the model i have instability type 1 problems for most of the mesh nodes in the last two panels which then give me ridiculous deflections as i'd be expecting the deflections to be double if i'm halving the flexural stiffness?

To create all the models i initially created one panel and then copied 5 times to form the basis of each wall panel before i then changed the material properties only. Why am i having instability problems with the final two wall panels only where the only thing i have modified are the stiffness properties? The model is a 3d type and the support conditions are pinned at the base and fixed in the X-direction (out-of-plane to the wall) at the top only.

Jpeg below (and attached) of problem with screenshots. Thanks for any help.

Attached is a pdf to show the main information. All panels are

[Rafal.Gaweda]

PLease send us this test rtd file

[Rafal.Gaweda · Accepted solution]

However I'm unsure they're working correctly and when i try and check them in a test model i have instability problems which gives me inappropriate results.

Remark 1

Orthotropy direction can not be perpendicular to panel plane.

Use Automatic or Dir Z

[Rafal.Gaweda]

Remark 2: after correcting model as mentioned in  remark 1 the results seem to be OK.

[mgaafar]

Rafal,

        I wanted to model a wall in a concrete building that I don't want it to attract any shear forces while keeping the wall load bearing. So I tried to use  Orthotropic material with f12 modified to 0.001 so the wall in-plane shear stiffeness is reduced. I tried it with ETABS and it works but changing it in ROBOT does not even change the deflection of the wall.

what am i doing wrong? and how come lowering f12 does not affect the wall deflection.

please see attached file. 

Happy new year 🙂

[Rafacascudo]

mgaafar ,

That´s a good point ! I tried here and came to the same results as you.

For an isostatic short wall , as in your example , the forces horizontal would have to the base no matter how tiny the 12 shear stiffness was . But for sure the displacements would have to be greater.

As an alternative while we wait an official word on this , on a real model with many walls connected you can apply shear linear releases (Ux) to the top edge of your wall. This will prevent this wall from getting horizontal in plane shear forces.

I made some modifications to your test model .

[mgaafar]

Thanks Rafacascudo, I actually did released the edges for shear for now but I hope RSA team can look into it and see.

[Artur.Kosakowski]

        I wanted to model a wall in a concrete building that I don't want it to attract any shear forces while keeping the wall load bearing. So I tried to use  Orthotropic material with f12 modified to 0.001 so the wall in-plane shear stiffeness is reduced. I tried it with ETABS and it works but changing it in ROBOT does not even change the deflection of the wall.

what am i doing wrong? and how come lowering f12 does not affect the wall deflection.

please see attached file. 

In RSA you would need to change values in the fields that are grayed out and therefore not accessible (the request for making them editable has been sent the development team). The workaround is to use the user defined orthotropy instead.

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[mgaafar]

Thanks Artur, it works now.

may I ask, the label of f12 and K12 is not correct then, because that is not consistant with shell terminology f12 and k12?

and in other words, what is the effect of changing f12 and k12 on the results then?

Thank again, much appreciated!

[Artur.Kosakowski]

The influence of strain in the direction X on stress in the direction Y. 

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[Tuctas]

Yes, we have to be carefull about these definitions and Robot’s terminology. Artur’s explanation was very illustrative as concerns the membrane stiffness. The 3rd term of the diagonal of flexural/bending stiffness matrix, does it concern the in-plane torsional moment? The two terms of shear stiffness matrix have to do with shear's orientation that is perpedicular to the shell’s plane, aren’t they?      

[Pawel.Pulak]

Hello,

The 3rd term of the diagonal of flexural/bending stiffness matrix corresponds to twisting moment. In case of planar finite element it causes deplanation/warping.

---------------------------------------------
If this post answers your question please click the "Accept as Solution" button. It will help everyone to find answer more quickly!

Regards,

[Tuctas]

Thank you for the answer Pawel (twisting moment is what i meant indeed, when i said torsional).

As concerns the shear stiffness matrix, is like that what i described?

[Pawel.Pulak]

Yes, it is.

"The two terms of shear stiffness matrix have to do with shear's orientation that is perpedicular to the shell’s plane"

---------------------------------------------
If this post answers your question please click the "Accept as Solution" button. It will help everyone to find answer more quickly!

Regards,