Modeling Super-Tall Buildings Shear Walls

[imurdereri]

Hi all

i want to modeling a large scale structure and i need to know if the best way to model shear walls is by using the element ShellMITC4. because, we have various ways to do that.

for example, ¿It is reasonable to model shear walls by using a fiber beam element? (dispbeamColumn or dispbeamColumnInt by Massone)

In the shell elements i am studying just the ShellMITC4 and the ShellNLDKGQ, ¿Which is the the best of this two?

someone know?

Thank you!

[fmk]

to model using shells or beam column elements is an open debate, comes down to modeling preferences. beams is easier and in opnesees there are more material options. using shells is a more rigorous modeling approach and can capture things that the beams cannot, e.g. shear lag, it is however more time consuming numerically.

the ShellNLDKGQ is new and te developer uses it to model super tall buildings in China. The ShellMITC4 has been around awhile, a possible bug was reported in it this morning. Switch between the 2 should be fairly easy

set shellType ShellMITC4
element $eleType 1 1 2 3 4 10
..

and even doing a little study to see differences between the two:

for set shellType {ShellMITC4 ShellDKQG} {
wipe
..
element $shellType 1 1 2 3 4 10
..
..
recorder Node -file nodes$shellType.out ....

}

[imurdereri]

Thanks a lot!.

i am already doing what you mention and yes, is easy to switch between the two types of shell element. I am trying with a different mesh size and varying the parameters of the concrete constitutive and comparing with a real data that i obtain from the NEEShub Data base "ACI445B Shear Wall Database" where you can find the test results of his shear walls tested by Zhang, 2007 and many others types of walls.

¿The developer that you said is Xinzheng Lu right? because i read is document "A shear wall element for nonlinear seismic analysis of super-tall buildings using Opensees" wicth is avaible online since 11 February 2015. He has a website where you can download examples of shear walls.

¿So, the ShellNLDKGQ is the newest element? ¿So i presume it is better than MITC4?, and ¿The developer of this new element is the same Xinzheng Lu? cause in his paper he uses ShellMITC4.

In his document he mention that modeling large-scale structures is been rarely used in Opensees, but if you look at his examples, he run a 142m R/C structure with shear walls, using ShellMITC4 in a really simple way.

¿Are you familiar with special recorders of these two elements? or i can just use de Node recorder?

Thanks you very much again fmk

[imurdereri]

I need to know if i can record the stress-strain relationship. You can obtain the stresses with the element recorder but not the strains?

Thanks!

[fmk]

you should be able to get strains by replacing "stresses" in the recorder command with "strains"

[imurdereri]

i already try that, it doesn't work. can you help me please? i read others posts and they have the same problem.

[fmk]

post the element command and the recorder command

[imurdereri]

I am modeling shear walls so all the element are ShellMITC4, for example:
element ShellMITC4 1 1 2 7 6 1

The sections are LayerdShell, for example:
# section LayeredShell $sectionTag $nLayers $matTag1 $thickness1...$matTagn $thicknessn
section LayeredShell 1 10 4 0.0125 11 0.0002403 11 0.0003676 9 0.001118 4 0.048274 4 0.048274 9 0.001118 11 0.0003676 11 0.0002403 4 0.0125

iam using this recorders to obtain the stress-strain:

recorder Element -file Stress.txt -time -ele 1 stresses
recorder Element -file strain.txt -time -ele 1 strain

i read in the ShellMITC4 link on the wiki that you can also ask for an specific material with this command: "material $matNum matArg1, matArg2..." but i don't know what are the "matArg1, matArg2..." that are mentioned in the description.

[fmk]

what version of OpenSees?

[imurdereri]

in the last version

[imurdereri]

i am using this nDmaterials:

PlaneStressUserMaterial
PlateFromPlaneStress
PlateRebar
LayeredShell

and the steel02 for the PlateRebar

[fmk]

it should be "strains" not "strain" .. you are missing an s

[imurdereri]

Oh!, thanks you! it works =)
Now i can plot the stres-strain relationship, but i have to understand the column order first.

Thanks again fmk

[imurdereri]

Last question please:

In the LayeredShell section i have 10 layers with different materials.
If i want the stress-strain relationship of a material in a specific layer, ¿how it works the "material $matNum matArg1 matArg2.." in the NOTE of the ShellMITC4 element?

Because when use the element recorder and ask for streeses os strains, i get 32 columns and i dont know what they mean.

i cant't find any information relative to this command "material" in the element recorder for a Quad element

[fmk]

put a -xml instead of file .. you are getting 8 outputs per gauss point, the -xml will show what they are.

you can ask the material at the individual gauss point with the material $gp args
e.g
... material 1 deformations

you can ask the individual materials in the layered material
.. material 1 fiber 8 ...