Relationship of mesh size and wall thickness

[chocomocha]

Hi! do you have any idea if there is a relationship between the mesh size and wall thickness?

I am designing a reaction wall and strong floor for an actuator and shaking table using etabs. This type of structure requires very thick wall and slab thickness. And do u think that using using "shell" and defining it as "thick plate" would be more appropriate? My reaction wall thickness is 1.7m and strong floor is 1.2m.

Please help. Thank you very much.

 

[kellnerp]

What is the ratio of the thickness to the length of the structure in either direction?

TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

"Node news is good news."

 

[chocomocha]

my reaction wall is 7m high, 17m long and 1.7 m thick... thickness to length ratio would be 0.1

 

[rb1957]

that sounds reasonable.

you can test this yourself, reduce the element size ... do the results change significantly ??

 

[GBor]

You are on the cusp of shell element theory, and I would think, as thick as your wall is, it would behave more in accordance with brick theory (no rotational transfer). I would consider bricks, but rb's suggestion is, as always, a good one.

 

[kellnerp]

OK, you have attachments to this wall and the wall will see high point loading.

For a start, if your wall is 7 x 17 meters then assuming 1m square elements would give 119 shell elements to cover the wall. If you used solid hexahedral elements you would need at a minimum 357 elements (3 through the thickness). If you double the mesh density twice you would have 22,848 elements. This is not a bad problem to do in solids even on a modest computer. Using solids would take into account that the point loads were applied on the surface of the wall or floor and not on the neutral axis .5 meter below the surface.



TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

"Node news is good news."

 

[nothing111]

dear chocomocha,
as per SAP tutorial, ideal mesh size is maximum 5 times the thickness or less. and angle must be maximum 135degrees between 2 sides of a mesh

 

[gurmeet2003]

Best way to check the mesh size is: do a mesh convergence study.

Gurmeet

Gurmeet
Time is an illusion. Now is the only thing there is.

 

[BlasMolero]

Hello,
Let's see if this helps you:

http://translate.google.es/translate?prev=hp&hl=es&js=y&u=http://

http://www.iberisa.com/soporte/femap/shell.htm&sl=es&tl=en&history_state0=

Best regards,
Blas.

 

[chocomocha]

TO ALL:

thank you very much for your help especially to nothing111.

i'll keep you posted to what the results are, currently i am off to a new project, so i have to set it aside until i'm done with the new one.

thanks again