Mesh Size for Tetra Mesh- structural Analaysis 2

[RathodAsit]

Dear All,

I am researching on what will be minimum Mesh size required for sheet metal?

If any one knows any research material,any paper please let me know...

Thanks in advance

Regards
Asit Rathod

 

[rb1957]

1) do your own literature search.

2) there is no "minimum size mesh for sheet metal"; none. minimum mesh size depends on the geometry and the accuracy required.

3) why tet mesh sheet metal ?

another day in paradise, or is paradise one day closer ?

 

[aerostress82]

1-) Try to have at least 2 or 3 nodes in the middle of your sheet metal.
Meaning, if you have a sheet metal of 6mm thickness, a tetra mesh size of 2mm (with 1st order elements) would be a good enough approximation. This would enable you observe the stresses along the thicknesses better. Also, try to have almost similar mesh on top and bottom faces of your solid, so your mesh tetra mesh between these faces will be of very high quality and the stresses will be observed better.


2-) However, you could mesh a 6mm thickness sheet metal with 3mm element sizes too with 2nd order elements. This would actually give you a more accurate model as you would end up having 5 nodes. But in nonlinear analysis, these elements tend to warp a lot resulting in fatal errors (in Nastran). If your analysis is a linear static analysis, this approach will also be easy to apply then.


The simplest and more convenient version would be the 1st advice above in my opinion.
Please correct me if I'm missing anything.

Spaceship!!


Aerospace Engineer, M.Sc. / Aircraft Stress Engineer with 7 years of experience
(United States)

 

[rb1957]

1) this is the common practice for plates, with bending ... not so sure it applies to sheet metal parts. it would also lead to very small elements.

2) agree that tet10s are way superior to tet4s ... results with tet4s deserve extra scrutiny.

another day in paradise, or is paradise one day closer ?

 

[RathodAsit]

Dear rb1957,

Is it not advisable to used 3D Tetra Mesh in sheet metal? I already solve few problems on that and I got very nearer result. what element you are using for solving sheet metal?

aerostress82,
I usually prefer to go with option2, as my 3.0 mm thk sheet metal I used to give two element of 1.5 mm and two nodes on edges, but it will take more time for meshing and solving. even if I go with option 1 then I believe it will be more & more time consuming, I solve problem in solidworks simulation.

Thanks & Regards
Asit Rathod

 

[rb1957]

i din't say it was inadvisable, I just asked "why?" why not use prismatic brick elements ? the two faces of the sheet metal volume are near enough parallel. I'd suggest 20node bricks as opposed 8node. I suggest we use tets because most 3D meshers default to that and we are generally lazy.

using tet10s is much preferred. whether you need to use 1/2 thickness or full thickness as the element dimension is something you can investigate (similar to "mesh convergence" modelling).

another day in paradise, or is paradise one day closer ?

 

[rb1957]

i notice you didn't post a thickness, and aerostress assumed 6mm. if your sheet metal is 6mm (1/4") thick then it could react bending, justifying more elements through the thickness.

another day in paradise, or is paradise one day closer ?

 

[stressebookllc]

Why the hell do you want to use tet elements for sheet metal?

Why not 2D Shell Elements, they are designed for this type of geometry.

Aerospace Stress Analysis and FEA Courses

http://www.stressebook.com

Stressing Stresslessly!

 

[RathodAsit]

Dear StressbookLLC,

the software which i used for simualation, in which auto option of auto Meshing of 3D CAD DAta and used Automesh 3D Tet elements, so used 3D Element,If I will used 2D Shell and 3D Tet Elments how far from result? Any suggestion.

Thanks & Regards
Asit Rathod

 

[inline6]

3D elements will give better results if you do it properly but it cost more computationally, sometimes orders of magnitude.

mesh density will also depend on the stress field and the gradients you need to capture. Also if there is no out of plane bending of the sheet metal you can use a coarser mesh and dont need as many elms through thickness.

tet4 should be phased out

 

[ThomasH]

Hi

I am not sure that I agree with the idea that 3D elements (volume elements) will give better results. It depends on the application.

They will probably always give you a better picture since everything we work with has some volume. A beam has a crossection even if the software only shows you simple line.

If I am working with a plate structure, like sheet metal, I would use plate elements (plane elements).
For a structure made of beams, like a frame, I would use beams (line elements).

The only reason I can think of to use solids (volume elements) for sheet metal is if the software can only handle solids. Then I might consider it, but I would also consider a different software, if possible .

To get results with comparable accuracy for sheet metal using volume elements instead of plane elements. I think that would require a tremendous amount of more elements for the volume mesh.

What results is it that you need that the plane elemens can't give you?
As for automeshing, I usually avoid it.

Good Luck

Thomas