volume tetra elementi mixed or trias crash

[mirko088]

Hi, I'm creating my model's mesh, but I'm starting to find crash issues. I'll explain in detail. I initially meshatto all the model (constituted from several parts) in step 1 however being at first weapons after finishing everything (materials,ply etc) I had different errors, so I was advised to work we say part by side and check every time if I had problems. I continued to work like this and I managed to complete about 3/4 model without errors, however now when I try to meshare the component with tetra volume and 2d:mixed the software crashes (this is the mesh that I used until now for parts in 3d ). I tried to throw a mesh2d with mixed elements and performs it quietly, as I did the test with the volume tetra with elements 2d: trias and performs it also. the problem arises only with volume tetra elements 2d: mixed or square. Does anyone know how to solve the problem? if I continue to use the mesh 2d:trias for the remaining part of the components instead for those already created I have a mesh 2d:mixed is a problem?
Thank you.

 

[Onda]

Maybe if you show us some pictures we understand better

 

[mirko088]

I have attached the image to the settings until now used for the creation of the mesh. Until now I have meshato without problems the components you see in the figure, now if I try to meshare the new component highlighted with those parameters it blocks hypermesh while if instead of using 2d type: mixed use 2d type: trias hypermesh creates mesh quietly . place also photo of when software stops working

 

[Onda]

not all solids are meshable from quad elements. If you want to make mesh simple use tetrahedri with midside knots.
Otherwise, there are special procedures for meshare to hexaedri but I do not know them in hypermesh.

 

[mirko088]

if I open a new file and try to meshare that component with mixed elements but it makes me do it. the problem arises with all the remaining parts and not only with that particular piece. in case I now use trias elements while for the components that I already meshato have used the mixed elements there would be problems at the level of analysis?

 

[Onda]

depends on how you interact with each other.
if they are connected with rbe or other means of connection, type glue, no problem. If you want to work with matching knots, then you can't
Looks like you have a corrupt database, so for some reason it doesn't make you move on with mesh.

 

[mirko088]

how do you know there's a mistake in the database?

 

[Onda]

If you work with matching knots, you have to make sure that all parts use the same elements, otherwise the transfer between one part and the other is not correct.
If a piece you can meshare it in a database, but not in the other, there is a chance that there is a mistake in what you can't meshare it. Sometimes the database gets corrupted.
Anyway, I don't know what you're trying to do. normally you do not work with solids at matching knots. solids are normally separate bodies, connecting to other bodies through connection systems. The coin knots are used more for shells.
then, to each his way of working. .

 

[mirko088]

the central shovel and the tip have an inside in honeycomb that then I go to meshare as solids, while the external skin I am treating it as a 2d element. therefore they are two solids in the construction of the model as divided into parts but in reality it represents a single solid

 

[Onda]

Okay, then clearer now. but therefore you should have a single solid, of which the shell models with planar elements and the inside with solid elements
Now, so that you can meshare with hexaedral elements you should have that the elements on the upper shell are identical to those on the lower shell. otherwise you can't always meshare. In other words, assuming that you can do exactly the same mesh on the two shells, then solid elements are one extrusion from one face to another. normally you do so to make blades. .

 

[mirko088]

I'm sorry, but I don't know what you mean. It's my first work on hypermesh . I'm working starting from the volumes that I create a tetra volume like mesh and then I write back the faces of the created mesh that will be so a mesh 2d . at that point erasing the faces that I don't care about the mesh 2d I have the skin that will surely have the knots matching with those of the solid and then I execute this work for every part. I just launched the test analysis to see if it encountered errors using the mesh trias on the last piece and ended the analysis without errors if it did not transmit the load properly I would have had some kind of error?

 

[Onda]

as a process is fine, but it is not said that between two matching surfaces of two different solids make you the same mesh and then you keep the same knots. There should be a way to get it done. I normally work with the reverse process, I first create the surface mesh and the solid one, but I do it in another pre-post. so it is not said to be the right way.
the tria problem is that these are generally not considered good elements.
or better:
shell mesh must be made with linear quad elements
solid mesh, or linear hexaedral elements or square tetrahedra elements.

in the case of mixed mesh, you have to use compatible elements and you can't mix linears with qudratrici, so, you should meshare to mixed shells and with dominant hexaedri the core.

this theory...
If you use square solid elements with linear plane elements, you have surface solid knots to which do not match shell knots.

 

[mirko088]

ok thanks I see if I can solve in that case I try to recreate the model again to avoid this problem related to the database.