Structured mesh of 3D model "particles in a box": how to do it? 1

[mika3049]

Hi,

I am trying to design 3D structured mesh of unit cell model of a material. The model is built as several particles (of complex shape) in a 3D cubic box. I generated Parts, Instances, Sections of the particles and the 3D box, and also produced a Part Matrix (by boolean cutting the instances "particles" from the instance "box").
The particles are much smaller than the box size.

I want to have fine mesh in particles, and more coarse mesh in the matrix outside of particles.
How can i do it?

(I dont want to use Partition because the particles have quite irregular shape). I tried to mesh first particles, then matrix and then merge them (by boolean operation), but it seems not to work well.
Can you advise how to generate such mesh on the basis of Instances/Parts and without Partition?

Thank you

Mika

 

[mika3049]

More exactly: I mesh a Part Matrix and parts Particles, and then Merge them (into new Instance Total). But when I apply load to Total, and run the model, the CAE gives an error - "TOTAL is not meshed".

 

[Mustaine3]

After the merge you have a new part with it's own instance. One of both you have to mesh. The meshes on the original parts are irrelevant, since these parts are not used anymore.

 

[mika3049]

Thank you. But how can I define the fine mesh in the particles inside the box, and rough mesh on the faces of box? In Patran, you make fine mesh seed on some parts and rough mesh seed on box, and so you have it. In 2D simple case, you can also make fine mesh seed on the boundaries inside the box (by selecting them by hand) and so you have it. But how do you do it in 3D case with many complex-shaped particles? Make mesh seed on particle parts, then mesh seed on box, then merge them and only after that mesh them - will the mesh seed retained in this case? Or by playing with dependent/independent/parts/instances? Or just using Patran, as more flexible?

 

[Mustaine3]

I assume you have one part with multiple cells.
Apply the coarse global mesh seed to the whole part. Be aware of the minimum mesh size option in that dialog.
Then hide the cell, that represents the overall region, so you just see your particles on screen. Now activate the local mesh seed option and change the selection filter to 'cells'. Select all your particle cells (drag-box, e.g.) and apply the finer local mesh seed. See also the minimum size option here.

 

[mika3049]

Thank you very much! How can I hide the box, and keep particles on display? Since the "matrix" and particles are merged, they are one instance/one part. (When I merged matrix and particles, the initial instances are supressed) Or do I miss something?

 

[Mustaine3]

In the merge operation is an option (Retain Boundaries), that keeps each region as separate cell. When you've used that, you have a Part/Instances consisting of multiple cells. This is also necessary, when you want to apply different section (materials) to each cell.

Then you can hide/show cells as you like with the display group option.
When using the quick access buttons to hide/show, you get an entity filter below the viewport, were you can choose between instances, cells, faces, edges, ... See CAE Manual 78.1.2 Understanding display group Boolean operations

 

[mika3049]

Thank you very much!

 

[mika3049]

Hi, Mustaine, now I tried to follow your advices. The techniques with "display group otion" works very well indeed.
So, I can hide/show Cells which were connected into the Total body (by boolean merging parts).
But when I try to define free tet mesh and Mesh seeds on the parts, the system says that the parts are dependent and I can mesh them only by making them Independent via Instances or making them Current via Parts. I can not make them independent via Instances because they are supressed (or should I resume these parts?).
So I made the particles current via Parts, and defined Tet Free meshing/fine mesh seed for the particles in the box. Then I made the Matrix Current via Parts, defined the coarse Mesh seed on the Matrix. Then I tried to mesh the full unit Cell. But the mesh created on the full unit cell (which is in fact merged Matrix + Particles) disregarded the fine mesh seed on particles, and simply created the coarse mesh on all the model. Did I do something wrong in the process?