[Danial Q]

Hi,
Could anyone please give me advice for meshing in this case, where I have two domains (fluid, solid). I am solving Eulierian homogenous phase, free surface is involoved and thermal energy model is employed also. So need a very fine mesh as domains are of micron sizes too and i have to capture phase change at domain interfaces.
I have attached pic below.

Thanks in advance

[PSYMN]

You could do it with a sweep or MultiZone mesh (probably would do that automatically anyway).

To control the distribution, switch the selection tool to select edges, then pick the 4 edges (hold down ctrl) perpendicular to the interface and insert method => Sizing. Setup some sort of edge biasing so the mesh will refine as it approaches the interface...

[Danial Q]

I have used sweep method and edge sizing. But after many tries ,the meshing window(small progress window) is stuck at the same point for almost like 14 hrs .
Could you please have a look at parameters and locate the error or unnecessary tighter parameters?

Thanks

[Gweher]

Could you please add your .agdb file

[Danial Q]

I have done sweep meshing but was unable to do edge sizing for unknown reasons. It got stuck somewhere when i selected Hard behavior with BIAS >>ON and BIAS factor 1.
Now the worry is, i have 8 million elements and my system needs multiprocessor. so the question is, should i do meshing again with less element size.would it work???

[Gweher]

I never encountered that error, but did you try to assign the edge sizing before adding the sweep method? Cause I did a quick cube of 20 microns edge size and assigned 100 elements in each direction, so a total of 1M and it took less than 10min (I have a 16gb Z210 workstation). You should try edge sizing first and then use mapped face, compute the surface mesh and then generate the volume mesh.

1MelementsAM.jpg

Regarding the multiprocessor did you assign multiple CPU in the Ansys Meshing Options (under >Tools>Options>Meshing) ?

[Danial Q]

No, i did not assign multiple CPU, glenn told me that it was not issue of CPU but CFX SOLVER capability. Anyways, could you please explain how did you set 100 elements, because I have option edge size element. No, I did sweep meshing first and then edge sizing.

[Gweher]

Well under >Sizing>Definition>Type you can choose between :

- Element size
- Number of divisions
- Sphere of influence

I assigned 100 elements using Number of divisions under >Type (see attachment)

Originally Posted by Danial Q

No, I did sweep meshing first and then edge sizing.

Have a try assigning Sizing first and then sweep.

And yes, I forgot for the SMP it depends on the method that you are using, sorry.

[Danial Q]

I have generated this mesh (attachments) and please comment on mesh statistics;
Mesh Statistics |
+--------------------------------------------------------------------+
| Domain Name | Orthog. Angle | Exp. Factor | Aspect Ratio |
+----------------------+---------------+--------------+--------------+
| | Minimum [deg] | Maximum | Maximum |
+----------------------+---------------+--------------+--------------+
| splat | 90.0 OK | 2 OK | 53 OK |
| substrate | 38.7 ok | 6 ok | 37 OK |
| Global | 38.7 ok | 6 ok | 53 OK |
+----------------------+---------------+--------------+--------------+
| | %! %ok %OK | %! %ok %OK | %! %ok %OK |
+----------------------+---------------+--------------+--------------+
| splat | 0 0 100 | 0 0 100 | 0 0 100 |
| substrate | 0 3 97 | 0 <1 100 | 0 0 100 |
| Global | 0 <1 100 | 0 <1 100 | 0 0 100 |
+----------------------+---------------+--------------+--------------+

[Far]

seems good

[Danial Q]

Could you please explain that how can I set the number of cells for meshing as I have some parameter like " 20 cells per 20microns". This parameter gives some grid independent results for my model. What is that supposed to mean?

[Gweher]

Well if your results are grid independent with that parameter for a particular setup then you do not need to continue refining your mesh for that case.

So you just need to generate your mesh according to the parameter you just mentioned in order to obtain grid independent results.

[Danial Q]

All i have as information about meshing is "Hexagonal mesh (grid independent) with 20cells per initial dia of drop (i.e. 20 micron). So, Could you please mention where can i set this "number of cells" parameter. I am unable to find it or may be if it is something different than normal meshing then how to employ this.

Thanks