CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Meshing & Mesh Conversion (https://www.cfd-online.com/Forums/openfoam-meshing/)
-   -   [Other] mesh generation of porous media (https://www.cfd-online.com/Forums/openfoam-meshing/149349-mesh-generation-porous-media.html)

liu-t11 March 3, 2015 02:18

mesh generation of porous media
 
hi, everyone! I am new to openFoam. Though some simple cases with simple geometry can be done, many solvers and utilities are not familiar.

I am trying to solve a 2D case of porous media, but stuck in generating the meshes. The blockMesh is not so useful for this problem.

I get a txt file to tell the position of the solid and the pore, but do really not know how to generate the meshes. Should I use snappyHexMesh or other mesh generators?

Anyone solving similar questions is welcome to give suggestions and I really appreciate your help! :)

Saideep May 27, 2015 05:55

Porous media generation!!
 
2 Attachment(s)
Hi Eric;

Though I realise yours is an old thread, did you succeed in accomplishing your task?

I am quite sure you cannot use blockMesh as it can only create more or less structured meshes with arc and splines etc but it would be useful if you can generate a 2D image of your pores or solids on inkscape or blender and convert it to a 3D .stl image and use it over snappyHexMesh.

I have the results something like this(figure 1) for a uniform pore geometry. But it is not of much help as when I inject a fluid at top it just flows quite fast and in the easiest path(straight).

So, next step I tried to create a non-homogeneous pore geometry. But it is surely not easy as when you try to generate the solids randomly there can be several connections inbetween the solids and your snappyhexmesh somehow covers only a small area.

Actually I am impressed by the way how Andrea Ferrari and Ivan Lunati created there random distributed solids in paper: Direct numerical simulations of interface dynamics to link capillary pressure and total surface energy.

If you already have any idea please do share.

Thanks
Saideep

Saideep May 27, 2015 06:16

1 Attachment(s)
I am interested to keep this thread alive and from my new model for a non-homogeneous porous medium it looks to get some sort of shape.

I had to refine the mesh so much (800*800) to get some sort of meaningful result for the porous medium that is randomly generated using inkscape(to generate random solids), blender (to convert .svg to .stl and from 2D to 3D) and snappyHexMesh to get the required model shape for analysis.

But the problem now is the time. As the dx is going so low, the Co number is too large and thereby ultimately it effects the time step size. However, I will just run basic simulation to see how the results vary from homogeneous case and will keep updated.

Thanks and waiting for any suggestions.
Saideep

liu-t11 May 28, 2015 08:18

1 Attachment(s)
hello, saideep:

i am glad that you reply to me. i am using snappyHexMesh to generate mesh for my porous media. the .stl file is obtained from the CT images, which i did some process.

the mesh is as follows.

the http://www.cfd-online.com/Forums/att...1&d=1432815464

Majed January 19, 2016 14:22

1 Attachment(s)
Hello Foamers,

I am new to OpenFOAM. I am trying to simulate a two-phase flow in porous medium using PorousInterFoam solver. I my case I have a permeable wall which I want to simulate it as a surface in which there are many holes so the the porous medium region is connected to the surface flow region. I am wondering if there is an easy way to do that. I am attaching a photo of my geometry for better understanding of my case.

Any help would be greatly appreciated.

KayGhana February 24, 2017 10:32

3-D porous media: twophase flow
 
Hello All,

I have a three phase (solid, water, air) 3-D pore structure.
I also have the STL files for all three phases (solid, water, air) generated.
Any ideas how these phases can go directly into OpenFOAM with boundaries correctly delineated would be appreciated.

For the boundaries,
The top boundary is atmospheric (later I would have an infiltration boundary), the bottom is a fixedpressure (water would either drain by gravity or rise by capillarity depending on the direction of pressure step down or step up) boundary and the sides are walls (no flow).

I look forward to your input.

Thank you.




Quote:

Originally Posted by Saideep (Post 547836)
Hi Eric;

Though I realise yours is an old thread, did you succeed in accomplishing your task?

I am quite sure you cannot use blockMesh as it can only create more or less structured meshes with arc and splines etc but it would be useful if you can generate a 2D image of your pores or solids on inkscape or blender and convert it to a 3D .stl image and use it over snappyHexMesh.

I have the results something like this(figure 1) for a uniform pore geometry. But it is not of much help as when I inject a fluid at top it just flows quite fast and in the easiest path(straight).

So, next step I tried to create a non-homogeneous pore geometry. But it is surely not easy as when you try to generate the solids randomly there can be several connections inbetween the solids and your snappyhexmesh somehow covers only a small area.

Actually I am impressed by the way how Andrea Ferrari and Ivan Lunati created there random distributed solids in paper: Direct numerical simulations of interface dynamics to link capillary pressure and total surface energy.

If you already have any idea please do share.

Thanks
Saideep


minzhang May 26, 2020 19:17

I have a 3D segmented image (0,1 data set)
 
Hello Liu,

I am wondering whether you could give more details about how to get the .stl file from the CT images. Now I have a 3D raw data of a segmented CT image. I am trying to generate the oepfoam mesh.

Thanks!



Quote:

Originally Posted by liu-t11 (Post 548017)
hello, saideep:

i am glad that you reply to me. i am using snappyHexMesh to generate mesh for my porous media. the .stl file is obtained from the CT images, which i did some process.

the mesh is as follows.

the http://www.cfd-online.com/Forums/att...1&d=1432815464



All times are GMT -4. The time now is 07:00.