snappyHexMesh doesnt snap
3 Attachment(s)
Hey Guys, i know you must be tired of this sort of question, but i havent found a similar case in this forum.
I am trying to create a mesh of a wall with several drillholes in it. The wall including the holes is a stl file, created with AutoCAD. When i run snappy, it refines the mesh but doesnt snap the cells to the stl surface, as you can see at the attached files. I tried to vary the refinement options (maybe too much cells), but that doesnt take effect. I am sure to look at the mesh at time "2". The Mesh of the bounding box is nearly cubic. Here are my blockMeshDict and snappyHexMeshDict: Code:
/*---------------------------------------------------------------------------*\ Code:
/*---------------------------------------------------------------------------*\ |
Greetings Thomas and welcome to the forum!
The only thing that I can find to be potentially wrong is the "locationInMesh", which is a value too round and likely to get placed on top of a face or vertex of a cell. I suggest the following:
Best regards, Bruno |
Your SHM-Dict looks good, but surfaces not beeing snapped is a bug in OF 2.0.x Change to the new OpenFOAM (2.1.x) and it will (most certainly) work!
Watch your overall cell count. You have quite a large domain there and a lot of refinement. G. Jan |
I solved the problem i think.
I added an additional stlsurface refinement region and lowered the refinement levels at higher distance to the stlsurface. Now i get a nice mesh. Quote:
Quote:
Quote:
The cell count is very high, indeed. It took some time to generate the mesh;) Now it works, thanks to both of you |
1 Attachment(s)
Hi everyone,
I am studying free round jet flow in detail. PFA the computational domain. My inlet is very small compared to rest of the domain. snappyHexMesh is not meshing the inlet pipe. I had tried a lot of things but of no use. :( I am starting from the begining. So Any suggestions guys? :confused: Thanks in advance Turbulence. |
Have you tried to define a refinementBox around your inletpipe?
|
Hi staustrahltriebwerk :)
I had tried the refinement box thing. If my locationInMesh is in inlet pipe, then only inlet pipe remains after meshing. If my locationInMesh is in the big cylinder ,then only big cylinder remains. This is what happened in my previous runs. Any idea why? :confused: Thanks Turbulence. |
seems as if the pipe and the cylinder arent connected to each other. Which program do you use for stl generation?
|
3 Attachment(s)
Hi
I had used solidworks. For the background mesh I had used two separate boxes for bounding mesh. That may be the problem. :) Fresh Story: Now, I had set all the parameters false except castellatedMesh and proceeded with meshing. I made my background mesh very fine so as to make sure that it intersects my inlet pipe. PFA the images i had attached. rough diagram of my blockMesh: http://www.cfd-online.com/Forums/att...-photo0221.jpg The coloured object is my original geometry, and the wireframe is the mesh generated.As you can see in screenshot 3 instead of whole face only a small portion is getting meshed. Any ideas? :confused: Thank You Turbulence |
Does my background mesh has anything to do with this problem? :confused:
Tried many things..still stuck with this issue.:( |
Greetings Turbulence,
I'm not 100% certain I understand the problem here, but my deduction is this:
Best regards, Bruno |
Hi Bruno :)
Thanks for replying. I just realized that the small cyclinder is not open to the big one:p. I will make this correction and run this case again , will get back to you. Please do let me know if there are any more suggestions before I run this case. :) Thank You Turbulence |
Hi Bruno :)
I ran the case with only castellated mesh set as true. The big cylinder is meshed, and small one is totally disappeared! Questions: 1.Should my background mesh be finer? Should I make sure that the back ground mesh cuts the inlet pipe(small cylinder) ? 2. Even If I make that background mesh finer, (its taking lot time to run bolckmesh) I am just able to cut the small cylinder at 2 points. Then the resulting inlet pipe turns to be a rectange/square. How can I make sure that the small circular inlet remains circular? :confused: Thank You Turbulence. |
Hi Turbulence,
Create a refinement box for the castellation phase around this cylinder. And on blockMesh, extend a bit more the box around this cylinder, by 1 or 2 mm if possible. Best regards, Bruno |
1 Attachment(s)
Hi Bruno :)
I took your advise but the error couldn't be rectified. There must be something wrong with my snappyHexMeshDict. PFA the essential files required to run the case. Kindly look in this problem. Thank you,:) Turbulence. |
Hi Turbulence,
I'm sorry to say, but this base mesh is really lousy :( Given the geometries at hand, you should create cylinders with blockMesh for the base mesh. Or at least, use cube-like cells, not parallelepipedic ones Otherwise, even if you do generate a mesh with snappyHexMesh, it won't be very good. I didn't even try running sHM because I don't have enough RAM on my machine :( 2.5 million cells for the base mesh is a lot and I'm expecting this to shoot up to 10 million during sHM. I advise you to use these two tools: They are for using with Blender and they should make your life really a lot easier! The other hypothesis is to look here in the forum for: Code:
cylinder blockMesh Best regards, Bruno |
Hi Bruno :)
Thanks for the reply.My prof expects me to mesh learn meshing a solid geometry generated by solid works/catia(the geometry is pretty complex) using snappyhexmesh. So I don't have lot of options here. I will try to use cylindrical background mesh as suggested by you. Will get back to you. Thank a Lot Turbulence. :D |
Hi Turbulence,
I think I didn't explain properly: SwiftBlock and SwiftSwak are helper tools for setting up "blockMeshDict" and "snappyHexMeshDict". They themselves are not meshers! But still, before these existed, people have to set the dictionaries manually. So search the forum and you should find a lot of pearls of wisdom! Good luck! Bruno |
4 Attachment(s)
Thanks Bruno :)
I have managed to mesh the geometry 'satisfactory'LY :p Still a problem is nagging me. :mad: After running the snappyhexmesh, a unwanted surfaces is formed inside my inlet pipe. (image1 patch d). Image 2(patch e) is the face of big cylinder in contact with my inlet pipe. The whole point of simulating a round jet is defeated here :(. This unwanted surface is not lost even after running the snap , add layers part. :( I am attaching the necessary code. I think it has something to do with my stl surface, but I cant find the mistake. Please look into it. Thanks a lot for your help :) Turbulence |
Hi Turbulence,
I won't be able to look at this anytime soon... maybe this weekend. But in the meantime, I've picked up on this post: Quote:
Best regards, Bruno |
All times are GMT -4. The time now is 20:28. |