hmmmmmmmmm ok ;) http://www.4shared.com/rar/LRL3929E/..._complete.html

Hello

I tried to refine the blocks around the cylinder and make a mesh file. But after making the .msh file, FLUENT is not reading the mesh and showing some error regarding the two interior circles which you mentioned yesterday.

It is saying creating wall between circles and then critical error occurred.

Can you please try yourself once and then guide me where I am going wrong

There is no error in new mesh. try again and let me know. four times refined mesh

The error is not in the mesh, but the way I am refining the mesh. Kindly give me a detailed step by step procedure to refine mesh using refinement option in blocking tab. I am going some where wrong and when I read the mesh in FLUENT, it is giving the following message.

Note: Separating interior zone 21 into zones 21 and 2.
int_circles -> int_circles (21) and int_circles: 002 (2)

Warning: Inappropriate zone type (interior) for one sided face zone 2.
Changing to wall.

I want to refine only some of the blocks and not entire mesh. Mainly blocks around cylinder wall.

I am doing following way. If you find anything wrong, let me know.

I am first going to blocking tab--> Refinement of selected blocks; refinement in all directions--> then Premesh; when I finish with refinement, it is asking me to recompute mesh, as it is out dated. I am clicking on yes. Then Convert to unstructured mesh. Saying mesh already exists, so I am clicking on replace. Earlier when I unselect Pre mesh, mesh disappeared. But after refining, even if I unselect it , mesh appears.

Then Output, selecting .uns file of cylinder and writing input file to FLUENT.

Thanks

Interesting-Never seen this before
 

I have also found same problem and tried many options but in vain. Shall try tonight and update as soon as possible.

PSYMN can easily solve this problem



http://img259.imageshack.us/img259/9...hcylinder2.png
http://img210.imageshack.us/img210/6...hcylinder5.png
http://img577.imageshack.us/img577/3...hcylinder7.png
http://img72.imageshack.us/img72/261...hcylinder8.png
http://img585.imageshack.us/img585/3...hcylinder9.png
http://img542.imageshack.us/img542/4...cylinder10.png

Hello Far

Did you get to know any reason behind the error?

not yet. May be some one else be able to find the answer. mean while you can refine the wake blocks by increasing the no of nodes on edges.

Hello

When I checked the mesh you uploaded without any refinement, it has shown problems near the cylinder wall (single edges problem). These single Edges being shown are one is cylinder wall, which has a mesh line coincide with it and next concentric circle edge. I deleted them when diagnostics prompted me to take action. Next time when I check the mesh, it is showing next two concentric circles around cylinder as uncovered faces and single edges.

What exactly happening here?

Also, give me directions briefly how you made the mesh, I mean the approach, as I am very new to ICEM CFD.

Thanks

Hello

When I try to increase no of nodes of wake side of cylinder, side opposite to wake also getting refined. Exactly when I tried to increase no of nodes on edge of inner circle, both wake side and inlet side nodes getting increased.

What might be reason for this?

Thanks

turn off the meshing to all parallel edges.

I never turned it on. Between please post the guidelines,

it was turned on when creating the project and saved. Now you need to turn off them

Hey

Also when I try to check the mesh you made in FLUENT, it leaves a message saying that mesh has high aspect ratio quadrilateral cells, it may effect your wall distance. Max aspect ratio getting is 145.

How to reduce it and make mesh of more quality.

Thanks

It is not the refinement that is killing you...

Here is an example with refinement that comes up in fluent without an issue...

You can ignore the "uncovered faces" error when you expect coupling. In fact, fixing it will wreck the coupling. Just make sure that during output, you set "ignore coupling" to "No".

For 2-D case we have this error. With 3-D models this function is working without any problem.

I'll look into it tomorrow.

Aspect ratio is cells' length divided by cell's width. Higher aspect ratio means that cells are stretched in length wise direction and which is normal in boundary layer, where gradients are in normal direction. However it is good to use AR=1000 for 3D and AR=100 for 2D. However with double precision solver AR= 145 shouldn't be any problem. You may check your results for AR = 80, 145 and 200 and be sure what is happening with important flow variables.



To reduce the AR either increase the no of nodes on circle or increase the height of first cell. Again you need to check the effects on important flow parameters.

I was asked about the uncoupled faced error on post #45...

Uncovered faces are volume elements that do not have another element against them... Usually, it means that the volume mesh is not covered by surface elements. This means there is no boundary for the solver, which usually will result in a failure. It could also mean that there are volume elements missing from the mesh, which also isn't good... A similar problem can happen in 2D meshing if there are not line elements around the 2D elements...

But in this case with the hexa refinement, you have two elements up against one with a "hanging node". The uncovered faces check sees this as the same sort of problem, but it is actually a coupled mesh situation.

I will see if we can tweak the check so it doesn't identify coupled elements as "uncovered". In the mean time, just look at the uncovered faces subset and, if they are all at this refinement location, don't worry about it.

Oh yea, make sure that during export to Fluent you select no for the question about "ignore coupled elements".

I am not able to avoid refinement in parallel edges. It always happens when I am refining one edge. Help me out.

Thank you

Also I refined blocks around cylinder and wake region to level 2. When I try to read the mesh in FLUENT, it showing error. I am attaching here the refined mesh files. I am going somewhere terribly wrong, may be these files will help you understand better where I am going wrong.




http://www.mediafire.com/download.php?p23wmnx2fkuxc49