[Anorky]

Hi all,

I am currently meshing a cornering wing geometry. For simplicity the pictures only contain a part of my geometry to mesh (i.e. one side and half the corner).

I want to make a full hexagonal mesh around the wing geometry.

Around the wing geometry, I created two Boundary layer geometries. Each of those is blocked seperatly and in a later stage merged together. The resulting blocking is shown in figure 1 and figure 2. First the innermost (closest to wing surfaces) blocking is made and associated correctly with the geometry. Next the outer blocking is made and associated with the geometry (vertices to points and edges to curves).

Next I load the geometry containing the wing and first and second boundary layer. I delete the interface surfaces between both boundary layer geomerties. Then I load the first and second blocking and merge both together.

The outer blocking for this geometry consists of single edges where the innermost edges have the same shape but exist out of 4 smaller edges.
To make sure that the outermost edges contain the same number of elements as the sum of the elements on the 4 smaller egdes of the inner most blocking I use linked edge bunching. I included a simple sketch of this which can be seen in figure 5.

Previously I succeeded in making a good quality mesh. Now however whenever I run the Check Mesh function in ICEM, I always get the error that Volume orientation errors are found, see figures 3 and 4.

Does anyone know how to prevent this? How can I assure or repair this current problem? Does anyone know the cause of this problem?

I tried to explain my problem as clear as possible, however I can imagine not everything might be clear! Please don't hesitate to contact me for more information!

Kind regards and thanks for reading my post.

[PSYMN]

It looks like the mesh is just self penetrating. This results in mixed up elements that give this orientation error.

Run a scan plane thru the affected area and take a look. Try to figure out what is causing the issue. Usually it is a projection problem, but if it is totally in the volume, you may have an interpolation issue or something. These can usually be fixed at the blocking level. You may need to use link shape or something to keep this problem from happening.



Simon

[Anorky]

Hi Simon,

Thanks for your reply, I am currently playing around with the link shape option. I am also looking for possible projection problems, I hope this will yield a solution. But so far without succes.

Could following error message be the cause of my problem?

A minor internal error has occurred.
You may continue to work but if problems arise they may be related to this error.
Try using the check/fix function to fix the problem
The following message was designed to aid developers fix the problem:

internal error in copy_linked

A minor internal error has occurred.
You may continue to work but if problems arise they may be related to this error.
Try using the check/fix function to fix the problem
The following message was designed to aid developers fix the problem:

can't unlink counts at this point

One way or another I also get this error when using the linked bunching option... I also included new figures, focused on the problematic elements.

Altough I was wondering, I have a mesh, blocking and geometry file in which the negative volume orientations are not present.

Is there any way to by using these files to find out which steps I made in generating the that mesh and blocking? I mean to know which edges I linked together, which shapes,etc.?

Kind regards and thanks for the help!

[PSYMN]

Comparing the cut planes thru these regions would probably help... Right click on "premesh" and select "Scan". Use the screen select option to click on of the near by edges and let the scan pass thru.

From what I can see in these new pics, your mesh is falling over (like a collapsed parallelogram). This suggests your edge params are suddenly pulling the mesh or perhaps your mesh is being pulled by association to only part of a curve (group curves may be the solution). Either way, Scan planes will be your quickest way to determine the problem.

[Anorky]

Hi Simon,

Thanks for the tip. In could see the difference in the scan plane between the correct mesh and the falsified mesh. You are right, the elements collapse...
I have tried to figure out what the cause is, but so far without succes.

I noticed that, when I open the blocking files with a text editor there is a difference between both.

At the end of the blocking file the following expression can be found

DEFAULT_BUNCHING_LAW default 2
EDGE_RATIOS node_nums
{

for the correct blocking file the { is followed by a list in the form of:

1639 1641 8.00492 2 8.00492 2 0 0 1e+010
1639 1640 101.571 2 101.571 2 0 0 1e+010
76 79 0.892828 2 1.60998 2 0 0 1e+010
76 77 8.00007 2 3.63934 2 0 0 0
172 76 2.02001 2 2.02001 2 0 0 1e+010
172 175 8.00492 2 8.00492 2 0 0 1e+010
172 1639 6.57 2 6.57 2 0 0 1e+010
172 173 101.571 2 101.571 2 0 0 1e+010
77 78 0.892828 2 1.60998 2 0 0 1e+010
77 80 2.36285 2 2.36285 2 0 0 1e+010
80 81 0.892825 2 1.60998 2 0 0 1e+010
.
.
.
1054 1055 linked 115 114
1054 1056 2.525 2 2.525 2 0 0 1e+010
1274 1275 linked 115 114
1074 1081 linked 115 114
1081 1082 4.09519 2 6.79838 2 0 0 1e+010
1081 1088 linked 127 131
1123 1130 linked 127 131
1130 1136 linked 207 210
.
.
.
}

While examining the incorrect blocking file I noticed that the expressions with "linked" are not included in the list.

So I checked by making a new blocking, linking the respective edges (linked bunching and shape linking) , etc. and even than these expressions are not saved in the blocking file.
Do you have any idea why this happens? Could this be the cause of the problem?

I have also looked into the curve grouping. However without succes, I will look in to this again first thing tommorrow, for the moment I am out of inspiration

Again, thanks a lot for helping me out!


kind regards

[PSYMN]

Well, I think you may have found the issue.

Your bunching is not linked in the new file. But I have no idea why couldn't just copy these lines to the new blocking file and have it save properly.

The first two numbers in each of these lines refers to the vertex numbers at the ends. If your blocking is all the same (same vertex numbers), you should be able to copy and paste this entire edge rations section from one to the other and fix your distribution problem.

However it it is a projection problem you may need to do something else...

Can you post a screen shot of the scan plane? I may see something there.

[Anorky]

Hi Simon,

Indeed copying the lines to the corrupted file would be a solution. But since I want to script the whole process I would like to figure out why these lines are not saved.

In the attachted files you see some scan plane views. In these figures only part of the geometry is visible compared to the earlier figures.

The linking not being saved in the files seems to be the problem.
I look into this further in the morning.

Do you have any idea why these errors appear while link bunching edges?
This always appear at the same point during the process, which is indicated by the red circle.


A minor internal error has occurred.
You may continue to work but if problems arise they may be related to this error.
Try using the check/fix function to fix the problem
The following message was designed to aid developers fix the problem:

internal error in copy_linked

A minor internal error has occurred.
You may continue to work but if problems arise they may be related to this error.
Try using the check/fix function to fix the problem
The following message was designed to aid developers fix the problem:

can't unlink counts at this point

Besides this error messages I also have two warning messages appearing:

Warning:family ORFN used to exist but no longer does
Warning:creating family FLUID

But I don't think these can be the cause of the volume orientation errors. Is this correct?

Thank you for replying and helping me finding the solution.

Kind regards

[Anorky]

Hi Simon,

I think I found the solution. I'll have to verify this on monday but for now it seems promising.

Something was indeed going wrong during the linked bunching.

Thanks a lot for your help!!!

Regards,

[Anorky]

Hi Simon,

The problem is solved. The problem was indeed the linked edge bunching.

Thanks for the tips and solutions you provided!

Kind regards