Boundary Layer creation error: Unrealistic large height computed

[Jawisr]

Hi all,
i'm trying to mesh a wing with master branch version but I'm experiencing some issues when boundary layer is created.
I've tried tons of parameters in "prismatic layer" window with no success. Some minutes after launching the volumetric creation this window appears:

This were my last parameters of boundary layer. I understand that the BL has too much height to be created, however those are the values that I gave to the mesh in version 1.4 and it was meshing it correctly.

regards!

[kateeisenhower]

Hi Jaswir,

I am very interested in this kind of application. Can you please post a picture of the successfully created boundary layer in enGrid 1.4. I am asking because I always failed in creating a decent prismatic layer in enGrid.

Thanks in advance,

Kate

Am 21.07.2015 um 07:44 schrieb Jawisr notifications@github.com:

Hi all,
i'm trying to mesh a wing with master branch version but I'm experiencing some issues when boundary layer is created.
I've tried tons of parameters in "prismatic layer" window with no success. Some minutes after launching the volumetric creation this window appears:
https://cloud.githubusercontent.com/assets/12989191/8793800/ee952902-2f7b-11e5-8ab0-50854bfe6ddc.png
This were my last parameters of boundary layer. I understand that the BL has too much height to be created, however those are the values that I gave to the mesh in version 1.4 and it was meshing it correctly.
https://cloud.githubusercontent.com/assets/12989191/8793817/228b5b28-2f7c-11e5-8ee8-49504c7addc1.png
regards!

—
Reply to this email directly or view it on GitHub #54.

[Jawisr]

Hi Kate,
Indeed I've been able to create some decent meshes with enGrid 1.4. Now, I am trying the new version but I think it is still in very early development and it is not stable since I couldn't mesh any BL at all.
Recently I've meshed some cases with 1.4 and they look quite good, however I get some fails in checkMesh, negative cells, skewness, etc. I'm trying to increase the number of iterations in surface, BL and tetras meshing to obtain better meshes.
I attach a HiLift case mesh with about 10M cells. The BL is divided in 11 layers, I could divide it in up to 24 layers but I have no pictures of that case.

Regards!

[kateeisenhower]

Hi Jawisr,

And thank you for the answer. I have been using 1.4 for some time now and experienced the same problems as you have plus the inability to create a decent mesh around sharp corners. Additionally the overall boundary layer height mainly seemed to depend on the size of the surface elements and not on my input parameters.

Best regards,

Kate

Am 22.07.2015 um 08:39 schrieb Jawisr notifications@github.com:

Hi Kate,
Indeed I've been able to create some decent meshes with enGrid 1.4. Now, I am trying the new version but I think it is still in very early development and it is not stable since I couldn't mesh any BL at all.
Recently I've meshed some cases with 1.4 and they look quite good, however I get some fails in checkMesh, negative cells, skewness, etc. I'm trying to increase the number of iterations in surface, BL and tetras meshing to obtain better meshes.
I attach a HiLift case mesh with about 10M cells. The BL is divided in 11 layers, I could divide it in up to 24 layers but I have no pictures of that case.
https://cloud.githubusercontent.com/assets/12989191/8819311/8bcaa13c-304b-11e5-94e9-0eefa6fb0c64.png
https://cloud.githubusercontent.com/assets/12989191/8819482/f96f3e18-304c-11e5-8d3e-c8c4f83c3c4c.png
https://cloud.githubusercontent.com/assets/12989191/8819497/162a9bec-304d-11e5-9077-09dd345659c4.png
Regards!

—
Reply to this email directly or view it on GitHub #54 (comment).

[Jawisr]

I have the same problem with BL height, I've tried to set it in absolute values with no success.
I think you can solve your problems in sharp corners by reducing the surface mesh in edges to lower values. Setting a lower grow ratio in 2D meshing also helped me, 1.05-1.2 is what I tend to use.
And finally I've find out that increasing the number of iterations until the fluctuation ratio is low (0-4%) also helps to create BL properly.
Regards,

[ftpronk]

Hello,
Jaswir, you get unrealistic heights because you haven't specified any mesh resolution. As a small note, the height specification in 1.5 is now absolute, and no longer relative.

Kate, it is recommended to transition to v1.5 as the prismatic layer creation is more robust.
Regards,
F.

[Jawisr]

Hi F,
thanks for your answer. I have some questions:

How can I specify the mesh resolution rules?
Is the heigh in 1.5 value in meters?
For example if I set: sphere = 0.001; will the BL total height be 1 mm?
Could you give us a brief example applied to the sphere tuto please?

Thanks again!!

[ftpronk]

Hello,
Yes, the resolution is in meters, and the syntax is the same as for the surface meshing. Surface resolution specification is advisable, yes. Edge resolution is not compulsory.
Unfortunately we currently do not have the time and resources to provide extended support for free, and we will therefore not be able to upload a more current tutorial. But the boundary layer meshing should be rather straightforward.

Kind regards,
F.

[Jawisr]

Thanks F,
I'll do some test and post it in the forum to help other users when I have some time. I think engrid is a very promising tool!
thanks!

[Jawisr]

Hi F,
could you give me some advice in order to avoid the h1<h0 error?. I don't even know what h1 or h0 means. I tried to set minimum edge length and relative face size (minimum value) to 0 but this seems not to be the problem.
Thanks!

[ftpronk]

Hello,
It means that you are probably missing a height definition somewhere, i.e. computing bogus heights.
Kind regards,
F.