PointWise T-rex non-orthogonality OpenFoam

LohengrinVB · January 9, 2019, 05:06

Dear PW an OF enthusiasts,

I am struggling with the mesh of a sharp corner for my thesis in PW V18.2.
There is a similar thread posted recently, but it couldn't help me.

Some pictures of the geometry (cross-section), two wings joined together.

junction_crossection.png

junction_crossection_detail.png

I want to create a BL mesh around this using T-rex. Overall it goes well, except for the sharp corner. Too many highly non-orthogonal cells are created for OF. I would to like keep the sharp edge since rounding it gives trouble at the trailing edge.

The question: How to use T-rex such that the non-orthogonality stays lower than 70°?
I already tried the collision buffer, seed layer and skew criteria.

Any suggestion is highly appreciated.
Thanks in advance.

Lohengrin

Jesus.M · January 14, 2019, 07:35

Hi Lohengrin VB.

Have you tried using the smoothing parameters? They usually help solving this kind of problems.

Hope this is of help to you.

Good luck!

LohengrinVB · January 16, 2019, 11:22

Hello Jesús,

Thank you for your reply! I boosted the smoothing and indeed it helps. The max non-ortho lowered by +- 2°. However, there still is a substantial problem. I had to switch to the rounded angle.

Thanks,
Lohengrin

LohengrinVB · January 16, 2019, 12:20

Dear All,

I would like to reformulate the question.
How to use T-rex in corners less than 90°?

I attached some pictures of the problem area. A section in the flow direction. A detail of the circle. Cut of the lower boundary.

Inkednonortho_section_LI.jpg

nonortho_section_detail.png

nonortho_z0.png

T-rex is applied on the wing and the lower surface is put to a match condition. This surface acts as boundary (wind tunnel floor/end plate) for the block.

Any suggestions on how to tackle this corner?

Thanks!
Lohengrin

Jesus.M · January 24, 2019, 01:56

Quote:

Originally Posted by LohengrinVB

Dear All,

I would like to reformulate the question.
How to use T-rex in corners less than 90°?

I attached some pictures of the problem area. A section in the flow direction. A detail of the circle. Cut of the lower boundary.

Attachment 67803

Attachment 67804

Attachment 67805

T-rex is applied on the wing and the lower surface is put to a match condition. This surface acts as boundary (wind tunnel floor/end plate) for the block.

Any suggestions on how to tackle this corner?

Thanks!
Lohengrin

Hi Lohengring,

Thank you for the comment, I'm glad I could be of help.

The thing is that T-REX build up layers until it reaches isotropy with the non structured cell types. That is the reason why at the sharp corners of the trailing edge there is almost nothing.

However, I have seen that you have cut both trailing edges, therefore I would suggest you to set a adjacent boundary condition on both blunt trailing edges.

Then you should figure out the right set-up of max and full layers and the smoothing parameters along with the distribution of the connector points in that region.

I hope that this can help you again.

Good luck

Jesus.M · January 25, 2019, 06:18

Quote:

Originally Posted by Jesus.M

Hi Lohengring,

Thank you for the comment, I'm glad I could be of help.

The thing is that T-REX build up layers until it reaches isotropy with the non structured cell types. That is the reason why at the sharp corners of the trailing edge there is almost nothing.

However, I have seen that you have cut both trailing edges, therefore I would suggest you to set a adjacent boundary condition on both blunt trailing edges.

Then you should figure out the right set-up of max and full layers and the smoothing parameters along with the distribution of the connector points in that region.

I hope that this can help you again.

Good luck

Hey,

I have found a video very interesting about meshing with PW. Perhaps it is of help to you. In about minute 4, it talks about colission buffers and more features of T-REX, that could solve your problem.

https://youtu.be/JIj97iUCB5Q

Good luck

LohengrinVB · January 25, 2019, 06:25

Hi Jesús,

Thanks a lot for the help!
Setting the TE to adjacent results in some serious non-ortho generated there.
But this could be related to the surface mesh.

However, putting the "floors" to adjacent in stead of match solves the problem locally, but moves it up higher in the mesh.

I am trying different surface domain to see the effect. I'll come back when I find an optimum.

Thanks again for the tips and inspiration.
Greatly appreciated.

Best Regards

January 9, 2019, 05:06	PointWise T-rex non-orthogonality OpenFoam	#1
LohengrinVB New Member Join Date: Jan 2019 Posts: 7 Rep Power: 7	Dear PW an OF enthusiasts, I am struggling with the mesh of a sharp corner for my thesis in PW V18.2. There is a similar thread posted recently, but it couldn't help me. Some pictures of the geometry (cross-section), two wings joined together. junction_crossection.png junction_crossection_detail.png I want to create a BL mesh around this using T-rex. Overall it goes well, except for the sharp corner. Too many highly non-orthogonal cells are created for OF. I would to like keep the sharp edge since rounding it gives trouble at the trailing edge. The question: How to use T-rex such that the non-orthogonality stays lower than 70°? I already tried the collision buffer, seed layer and skew criteria. Any suggestion is highly appreciated. Thanks in advance. Lohengrin Last edited by LohengrinVB; January 9, 2019 at 06:46. Reason: add release

January 14, 2019, 07:35		#2
Jesus.M New Member Jesús Miguel Sánchez Join Date: Jan 2019 Posts: 9 Rep Power: 7	Hi Lohengrin VB. Have you tried using the smoothing parameters? They usually help solving this kind of problems. Hope this is of help to you. Good luck! LohengrinVB likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[Commercial meshers] Inconsistent OpenFOAM results using Pointwise & blockMesh	a.runnerdude	OpenFOAM Meshing & Mesh Conversion	6	November 19, 2018 10:35
OpenFOAM Training: Programming CFD Course 12-13 and 19-20 April 2016	cfd.direct	OpenFOAM Announcements from Other Sources	0	January 14, 2016 10:19
OpenFOAM v3.0.1 Training, London, Houston, Berlin, Jan-Mar 2016	cfd.direct	OpenFOAM Announcements from Other Sources	0	January 5, 2016 03:18
[Commercial meshers] Native OpenFOAM interface in Pointwise	cnsidero	OpenFOAM Meshing & Mesh Conversion	41	May 20, 2012 18:30
Native OpenFOAM interface in Pointwise	Chris Sideroff	Main CFD Forum	0	January 16, 2009 12:37

January 16, 2019, 11:22		#3
LohengrinVB New Member Join Date: Jan 2019 Posts: 7 Rep Power: 7	Hello Jesús, Thank you for your reply! I boosted the smoothing and indeed it helps. The max non-ortho lowered by +- 2°. However, there still is a substantial problem. I had to switch to the rounded angle. Thanks, Lohengrin

January 16, 2019, 12:20		#4
LohengrinVB New Member Join Date: Jan 2019 Posts: 7 Rep Power: 7	Dear All, I would like to reformulate the question. How to use T-rex in corners less than 90°? I attached some pictures of the problem area. A section in the flow direction. A detail of the circle. Cut of the lower boundary. Inkednonortho_section_LI.jpg nonortho_section_detail.png nonortho_z0.png T-rex is applied on the wing and the lower surface is put to a match condition. This surface acts as boundary (wind tunnel floor/end plate) for the block. Any suggestions on how to tackle this corner? Thanks! Lohengrin

January 25, 2019, 06:25		#7
LohengrinVB New Member Join Date: Jan 2019 Posts: 7 Rep Power: 7	Hi Jesús, Thanks a lot for the help! Setting the TE to adjacent results in some serious non-ortho generated there. But this could be related to the surface mesh. However, putting the "floors" to adjacent in stead of match solves the problem locally, but moves it up higher in the mesh. I am trying different surface domain to see the effect. I'll come back when I find an optimum. Thanks again for the tips and inspiration. Greatly appreciated. Best Regards