[Optimizer]

Hi all.

I am meshing a turbomachinery blade passage with Pointwise T-rex.

Please see attached for the details of the geometry and the mesh. The red lines I added represent the solid walls on which the T-Rex boundary conditions are expected.

When I Creates a volume mesh and initialize the block with T-Rex, The layers of tetrahedra from boundaries can't be created. Through the cutting surface, it can be found that, it is only on the outer surfaces of the computational domain that the anisotrpic tetrahdera generated (shown in the green circles), and there is no anisotrpic tetrahdera generated along the solid walls.

I have tried many times through adapting some parameters for T-Rex, but the prolem can't be solved.

So, could you give me some suggestion or hints for this issue?
Any comments and suggestions will be greatly appreciated. Thank you.



BTW, the parameter settings are as following in my T-Rex meshing:
Max. layers: 23
Full layers: 19
Growth Rate: 1.2
Push Attributes is checked.
Collision Buffer: 2
Max. Angle: 165
Centroid: 0.7

[cnsidero]

Sorry, I'll have to ask the obvious. Have you created a T-Rex BC for the domains you want to create layers on, changed the type to Wall and entered an appropriate ds? If so, is the ds small enough that layers can be created, i.e. the first cell is non-isotropic?

[Optimizer]

Originally Posted by cnsidero

Sorry, I'll have to ask the obvious. Have you created a T-Rex BC for the domains you want to create layers on, changed the type to Wall and entered an appropriate ds? If so, is the ds small enough that layers can be created, i.e. the first cell is non-isotropic?

Thank you, Chris. You are always there giving us help.

I created a T-Rex BC and the ds on the wall for the first layer is small enough.
I really have no idea how I can do further to get the mesh I am looking forward to. Is there the possiblity that the undulate wall shape would cause the failure of T-Rex mesh generation?

[cnsidero]

Ok, thanks. I have another question. You set the following parameters:

Max. layers: 23
Full layers: 19
Growth Rate: 1.2
Push Attributes is checked.
Collision Buffer: 2
Max. Angle: 165
Centroid: 0.7

When you set Full Layers to non-zero (i.e. 19) with quality checking on (Max Angle and Centroid), if T-Rex can't create a full layer (before it reaches 19) it will pause and ask if you want to continue without full layers, generate isotropic cells, or stop. Which did you choose?

[Optimizer]

When I initialize the grid, the stopped points firstly appear in the conner of blade wall and endall. My endwall is a non-axisymmetric profiling endwall.

I don't kow whether this information is helpful or not for my prolem.

Thank you in advance.

[cnsidero]

Not quite what I was asking. I can continue asking questions or if you want a quicker solution, you can send me the grid and I'll take a look.

[Optimizer]

Originally Posted by cnsidero

Ok, thanks. I have another question. You set the following parameters:



When you set Full Layers to non-zero (i.e. 19) with quality checking on (Max Angle and Centroid), if T-Rex can't create a full layer (before it reaches 19) it will pause and ask if you want to continue without full layers, generate isotropic cells, or stop. Which did you choose?

To generate isotropic cells is my choice. Cheers.

[cnsidero]

Originally Posted by Optimizer

To generate isotropic cells is my choice. Cheers.

That's why. Choose to continue making layers (albeit non-full). You will still be able to recover prisms with non-full layers.

[Optimizer]

Originally Posted by cnsidero

That's why. Choose to continue making layers (albeit non-full). You will still be able to recover prisms with non-full layers.

Got it. That's why, indeed.

Thank you for your help,

[cnsidero]

Originally Posted by Optimizer

Thank you for your help,

Glad I could help. It's nice when the fixes are straightforward.