[Aeronautics El. K.]

Hi all,

I'd like some help (again ).

I am meshing an aircraft quite similar to the Lockheed Tristar.
I finished the (unstructured) surface mesh and assembled the final block.

I've tried two approaches. In my first attempt I had already created and initialised a block inside the engine duct (on the top of the aircraft) so the final block did not include the already created one. They only shared the domain on the engine's inlet. Then, I set up the T-Rex parameters. The symmetry plane and the shared domain (at the engine's inlet) are set to match (push attribute is enabled), everything else is set to wall and the farfield boundaries are unspecified. I set the collision buffer to 2, and changed the maximum angle in the skewness criteria to 120. Boundary decay is 0.85. (I've made sure that the orientation of the domains is correct.)
Then I go to Solve --> Initialize. After the first layer is created I get this message:
"Warning: The creation of full anisotropic layers has stopped for blk-1. Partial anisotropic layers and/or isotropic cells will be created for the remainder of the mesh."
and the choices are to either resume or to create isotropic cells. I chose resume.
After the solver has finished with the layer creation (I get the maximum number of specified layers) the solver actually stops with the message:
"Could not import tets!"
I tried changing various parameters but it seems that it doesn't have to do with the T-Rex solver as every time I do get the layers but it goes wrong just after that.

The next thing I tried was to delete the block inside the engine's duct and assemble just one unstructured block. I didn't use T-Rex this time. Just a plain unstructured block. This approach failed again with a message saying that some entity could not be initialized, tets could not be imported due to poor quality surface mesh and it also gave me the coordinates of point which it seems to be somewhere on the nose of the aircraft. That was somewhat a surprise as the surface mesh in that region is quite good (unstructured, advancing front algorithm, low skewness, very specific edge length (which I had set up in the Solve --> Attributes)).

Right now I'm trying the same as I mentioned just above but using T-Rex and I'll update the post with the result.
Update: This approach was unsuccessful too.... Error: Could not Import Tets! again...

However, I'm still curious as to why the block initialisation fails. I cannot understand what is going wrong in this case.

I would appreciate it if someone could help!


Thanks,

Lefters

ps
Quick update: I'm checking the memory usage and it doesn't seem to be going above 30% up to the point the solver stops.

pps
New Update: This time I did exactly what I did before, only this time I had changed the number of Full Layers to 0. Again after T-Rex finished with the layers, the Iso-Tet solver failed (could not import tets) and it also gave me some feedback this time:
Error: Could not import the tets! Error: Block blk-1:
Error:
Error: Edge:
Error: 19.227013 -1.9838768e-08 0.22998724
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Edge:
Error: 19.228339 -1.9733638e-08 0.22678189
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Edge:
Error: 19.227036 -1.9882948e-08 0.22811863
Error: 19.22571 -2.0055662e-08 0.22865993
Error:
Error: Edge:
Error: 19.226642 -1.9972134e-08 0.22678189
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Edge:
Error: 19.228115 -1.9714556e-08 0.22877648
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error:
Error: Triangle:
Error: 19.226642 -1.9972134e-08 0.22678189
Error: 19.228339 -1.9733638e-08 0.22678189
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Triangle:
Error: 19.226642 -1.9972134e-08 0.22678189
Error: 19.227036 -1.9882948e-08 0.22811863
Error: 19.22571 -2.0055662e-08 0.22865993
Error:
Error: Triangle:
Error: 19.227013 -1.9838768e-08 0.22998724
Error: 19.228115 -1.9714556e-08 0.22877648
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Triangle:
Error: 19.228339 -1.9733638e-08 0.22678189
Error: 19.228115 -1.9714556e-08 0.22877648
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Triangle:
Error: 19.227013 -1.9838768e-08 0.22998724
Error: 19.227036 -1.9882948e-08 0.22811863
Error: 19.22571 -2.0055662e-08 0.22865993
Error:
Error: Triangle:
Error: 14.496479 -1.0338833e-06 1.4004085
Error: 14.495898 -1.02183e-06 1.4006792
Error: 14.49593 -1.0123238e-06 1.4002735
Error: The surface triangulation could not be recovered

None of these points are on a surface, they are just where the T-Rex layers stop (which makes sense since the iso-tet solver crashes right after the T-Rex has finished).

[PeterShi]

Originally Posted by Aeronautics El. K.

Hi all,

I'd like some help (again ).

I am meshing an aircraft quite similar to the Lockheed Tristar.
I finished the (unstructured) surface mesh and assembled the final block.

I've tried two approaches. In my first attempt I had already created and initialised a block inside the engine duct (on the top of the aircraft) so the final block did not include the already created one. They only shared the domain on the engine's inlet. Then, I set up the T-Rex parameters. The symmetry plane and the shared domain (at the engine's inlet) are set to match (push attribute is enabled), everything else is set to wall and the farfield boundaries are unspecified. I set the collision buffer to 2, and changed the maximum angle in the skewness criteria to 120. Boundary decay is 0.85. (I've made sure that the orientation of the domains is correct.)
Then I go to Solve --> Initialize. After the first layer is created I get this message:
"Warning: The creation of full anisotropic layers has stopped for blk-1. Partial anisotropic layers and/or isotropic cells will be created for the remainder of the mesh."
and the choices are to either resume or to create isotropic cells. I chose resume.
After the solver has finished with the layer creation (I get the maximum number of specified layers) the solver actually stops with the message:
"Could not import tets!"
I tried changing various parameters but it seems that it doesn't have to do with the T-Rex solver as every time I do get the layers but it goes wrong just after that.

The next thing I tried was to delete the block inside the engine's duct and assemble just one unstructured block. I didn't use T-Rex this time. Just a plain unstructured block. This approach failed again with a message saying that some entity could not be initialized, tets could not be imported due to poor quality surface mesh and it also gave me the coordinates of point which it seems to be somewhere on the nose of the aircraft. That was somewhat a surprise as the surface mesh in that region is quite good (unstructured, advancing front algorithm, low skewness, very specific edge length (which I had set up in the Solve --> Attributes)).

Right now I'm trying the same as I mentioned just above but using T-Rex and I'll update the post with the result.
Update: This approach was unsuccessful too.... Error: Could not Import Tets! again...

However, I'm still curious as to why the block initialisation fails. I cannot understand what is going wrong in this case.

I would appreciate it if someone could help!


Thanks,

Lefters

ps
Quick update: I'm checking the memory usage and it doesn't seem to be going above 30% up to the point the solver stops.

pps
New Update: This time I did exactly what I did before, only this time I had changed the number of Full Layers to 0. Again after T-Rex finished with the layers, the Iso-Tet solver failed (could not import tets) and it also gave me some feedback this time:
Error: Could not import the tets! Error: Block blk-1:
Error:
Error: Edge:
Error: 19.227013 -1.9838768e-08 0.22998724
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Edge:
Error: 19.228339 -1.9733638e-08 0.22678189
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Edge:
Error: 19.227036 -1.9882948e-08 0.22811863
Error: 19.22571 -2.0055662e-08 0.22865993
Error:
Error: Edge:
Error: 19.226642 -1.9972134e-08 0.22678189
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Edge:
Error: 19.228115 -1.9714556e-08 0.22877648
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error:
Error: Triangle:
Error: 19.226642 -1.9972134e-08 0.22678189
Error: 19.228339 -1.9733638e-08 0.22678189
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Triangle:
Error: 19.226642 -1.9972134e-08 0.22678189
Error: 19.227036 -1.9882948e-08 0.22811863
Error: 19.22571 -2.0055662e-08 0.22865993
Error:
Error: Triangle:
Error: 19.227013 -1.9838768e-08 0.22998724
Error: 19.228115 -1.9714556e-08 0.22877648
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Triangle:
Error: 19.228339 -1.9733638e-08 0.22678189
Error: 19.228115 -1.9714556e-08 0.22877648
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Triangle:
Error: 19.227013 -1.9838768e-08 0.22998724
Error: 19.227036 -1.9882948e-08 0.22811863
Error: 19.22571 -2.0055662e-08 0.22865993
Error:
Error: Triangle:
Error: 14.496479 -1.0338833e-06 1.4004085
Error: 14.495898 -1.02183e-06 1.4006792
Error: 14.49593 -1.0123238e-06 1.4002735
Error: The surface triangulation could not be recovered

None of these points are on a surface, they are just where the T-Rex layers stop (which makes sense since the iso-tet solver crashes right after the T-Rex has finished).

Hey,

Have you solved this problem yet? It will be great if you can share how you solve it.

Best regards,
Peter

[RcktMan77]

If you can share your Pointwise project file, then that would help. When initializing the two blocks separately, you're going to want to turn off the Push Attributes option on the domain(s) shared by both blocks for the 2nd block you initialize. Not doing this is going to blow away the volume grid for the first block.

Before even trying T-Rex, ensure that the area ratio of all of the domains on your surface mesh is below 4.0 for a tri-mesh and below 10.0 for a quad-dominant surface mesh.

Relax your max included angle skewness criteria. 120.0 degrees is way too strict. Start with a value around 165.0 degrees, and relax it to something more typical of 170.0 degrees if 165.0 is still too prohibitive.

One thing you didn't mention was the initial cell height that you're using for wall boundaries. If this value is much lower than the grid point tolerance, then you might want to adjust the grid point tolerance under File -> Properties

If your geometry has sharp edges in which the normal direction is difficult to compute, then that is when you keep Full Layers set to 0 to enable multiple normals. Otherwise, it's usually best to toggle them off by setting Full Layers to 1 (yes, I know it isn't intuitive).

If you're a customer with current maintenance, then you can always use support@pointwise.com for these types of questions.

[dgarlisch]

Original post is from December 4, 2015.

[rookie34]

Originally Posted by Aeronautics El. K.

Hi all,

I'd like some help (again ).

I am meshing an aircraft quite similar to the Lockheed Tristar.
I finished the (unstructured) surface mesh and assembled the final block.

I've tried two approaches. In my first attempt I had already created and initialised a block inside the engine duct (on the top of the aircraft) so the final block did not include the already created one. They only shared the domain on the engine's inlet. Then, I set up the T-Rex parameters. The symmetry plane and the shared domain (at the engine's inlet) are set to match (push attribute is enabled), everything else is set to wall and the farfield boundaries are unspecified. I set the collision buffer to 2, and changed the maximum angle in the skewness criteria to 120. Boundary decay is 0.85. (I've made sure that the orientation of the domains is correct.)
Then I go to Solve --> Initialize. After the first layer is created I get this message:
"Warning: The creation of full anisotropic layers has stopped for blk-1. Partial anisotropic layers and/or isotropic cells will be created for the remainder of the mesh."
and the choices are to either resume or to create isotropic cells. I chose resume.
After the solver has finished with the layer creation (I get the maximum number of specified layers) the solver actually stops with the message:
"Could not import tets!"
I tried changing various parameters but it seems that it doesn't have to do with the T-Rex solver as every time I do get the layers but it goes wrong just after that.

The next thing I tried was to delete the block inside the engine's duct and assemble just one unstructured block. I didn't use T-Rex this time. Just a plain unstructured block. This approach failed again with a message saying that some entity could not be initialized, tets could not be imported due to poor quality surface mesh and it also gave me the coordinates of point which it seems to be somewhere on the nose of the aircraft. That was somewhat a surprise as the surface mesh in that region is quite good (unstructured, advancing front algorithm, low skewness, very specific edge length (which I had set up in the Solve --> Attributes)).

Right now I'm trying the same as I mentioned just above but using T-Rex and I'll update the post with the result.
Update: This approach was unsuccessful too.... Error: Could not Import Tets! again...

However, I'm still curious as to why the block initialisation fails. I cannot understand what is going wrong in this case.

I would appreciate it if someone could help!


Thanks,

Lefters

ps
Quick update: I'm checking the memory usage and it doesn't seem to be going above 30% up to the point the solver stops.

pps
New Update: This time I did exactly what I did before, only this time I had changed the number of Full Layers to 0. Again after T-Rex finished with the layers, the Iso-Tet solver failed (could not import tets) and it also gave me some feedback this time:
Error: Could not import the tets! Error: Block blk-1:
Error:
Error: Edge:
Error: 19.227013 -1.9838768e-08 0.22998724
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Edge:
Error: 19.228339 -1.9733638e-08 0.22678189
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Edge:
Error: 19.227036 -1.9882948e-08 0.22811863
Error: 19.22571 -2.0055662e-08 0.22865993
Error:
Error: Edge:
Error: 19.226642 -1.9972134e-08 0.22678189
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Edge:
Error: 19.228115 -1.9714556e-08 0.22877648
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error:
Error: Triangle:
Error: 19.226642 -1.9972134e-08 0.22678189
Error: 19.228339 -1.9733638e-08 0.22678189
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Triangle:
Error: 19.226642 -1.9972134e-08 0.22678189
Error: 19.227036 -1.9882948e-08 0.22811863
Error: 19.22571 -2.0055662e-08 0.22865993
Error:
Error: Triangle:
Error: 19.227013 -1.9838768e-08 0.22998724
Error: 19.228115 -1.9714556e-08 0.22877648
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Triangle:
Error: 19.228339 -1.9733638e-08 0.22678189
Error: 19.228115 -1.9714556e-08 0.22877648
Error: 19.227036 -1.9882948e-08 0.22811863
Error:
Error: Triangle:
Error: 19.227013 -1.9838768e-08 0.22998724
Error: 19.227036 -1.9882948e-08 0.22811863
Error: 19.22571 -2.0055662e-08 0.22865993
Error:
Error: Triangle:
Error: 14.496479 -1.0338833e-06 1.4004085
Error: 14.495898 -1.02183e-06 1.4006792
Error: 14.49593 -1.0123238e-06 1.4002735
Error: The surface triangulation could not be recovered

None of these points are on a surface, they are just where the T-Rex layers stop (which makes sense since the iso-tet solver crashes right after the T-Rex has finished).

Hey,

Could you solve the problem?

[Aeronautics El. K.]

Wow that's almost ancient history...
Even if I had solved that, I really don't recall anything

[rookie34]

Okay, thanks for your reply.