Replies: 11 comments
-
|
I had a similar issue once and found this article https://cfdmonkey.com/snappyhexmesh/ If you turn off the mesh quality dictionary for the boundary layer generation part, it might work |
Beta Was this translation helpful? Give feedback.
-
|
I tried as described, but it didn't help. |
Beta Was this translation helpful? Give feedback.
-
|
Hello @gocfd. This is Donggyu from NEXTFOAM. Could you check the size of the boundary layer? If you do not check "Relative", it means you will generate the layer with absoulte unit (m). Best regards Donggyu kim |
Beta Was this translation helpful? Give feedback.
-
|
Hello Donggyu. the whole size of domain is fi 0.12 x 0.07 m. Max. blade thickness is around 1,5 mm. Best regards. |
Beta Was this translation helpful? Give feedback.
-
|
Hello @gocfd. I think the size is toooo small to configure whether it generates properly. Best regards Donggyu Kim |
Beta Was this translation helpful? Give feedback.
-
|
Hello, Could you share one document giving clearly which parameter is where in snappyHexMesh compared to BaramMesh, it is difficult to follow user CFD online forum advises. Thank you Julien |
Beta Was this translation helpful? Give feedback.
-
|
Hi Julien,
Thank you. Jake |
Beta Was this translation helpful? Give feedback.
-
|
Hello, I succeed to produre boundary layer on more than 98% of the wall surfaces. No illegal faces detected. But the simulation crashes everytime, without boundary layers it runs but 0.001 convergence is not reached. I hoped that with BL convergence will be better with more cells close to the surfaces walls... See the checkMesh message output:
CheckMesh output: Create polyMesh for time = 0 Time = 0 Mesh stats Overall number of cells of each type: Checking topology... Checking patch topology for multiply connected surfaces... Checking geometry... Mesh OK. End What is wrong with my mesh. I use ke model with first cells y+ around 20 with around 6-8 cells 30<y+<100. But simulations crashes everytimes it first steps. I try to reduce relaxation for p and U to 0.1 and less but it crashes... I initialize the case before start the run. Do you have an idea how to improve the convergence of my simulation? Julien |
Beta Was this translation helpful? Give feedback.
-
Hello,I see that during the boundary meshing operation a lot of messages from baramMesh:[1] Cannot find point in pts1 matching point 247 coord:(-0.170000001788 0.405480425711 0.154651596094) in pts0 when using tolerance 2.09595597196e-007 ................... Do you think that it could correspond to some meshing error or bad mesh quality?I see also this message at the end of the meshing sequence: it gives some bad edges, but it has not impact on the mesh quality output...Create time Other question where is writen the log of the meshing operation?Best regards Julien |
Beta Was this translation helpful? Give feedback.
-
|
I check the internal fanand porous jump patches and I see that there is a gap between with the wall corresponding to the thickness of the boundary layers. Do you think it is a display bug or the boundary layer is not included into the patch? Maybe that's why the simulation doesn't want to start betcause of these gaps everwhere around patched inside the duct walls. Picture from baramMesh Internal mesh at the patch location Geometric STL patches Picture from baramFlow |
Beta Was this translation helpful? Give feedback.
-
|
Maybe I could modify the fvSolution or fvScheme files to improve the stability of the convergence? |
Beta Was this translation helpful? Give feedback.
-
Hello,
I am facing problems with boundary layer generation on a fan blade. I am trying to figure out the optimal parameters for boundary layer mesh, but I am unable to do so. Layers on balde surface are not generated at all.
Here are the boundary layer settings and generated mesh:
Thank you for your support.
Best regards,
G
Beta Was this translation helpful? Give feedback.
All reactions