[GAMBIT] Can't refine face mesh enough

mbeals · April 3, 2009, 14:37

I am attempting to model an aircraft probe that is a rather complex shape. I have an image below:

I have manipulated the face meshes to get worst element skewness to around 0.7, but due to the sharp tips (below), I can't improve beyond that. This is with the tips being meshed on a 0.25 spacing. I have a boundary layer attached to the faces of the entire probe, which starts at .5 and grows at 1.4 for 8 layers. When I attempt to volume mesh this (subtracted from a cube), using T-grid, it always fails, telling me to check skewness in areas of small gaps. No amount of refinement seems to get it to work. The only way I have successfully meshed it was when I used stairstep. However that ignores the boundary layer and produces a highly skewed volume mesh. Any tips or tricks on how to approach this would be greatly appreciated.

-mAx- · April 3, 2009, 14:55

Merge the 2 faces in front of your geometry, it will delete the diagonal edge which generates the small angle (and produces high skewness).
Do the same at the adjacent side (right side on the picture) with the other diagonal edge.

mbeals · April 3, 2009, 15:48

awesome, thanks.

Quick question. Could I merge the other two flat faces with the front one too without losing the shape of the tip?

-mAx- · April 3, 2009, 16:02

Gambit won't allow you to do that, because the edge is needed to describe the shape.
What should be the aim for this? if the diagonals are gone, you are in the case of 90° corner (ideal)
But you can try (save before)

April 3, 2009, 14:37	Can't refine face mesh enough	#1
mbeals New Member Matthew Join Date: Apr 2009 Posts: 12 Rep Power: 17	I am attempting to model an aircraft probe that is a rather complex shape. I have an image below: I have manipulated the face meshes to get worst element skewness to around 0.7, but due to the sharp tips (below), I can't improve beyond that. This is with the tips being meshed on a 0.25 spacing. I have a boundary layer attached to the faces of the entire probe, which starts at .5 and grows at 1.4 for 8 layers. When I attempt to volume mesh this (subtracted from a cube), using T-grid, it always fails, telling me to check skewness in areas of small gaps. No amount of refinement seems to get it to work. The only way I have successfully meshed it was when I used stairstep. However that ignores the boundary layer and produces a highly skewed volume mesh. Any tips or tricks on how to approach this would be greatly appreciated.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Convergence problem when refine the mesh...	TB	CFX	37	May 13, 2019 19:50
[Netgen] Import netgen mesh to OpenFOAM	hsieh	OpenFOAM Meshing & Mesh Conversion	32	September 13, 2011 06:50
[mesh manipulation] TranformPoints gives skewed mesh Possible Bug	andersking	OpenFOAM Meshing & Mesh Conversion	3	March 25, 2008 22:33
Numerical calculation interFoam never finishes	unoder	OpenFOAM Running, Solving & CFD	9	October 18, 2005 06:00
how to increase face mesh no.?	Shankar	FLUENT	4	August 9, 2002 16:48

April 3, 2009, 14:55		#2
-mAx- Super Moderator Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 3,297 Rep Power: 41	Merge the 2 faces in front of your geometry, it will delete the diagonal edge which generates the small angle (and produces high skewness). Do the same at the adjacent side (right side on the picture) with the other diagonal edge.

April 3, 2009, 15:48		#3
mbeals New Member Matthew Join Date: Apr 2009 Posts: 12 Rep Power: 17	awesome, thanks. Quick question. Could I merge the other two flat faces with the front one too without losing the shape of the tip?

April 3, 2009, 16:02		#4
-mAx- Super Moderator Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 3,297 Rep Power: 41	Gambit won't allow you to do that, because the edge is needed to describe the shape. What should be the aim for this? if the diagonals are gone, you are in the case of 90° corner (ideal) But you can try (save before)