[snappyHexMesh] castellatedMesh doesn't refine in y direction

KTG · January 22, 2021, 21:02

Hi everyone,

I am having a hard time imagining this is not a bug. When I get past two levels of refinement, snappy stops refining in the y direction! Its clear in the attached image - the right face has squares, the left has rectangles. I can't imaging what is causing this. The block mesh is all cubes. Someone please give me a sanity check...

Thanks

KTG · January 22, 2021, 22:42

Ok, I figured this out and it feels like a cruel joke. I was working in Geographic coordinates with bounding box (715898 3.8793e+06 2.79123) (716766 3.88e+06 26.3115), and its looks like the size of the coordinates in the y direction overloaded something. I shifted everything back to the origin, and the problem is gone. As far as I can tell, snappy never threw an error.

I am not a C/C++ expert, but it seems strange that e+06 would cause this, but e+05 wouldn't.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[Other] Contribution a new utility: refine wall layer mesh based on yPlus field	lakeat	OpenFOAM Community Contributions	58	December 23, 2021 02:36
atmospheric boundary layer (ABL) - implementing wind direction	rob3rt 0ng	OpenFOAM Running, Solving & CFD	0	November 4, 2015 20:57
ERROR: Flow direction on the boundaries must not be tangential to the boundary.	turbomax	CFX	2	August 6, 2015 11:05
[DesignModeler] DesignModeler Scripting: Direction vector in extrude command	Tobias_H	ANSYS Meshing & Geometry	3	July 27, 2015 18:36
[ICEM] Changing edge direction	la7low	ANSYS Meshing & Geometry	2	June 7, 2010 13:26

January 22, 2021, 22:42		#2
KTG Senior Member Abe Join Date: May 2016 Posts: 119 Rep Power: 9	Ok, I figured this out and it feels like a cruel joke. I was working in Geographic coordinates with bounding box (715898 3.8793e+06 2.79123) (716766 3.88e+06 26.3115), and its looks like the size of the coordinates in the y direction overloaded something. I shifted everything back to the origin, and the problem is gone. As far as I can tell, snappy never threw an error. I am not a C/C++ expert, but it seems strange that e+06 would cause this, but e+05 wouldn't.