oscilated wallHeatFlux on a curved surface

nucat · January 16, 2015, 12:01

Hallo foamers,

I have a problem with the postprocessing of the wallHeatFlux on my curved surface patch. On non curved surface patch the wallHeatFlux looks good. But if the patch become curved, the wallHeatFlux oscillating.

Here is the visualized image of the problem:
http://picload.org/image/ccpcigp/wallheatflux.png

Did anyone you had the same problem and know how to solv it or what to try? I am using as snGrad scheme "orthogonal fourth", but also tried other schemes.

I am using OF 2.1.1. I read the change log of OF2.3.0. There I read that there are two new schemes: quadraticFit and linearFit. It sound good and I wanted to tray them for an snGrad scheme. I didn't work because that schemes need an argument in the fvSchmes file and I do not know which kind of. Do anybody using that scheme?

Thanks and regards,
Nucat

nucat · January 18, 2015, 11:17

Hey,

I understood the problem. The reason for this oscillation is the small thickness of the first boundary cell. If the boundary cells are too small, you will get oscillation while using snGrad(T) or grad(T).

Regards,
Nucat

ssss · April 16, 2015, 20:28

I'm also interested in lowering the oscillations that appear when discretizing the surface normal gradient, in a curve patch and with a very thin first layer of cells. Has anyone been able of reducing this oscillations?

Thank you very much

January 16, 2015, 12:01	oscilated wallHeatFlux on a curved surface	#1
nucat New Member Join Date: Jan 2013 Posts: 10 Rep Power: 13	Hallo foamers, I have a problem with the postprocessing of the wallHeatFlux on my curved surface patch. On non curved surface patch the wallHeatFlux looks good. But if the patch become curved, the wallHeatFlux oscillating. Here is the visualized image of the problem: http://picload.org/image/ccpcigp/wallheatflux.png Did anyone you had the same problem and know how to solv it or what to try? I am using as snGrad scheme "orthogonal fourth", but also tried other schemes. I am using OF 2.1.1. I read the change log of OF2.3.0. There I read that there are two new schemes: quadraticFit and linearFit. It sound good and I wanted to tray them for an snGrad scheme. I didn't work because that schemes need an argument in the fvSchmes file and I do not know which kind of. Do anybody using that scheme? Thanks and regards, Nucat

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[ICEM] surface mesh merging problem	everest	ANSYS Meshing & Geometry	44	April 14, 2016 07:41
[Gmsh] Layers in a curved surface	Artex85	OpenFOAM Meshing & Mesh Conversion	1	December 18, 2011 09:51
Plotting the velocity profile in normal direction along a curved surface	StSchiff	OpenFOAM	0	December 1, 2011 06:01
[ICEM] Automatic mesh generation script surface intersection problem	stuart23	ANSYS Meshing & Geometry	0	May 13, 2011 02:10
How to generate grid system with a curved surface	Jing Ren	Siemens	3	March 13, 2001 12:58

January 18, 2015, 11:17		#2
nucat New Member Join Date: Jan 2013 Posts: 10 Rep Power: 13	Hey, I understood the problem. The reason for this oscillation is the small thickness of the first boundary cell. If the boundary cells are too small, you will get oscillation while using snGrad(T) or grad(T). Regards, Nucat

April 16, 2015, 20:28		#3
ssss Senior Member anonymous Join Date: Aug 2014 Posts: 205 Rep Power: 13	I'm also interested in lowering the oscillations that appear when discretizing the surface normal gradient, in a curve patch and with a very thin first layer of cells. Has anyone been able of reducing this oscillations? Thank you very much