Pressure far field boundary conditions not respected?!

Bedotto · August 5, 2010, 13:45

Hello everyone,

I'm working on the expansion of a thermical jet of a motor. To do this I've created a mesh in 2d of the profile of this motor (see the scheme about boundary attached)). But I have an odd issue. The parameters choosen of the pressure far field boundary conditions are not respected. Here an example: I put axial component=1 et radial component=0 but after many iterations when I plot velocity vectors on theses pressure far field boundary I get vectors which haven't the same components that I fixed in parameters panel.

I've tried to modify the mesh on boundary conditions (more or less coarsed), and I tried to expand the limits of my numerical domain but I still have the same problem. I really don't understand the origin of this problem. Does anyone here have faced this kind of problem with Fluent?

(I work in axisymmetric with Spalart Almaras.)

My best regards

Quentin

doki · August 5, 2010, 16:07

hi
are you using the cartesian or radial components?
I have had such a problem and I realized thad fluent does not identify the radial and axial directions as we see!
I have fixed the problem using cartesian components of the flow.
So I guess that would be helpful to set the direction in cartesian coordinates.

davedong · August 5, 2010, 22:39

doki is correct.
i get some tricks for this problem. the pressure-far-field bound is pretty, just seems... if your mesh is narrow or unreasonbale, reversed flow and oddy boundary data will crazy you.
you can build the mesh wisely and widely enough (even in the supersonic aircraft problem, the width of tunnel is just 30 times than the wings.). then, you can set the bound as wall.

Bedotto · August 16, 2010, 04:30

doki: I would like to use cartesian coordinates too but because I work in axis symmetry, I've not the choice, I have to work with axial and radial coordinates when I define the direction of velocity on far field pressure boundary conditions. It seems I can't work with cartesian coordinates. Or am I wrong?

davedong: I'm not sure to understand: I should "stretch" my mesh, that's what you mean? Or the contrary?

davedong · August 17, 2010, 05:42

hi,
i'm sorry for your misunderstanding. english is not my native language.
i mean maybe you can stretch your computing domain. if it is big enough, the boundary set is not determinative.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
CFX doesn't continue calculation...	mactech001	CFX	6	November 15, 2009 22:25
Pressure boundary conditions	Vijay	FLUENT	2	January 14, 2009 13:24
Neumann pressure BC and velocity field	Antech	Main CFD Forum	0	April 25, 2006 03:15
Please help with flow around car modelling!	Tudor Miron	CFX	17	March 19, 2004 20:23
Boundary conditions?	Tom	Main CFD Forum	0	November 5, 2002 02:54

August 5, 2010, 16:07		#2
doki Member Join Date: Apr 2009 Posts: 41 Rep Power: 17	hi are you using the cartesian or radial components? I have had such a problem and I realized thad fluent does not identify the radial and axial directions as we see! I have fixed the problem using cartesian components of the flow. So I guess that would be helpful to set the direction in cartesian coordinates.

August 5, 2010, 22:39		#3
davedong New Member Dong Liang Join Date: Jan 2010 Location: Shanghai, China Posts: 12 Rep Power: 16	doki is correct. i get some tricks for this problem. the pressure-far-field bound is pretty, just seems... if your mesh is narrow or unreasonbale, reversed flow and oddy boundary data will crazy you. you can build the mesh wisely and widely enough (even in the supersonic aircraft problem, the width of tunnel is just 30 times than the wings.). then, you can set the bound as wall.

August 16, 2010, 04:30		#4
Bedotto New Member Quentin Join Date: May 2009 Posts: 22 Rep Power: 17	doki: I would like to use cartesian coordinates too but because I work in axis symmetry, I've not the choice, I have to work with axial and radial coordinates when I define the direction of velocity on far field pressure boundary conditions. It seems I can't work with cartesian coordinates. Or am I wrong? davedong: I'm not sure to understand: I should "stretch" my mesh, that's what you mean? Or the contrary?

August 17, 2010, 05:42		#5
davedong New Member Dong Liang Join Date: Jan 2010 Location: Shanghai, China Posts: 12 Rep Power: 16	hi, i'm sorry for your misunderstanding. english is not my native language. i mean maybe you can stretch your computing domain. if it is big enough, the boundary set is not determinative.