Boundary condition for conic structure for DES simulations

quarkz · May 13, 2020, 22:19

Hi,

I need to simulate a conical structure for DES simulations using OpenFOAM. Instead of simulating an entire cone, I'm thinking of simulating half, a quarter or even just a wedge.

1st of all, is this acceptable for DES simulation?

Next, for the boundary conditions at the interface, what should be the boundary condition?

Should it be symmetry or cyclic or periodic?

Lastly, is a 10 or 25 degree wedge acceptable for DES?

Or does it depend on the problem? Then I may need to do a grid convergent test to find out.

LuckyTran · May 14, 2020, 01:36

You need two-point correlations in order to determine what ° is acceptable. Because the correlation is some finite distance, it tends to be violated very easily near the axis. So in general, you shouldn't be doing sectors. But if you must,the cyclic/periodic BC is much more acceptable. The symmetry boundary condition adds an impermeability condition which severely constrains what can happen there.

FMDenaro · May 14, 2020, 02:58

A pure RANS simulation could be performed using a half geometry but the LES contribution in DES requires the full 3D case.

MikeBravo · May 14, 2020, 20:15

Hi quarkz,
No, that is not really acceptable, for DES or any scale resolving simulation you should have the full 3D geometry. You could do a quarter or a half with rotational periodic interfaces, but why not do the full geometry? Are you limited with computing resources?

quarkz · May 14, 2020, 21:17

thanks for the recommendations. yes, i'm indeed limited by computing resources. nevertheless, i think i will do a cyclic one and then a full 3d and compare.

MikeBravo · May 15, 2020, 19:32

Hi quarkz,

I thought about this a little more, and you may have trouble with the interfaces sharing the same edge, which they would at the axis. I'm not sure if OpenFoam would have a way of handling that. Alternative would be free-slip walls, which is further from reality.

Honestly, I would trust a full 3D DES on a coarser mesh and run for a shorter flow time before a segmented DES on a fine mesh and a longer flow time. I strongly recommend against anything less than the full 3D. If you're looking to get any useful results from this, that is the way to go.

May 13, 2020, 22:19	Boundary condition for conic structure for DES simulations	#1
quarkz Senior Member TWB Join Date: Mar 2009 Posts: 400 Rep Power: 19	Hi, I need to simulate a conical structure for DES simulations using OpenFOAM. Instead of simulating an entire cone, I'm thinking of simulating half, a quarter or even just a wedge. 1st of all, is this acceptable for DES simulation? Next, for the boundary conditions at the interface, what should be the boundary condition? Should it be symmetry or cyclic or periodic? Lastly, is a 10 or 25 degree wedge acceptable for DES? Or does it depend on the problem? Then I may need to do a grid convergent test to find out.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Centrifugal fan	j0hnny	CFX	13	October 1, 2019 13:55
Out File does not show Imbalance in %	Mmaragann	CFX	5	January 20, 2017 10:20
An error has occurred in cfx5solve:	volo87	CFX	5	June 14, 2013 17:44
External Radiation Boundary Condition for Grid Interface	CFD XUE	FLUENT	0	July 9, 2010 02:53
External Radiation Boundary Condition (Two sided wall), Grid Interface	CFD XUE	FLUENT	0	July 8, 2010 06:49

May 14, 2020, 01:36		#2
LuckyTran Senior Member Lucky Join Date: Apr 2011 Location: Orlando, FL USA Posts: 5,674 Rep Power: 66	You need two-point correlations in order to determine what ° is acceptable. Because the correlation is some finite distance, it tends to be violated very easily near the axis. So in general, you shouldn't be doing sectors. But if you must,the cyclic/periodic BC is much more acceptable. The symmetry boundary condition adds an impermeability condition which severely constrains what can happen there.

May 14, 2020, 02:58		#3
FMDenaro Senior Member Filippo Maria Denaro Join Date: Jul 2010 Posts: 6,768 Rep Power: 71	A pure RANS simulation could be performed using a half geometry but the LES contribution in DES requires the full 3D case.

May 14, 2020, 20:15		#4
MikeBravo New Member Mike Join Date: Jan 2018 Posts: 29 Blog Entries: 6 Rep Power: 8	Hi quarkz, No, that is not really acceptable, for DES or any scale resolving simulation you should have the full 3D geometry. You could do a quarter or a half with rotational periodic interfaces, but why not do the full geometry? Are you limited with computing resources?

May 14, 2020, 21:17		#5
quarkz Senior Member TWB Join Date: Mar 2009 Posts: 400 Rep Power: 19	thanks for the recommendations. yes, i'm indeed limited by computing resources. nevertheless, i think i will do a cyclic one and then a full 3d and compare.

May 15, 2020, 19:32		#6
MikeBravo New Member Mike Join Date: Jan 2018 Posts: 29 Blog Entries: 6 Rep Power: 8	Hi quarkz, I thought about this a little more, and you may have trouble with the interfaces sharing the same edge, which they would at the axis. I'm not sure if OpenFoam would have a way of handling that. Alternative would be free-slip walls, which is further from reality. Honestly, I would trust a full 3D DES on a coarser mesh and run for a shorter flow time before a segmented DES on a fine mesh and a longer flow time. I strongly recommend against anything less than the full 3D. If you're looking to get any useful results from this, that is the way to go.