Interpretation of pressure icoFoam

billy · May 5, 2007, 23:36

Hi,

I was running icoFoam with lid-drive cvavity and I noticed that pressure changed in same degree, as I refined the mesh. Doubling the cells in xy direction, the pressure doubles also. Documentation says:

"In a closed incompressible system such as the cavity, pressure is relative: it is the pressure range that matters not the absolute values"

But why does the pressure range change? With 10x10 grid pressure range is -20-20, with 20x20 grid it is -40-40, with 40x40 it is -80-80. Shouldn't it converge to a certain range?

Thanks in advance,

Billy.

billy · May 6, 2007, 10:46

Increasing even more, using a 60x60 grid gives very high pressure range like -1E+4-1E+4.

Billy.

billy · May 6, 2007, 19:15

I also tried a grid with one and two layers, the pressure is increases in the same degree that the layers increase.

billy · May 6, 2007, 19:15

I also tried a grid with one and two layers, the pressure is increases in the same degree that the layers increase.

billy · May 7, 2007, 11:16

The figure shows pressure contours around the inlet. The outlet is located on other side.
Both domains have same thickness.

billy · May 7, 2007, 15:41

Basically the pressure seems to increase at the same rate the volume of the cell decreases.

billy · May 8, 2007, 17:45

Does anyone know why this happens? Don't want this topic to die.

alberto · May 8, 2007, 17:51

Please, post a case. I'll try to run it.

A.

rafal · May 8, 2007, 18:00

I would think about two things:
1. where is the reference pressure cell and
what is the value you are setting in your case.
2. Dont you think that it would be better to
have more dense mesh in the region of big
gradients(4 cells is not enough, remember you
calculate everything on discrete values).

hope this helps
rafal

billy · May 8, 2007, 18:38

Thanks. Two test cases are included: one and two layer hex meshes.

hexlayer1.tgz

billy · May 8, 2007, 18:54

I wasn't able to include second case because it is bigger than 50k. But if you have Gmsh, the number of layer can be changed edting line 3:

l = 2.

Then running gmshToFoam.

billy · May 8, 2007, 19:12

You may have to set inlet velocity like

0.1, 0, 0.

billy · May 8, 2007, 19:15

Patch 0 is inlet,
Patch 1 is outlet,
Patch 2 is wall,
Patch 3 is wall.

Rafael,

I don't think it has to do with mesh density. The figures were done with interFoam and two fluids but this happens also with the icoFoam case I posted and the pressure variation is smoother.

billy · May 8, 2007, 19:20

I post two more figures with icoFoam.

billy · May 8, 2007, 19:30

I believe the reference pressure is only important in closed cavities. This case the outlet is set to zero.

billy · May 9, 2007, 17:42

The gradient seems ok. The problem is the value. Anyone have suggestions why this happens? It almost looks like the pressure is per layer.

joakim · May 9, 2007, 18:06

Hi Billy

I should state that I do not know the details of your simulation, but I must ask what your are trying to simulate:

1) From your pictures, it looks to me that your are trying to run a 2D simulation. Why use more than one layer?

2) Have you reached grid-independence? You have walls in your simulation, but I can't see that you have resolved any boundarylayers. When
you change the mesh-density, how does the velocities change?

Best regards

/Joakim

billy · May 9, 2007, 19:44

1) I am trying to simulate the flow inside a thin plate. I want to use several layers because I might have a temperature variation across thickness that I want to capture. This way I would solve equations for flow as well as temperature. However, I want to know also how much pressure is needed, so pressure shouldn't vary so much with the amount of layers I have.

2) This is for laminar flow not turbulent flow. I don't think this difference is due to grid indpendence, there is no cross thickness flow and I don't think pressure should be 1x, 2x, 3x higher as I refine the mesh.

With 3 layers, pressure is 3 times higher. I think this has to do with the pressure correction scheme.

joakim · May 9, 2007, 23:41

Hi Billy

Isn't your problem a 3D problem?

I would assume that you get a parabolic flow profile in the xy-plane between the plates. Your strange results is probobly due to your b.c. I assume that you are currently using empty as b.c. at bottom and top since you started with one layer. Shouldn't you use wall- or atleat slipwall conditions instead and something like 10 cells in the z-direction?

Regards

/Joakim

billy · May 10, 2007, 10:35

Actually I tried with 10 layers and it blows up. Courant just increases to infinity. There is an inlet, outlet and all other surfaces are walls. I will try another geometry to see what happens.

Billy.

May 5, 2007, 23:36	Hi, I was running icoFoam w	#1
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	Hi, I was running icoFoam with lid-drive cvavity and I noticed that pressure changed in same degree, as I refined the mesh. Doubling the cells in xy direction, the pressure doubles also. Documentation says: "In a closed incompressible system such as the cavity, pressure is relative: it is the pressure range that matters not the absolute values" But why does the pressure range change? With 10x10 grid pressure range is -20-20, with 20x20 grid it is -40-40, with 40x40 it is -80-80. Shouldn't it converge to a certain range? Thanks in advance, Billy.

May 6, 2007, 10:46	Increasing even more, using a	#2
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	Increasing even more, using a 60x60 grid gives very high pressure range like -1E+4-1E+4. Billy.

May 6, 2007, 19:15	I also tried a grid with one a	#3
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	I also tried a grid with one and two layers, the pressure is increases in the same degree that the layers increase.

May 6, 2007, 19:15	I also tried a grid with one a	#4
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	I also tried a grid with one and two layers, the pressure is increases in the same degree that the layers increase.

May 7, 2007, 11:16	The figure shows pressure cont	#5
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	The figure shows pressure contours around the inlet. The outlet is located on other side. Both domains have same thickness.

May 7, 2007, 15:41	Basically the pressure seems t	#6
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	Basically the pressure seems to increase at the same rate the volume of the cell decreases.

May 8, 2007, 17:45	Does anyone know why this happ	#7
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	Does anyone know why this happens? Don't want this topic to die.

May 8, 2007, 17:51	Please, post a case. I'll try	#8
alberto Senior Member Alberto Passalacqua Join Date: Mar 2009 Location: Ames, Iowa, United States Posts: 1,912 Rep Power: 36	Please, post a case. I'll try to run it. A. __________________ Alberto Passalacqua GeekoCFD - A free distribution based on openSUSE 64 bit with CFD tools, including OpenFOAM. Available as in both physical and virtual formats (current status: http://albertopassalacqua.com/?p=1541) OpenQBMM - An open-source implementation of quadrature-based moment methods. To obtain more accurate answers, please specify the version of OpenFOAM you are using.

May 8, 2007, 18:00	I would think about two things	#9
rafal Member rafal zietara Join Date: Mar 2009 Location: Manchester, UK Posts: 60 Rep Power: 17	I would think about two things: 1. where is the reference pressure cell and what is the value you are setting in your case. 2. Dont you think that it would be better to have more dense mesh in the region of big gradients(4 cells is not enough, remember you calculate everything on discrete values). hope this helps rafal

May 8, 2007, 18:38	Thanks. Two test cases are inc	#10
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	Thanks. Two test cases are included: one and two layer hex meshes. hexlayer1.tgz

May 8, 2007, 18:54	I wasn't able to include secon	#11
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	I wasn't able to include second case because it is bigger than 50k. But if you have Gmsh, the number of layer can be changed edting line 3: l = 2. Then running gmshToFoam.

May 8, 2007, 19:12	You may have to set inlet velo	#12
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	You may have to set inlet velocity like 0.1, 0, 0.

May 8, 2007, 19:15	Patch 0 is inlet, Patch 1 is	#13
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	Patch 0 is inlet, Patch 1 is outlet, Patch 2 is wall, Patch 3 is wall. Rafael, I don't think it has to do with mesh density. The figures were done with interFoam and two fluids but this happens also with the icoFoam case I posted and the pressure variation is smoother.

May 8, 2007, 19:20	I post two more figures with i	#14
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	I post two more figures with icoFoam.

May 8, 2007, 19:30	I believe the reference pressu	#15
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	I believe the reference pressure is only important in closed cavities. This case the outlet is set to zero.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Two fundamental questions about icoFoam while updating the velocities and pressure	dbxmcf	OpenFOAM Running, Solving & CFD	6	April 12, 2016 16:05
Density in icoFoam Densidad en icoFoam	manuel	OpenFOAM Running, Solving & CFD	8	September 22, 2010 04:10
UDF Interpretation	Suvash	FLUENT	0	October 31, 2005 23:10
UDF interpretation trough TUI	David	FLUENT	2	March 18, 2005 14:26
Result interpretation	Pilli	FLUENT	2	November 19, 2000 16:58

May 9, 2007, 17:42	The gradient seems ok. The pro	#16
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	The gradient seems ok. The problem is the value. Anyone have suggestions why this happens? It almost looks like the pressure is per layer.

May 9, 2007, 18:06	Hi Billy I should state tha	#17
joakim New Member Joakim Möller Join Date: Mar 2009 Posts: 26 Rep Power: 17	Hi Billy I should state that I do not know the details of your simulation, but I must ask what your are trying to simulate: 1) From your pictures, it looks to me that your are trying to run a 2D simulation. Why use more than one layer? 2) Have you reached grid-independence? You have walls in your simulation, but I can't see that you have resolved any boundarylayers. When you change the mesh-density, how does the velocities change? Best regards /Joakim

May 9, 2007, 19:44	1) I am trying to simulate the	#18
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	1) I am trying to simulate the flow inside a thin plate. I want to use several layers because I might have a temperature variation across thickness that I want to capture. This way I would solve equations for flow as well as temperature. However, I want to know also how much pressure is needed, so pressure shouldn't vary so much with the amount of layers I have. 2) This is for laminar flow not turbulent flow. I don't think this difference is due to grid indpendence, there is no cross thickness flow and I don't think pressure should be 1x, 2x, 3x higher as I refine the mesh. With 3 layers, pressure is 3 times higher. I think this has to do with the pressure correction scheme.

May 9, 2007, 23:41	Hi Billy Isn't your problem	#19
joakim New Member Joakim Möller Join Date: Mar 2009 Posts: 26 Rep Power: 17	Hi Billy Isn't your problem a 3D problem? I would assume that you get a parabolic flow profile in the xy-plane between the plates. Your strange results is probobly due to your b.c. I assume that you are currently using empty as b.c. at bottom and top since you started with one layer. Shouldn't you use wall- or atleat slipwall conditions instead and something like 10 cells in the z-direction? Regards /Joakim

May 10, 2007, 10:35	Actually I tried with 10 layer	#20
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	Actually I tried with 10 layers and it blows up. Courant just increases to infinity. There is an inlet, outlet and all other surfaces are walls. I will try another geometry to see what happens. Billy.