CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > OpenFOAM > OpenFOAM Running, Solving & CFD

p_rgh issues with interFoam (2.0.x)

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   November 22, 2011, 13:37
Default p_rgh issues with interFoam (2.0.x)
  #1
Member
 
Join Date: May 2009
Posts: 54
Rep Power: 17
gfilip is on a distinguished road
I am simulating a simple 2D laminar flow of a submerged cylinder near a free-surface (domain figure attached). After testing many pressure and velocity boundary conditions, I cannot obtain a reasonable p_rgh field with interFoam. The stagnation pressure should be somewhere around 500, but the p_rgh throughout the water fraction is orders of magnitude greater (see attached screenshot). If I set the gravity vector to zero, I get a reasonable p_rgh result where the stagnation pressure matches the expected value (see attached). Is this be due to the way the gravity term is treated in interFoam?

Some things I have tried in various combinations:
-pRefValue in both water and air and zeroGradient for p on boundaries
-buoyantPressure on the cylinder, outlet, both inlets
-totalPressure on the outlet and atmosphere
-fixedValue 0 for p on outlet
-extending the domain (the one pictured is ~100D upstream and downstream)
-cAlpha of 0, 1, 2
-p_rgh and p_rghFinal relaxation factors of 0.3 and off
-cranking up the p tol to e-12
-orienting the mesh so that the g vector is either in -y or -z (desperation mode)

I would appreciate any help.
Attached Images
File Type: jpg domain.jpg (20.4 KB, 165 views)
File Type: jpg p_rgh.jpg (16.3 KB, 151 views)
File Type: jpg p_rgh_noG.jpg (14.8 KB, 141 views)
File Type: jpg alpha.jpg (14.4 KB, 129 views)
gfilip is offline   Reply With Quote

Old   November 22, 2011, 14:03
Default
  #2
ngj
Senior Member
 
Niels Gjoel Jacobsen
Join Date: Mar 2009
Location: Copenhagen, Denmark
Posts: 1,902
Rep Power: 37
ngj will become famous soon enoughngj will become famous soon enough
Hi Grezgorz

Warning: Only using 1.6, but hope this helps anyway.

Looking at your results, I can think of a possible reason. Is your free surface positioned at z = 0 m? In the momentum equation, the following will dominate the reference value of p_rgh

\mathbf{x}\boldsymbol{\cdot}\mathbf{g}\nabla\rho

In the case z = 0 m, then the term \mathbf{x}\boldsymbol{\cdot}\mathbf{g} vanishes at the surface, which is the only place, where \nabla \rho is non-zero. If z is non-zero, this results effectively in a excess pressure in the water column which differs zero and thus produce the depicted results. There are two possible solutions:

1. Move the free surface.

2. Substract some far field p_rgh from your field in e.g. paraView and check that the stagnation pressure is in the correct order of magnitude.

Good luck,

Niels
ngj is offline   Reply With Quote

Old   November 22, 2011, 16:48
Default
  #3
Member
 
Join Date: May 2009
Posts: 54
Rep Power: 17
gfilip is on a distinguished road
Niels,

You are right, setting the f-s at z=0 did the trick! Thank you very much.


Greg
gfilip is offline   Reply With Quote

Old   April 17, 2012, 08:05
Question problems with interFOAM
  #4
New Member
 
Sam-CFD's Avatar
 
Sam Mathew
Join Date: Apr 2010
Location: India
Posts: 19
Rep Power: 16
Sam-CFD is on a distinguished road
Hi,

I was trying a simple tank draining problem and was trying to specify the right boundary conditions. I am implementing my case in 2.1.0 and it is basically a tank with water filled up to a specific height with a small outlet port at the bottom center of the tank.

At the top I gave the same boundary conditions as in the Dam_break case.

I am having difficulty with specifying appropriate boundary conditions for the outlet port. I have been basically playing around with values for alpha, p_rgh and U.

1. For p_rgh = 0 (as suggested here): The water does not flow out of the domain. The other parameters were U = (0,0,0), zeroGradient and alpha = {InletOutlet}, 0.

2. For p_rgh - zeroGradient: The flow is going out but still the results are quite strange. While keeping alpha = {InletOutlet} and U - zeroGradient, the water suddenly jumps up and after a few time steps, it starts draining. It seems as if, the solver detects a high pressure at the outlet port and causes flow to occur from higher to lower pressure, but then realizes actually there is gravity acting against and then the liquid flows out.

The reason I gave p_rgh to be zeroGradient is because I understood that it is the dynamic pressure and the dynamic pressure at the outlet cannot be zero but rather the normal gradient should be zero.

I would be thankful for any help since I am not able to grasp yet the right implementation. In other solvers (like FLUENT, CFX), I would just specify the static pressure to be zero at the outlet with the possibility for reverse flow of air into the domain.
Sam-CFD is offline   Reply With Quote

Old   April 18, 2012, 00:30
Default p_rgh and initial velocity specification
  #5
New Member
 
Sam-CFD's Avatar
 
Sam Mathew
Join Date: Apr 2010
Location: India
Posts: 19
Rep Power: 16
Sam-CFD is on a distinguished road
Hi,

I was finally able to solve the problem using the second formulation but with a finer mesh. <This approach has finally been deduced to be wrong. Please read further down for the way I solved it.>

I have another question with regard to the p_rgh formulation in OpenFOAM.

If I want to specify some initial velocity in the fluid (e.g., due to rigid body motion of the tank), do I only need to specify it as the internal field in the U file or also the p_rgh file?

Regards,

Sam

Last edited by Sam-CFD; April 23, 2012 at 01:38.
Sam-CFD is offline   Reply With Quote

Old   April 19, 2012, 15:07
Default
  #6
New Member
 
Nikhil
Join Date: Sep 2011
Posts: 11
Rep Power: 15
Nikhilcfd is on a distinguished road
Hello Sam,

p_rgh is not dynamic pressure. It is defined as p_rgh = p - rho*gh (look up pEqn.H in interFoam). It is a variable constructed for numerical advantage.

You have to specify the initial velocity in the U file only.

Regards,
Nikhil
Nikhilcfd is offline   Reply With Quote

Old   April 23, 2012, 01:36
Smile Re: p_rgh and initial velocity specification
  #7
New Member
 
Sam-CFD's Avatar
 
Sam Mathew
Join Date: Apr 2010
Location: India
Posts: 19
Rep Power: 16
Sam-CFD is on a distinguished road
Thanks Nikhil. I figured it out and have realized a more appropriate strategy for defining such problems.

The right approach is actually to define p_rgh as zero at the outlet. This is already explained in Dr. Rusche's PhD thesis.

Actually, I observed that in VOF simulations, especially, if you have pressure B.C. at inlet and outlet it is helpful to patch one cell layer at the inlet/outlet (whichever is relevant) with the liquid phase fraction of 1 if it is entering the domain, or 0 if it is leaving the domain, respectively. I remembered that around 3 years ago, I got this strategy from one of the posts on the FLUENT forum while I was implementing a swirl-injector problem in FLUENT. There seems to be some commonality between the numerical codes as it worked now in OpenFOAM as well.

Sam

Last edited by Sam-CFD; April 23, 2012 at 01:52.
Sam-CFD is offline   Reply With Quote

Old   January 10, 2019, 07:47
Default
  #8
New Member
 
elahe mirzade
Join Date: Apr 2017
Posts: 2
Rep Power: 0
E.Mrz is on a distinguished road
Quote:
Originally Posted by Sam-CFD View Post
Thanks Nikhil. I figured it out and have realized a more appropriate strategy for defining such problems.

The right approach is actually to define p_rgh as zero at the outlet. This is already explained in Dr. Rusche's PhD thesis.

Actually, I observed that in VOF simulations, especially, if you have pressure B.C. at inlet and outlet it is helpful to patch one cell layer at the inlet/outlet (whichever is relevant) with the liquid phase fraction of 1 if it is entering the domain, or 0 if it is leaving the domain, respectively. I remembered that around 3 years ago, I got this strategy from one of the posts on the FLUENT forum while I was implementing a swirl-injector problem in FLUENT. There seems to be some commonality between the numerical codes as it worked now in OpenFOAM as well.

Sam
Hi Sam,
I am working on a tow-phase air-water flow in an elbow. In zero time the elbow is full of water, and then from inlet both water and air are injected. There is gravity in the opposite direction of inlet flow . My problem is that the gas flow does not go out from outlet. It goes next to the outlet and then divergence occurs.
I tried your advise and patched water phase fraction=0 in the cell layer near outlet,but It didn't work, and when the fluids flow toward outlet that cell layer near the outlet becomes full of water again. (I couldn't attach the geometry picture )
Please help me, Thank you
E.Mrz is offline   Reply With Quote

Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
InterFoam stops after deltaT goes to 1e14 francesco_b OpenFOAM Running, Solving & CFD 9 July 25, 2020 07:36
Segmentation fault in interFoam run through openMPI voingiappone OpenFOAM 16 November 2, 2011 07:49
OpenFOAM 2.0.x libscotch issues gfilip OpenFOAM Installation 4 July 11, 2011 12:23
Slow interFoam compared with other CFD tools? Ralph M OpenFOAM Programming & Development 1 November 17, 2010 07:46
Open Channel Flow using InterFoam type solver sxhdhi OpenFOAM Running, Solving & CFD 3 May 5, 2009 22:58


All times are GMT -4. The time now is 05:26.