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

Initial Condition for alpha (VOF) using funkySetFields/swak4Foam

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

Like Tree1Likes
  • 1 Post By gschaider

Reply
 
LinkBack Thread Tools Display Modes
Old   July 29, 2013, 07:04
Default Initial Condition for alpha (VOF) using funkySetFields/swak4Foam
  #1
New Member
 
Stefan Hildenbrand
Join Date: Aug 2009
Location: Stuttgart, Germany
Posts: 20
Rep Power: 7
shildenbrand is on a distinguished road
Hi all,

I have following problem:
I want to simulate how a liquid runs out of a (car) door sill. The door sill is kind of box-shaped, but unfortunately only kind of: The left and the right side have different heights, the left side and right side are not perpendicular to the bottom, with each side being differently. Unfortunately I am not allowed to post picture here.

The door sill itself is located in a larger box which is my computational area.

I now want to impose a initial condition for alphawater such that the door sill is filled to a certain percentage with my liquid.

Basically I need to know which cell is inside the door sill and which cell is outside.

I generate my mesh using snappyHexMesh and reading in STL data.

With funkySetFields I can make a logical expression in the form of "(pos().x > 1.487) && (pos().x < 3.052)" [...] but that is not sufficient.

Any ideas anyone?

Thanks,
Stefan
shildenbrand is offline   Reply With Quote

Old   July 29, 2013, 13:14
Default
  #2
Assistant Moderator
 
Bernhard Gschaider
Join Date: Mar 2009
Posts: 3,803
Rep Power: 38
gschaider will become famous soon enoughgschaider will become famous soon enough
Quote:
Originally Posted by shildenbrand View Post
Hi all,

I have following problem:
I want to simulate how a liquid runs out of a (car) door sill. The door sill is kind of box-shaped, but unfortunately only kind of: The left and the right side have different heights, the left side and right side are not perpendicular to the bottom, with each side being differently. Unfortunately I am not allowed to post picture here.

The door sill itself is located in a larger box which is my computational area.

I now want to impose a initial condition for alphawater such that the door sill is filled to a certain percentage with my liquid.

Basically I need to know which cell is inside the door sill and which cell is outside.

I generate my mesh using snappyHexMesh and reading in STL data.

With funkySetFields I can make a logical expression in the form of "(pos().x > 1.487) && (pos().x < 3.052)" [...] but that is not sufficient.

Any ideas anyone?

Thanks,
Stefan
Sorry. Without even a sketch it's hard to understand what you're saying.

And in case of "I am not allowed to post of picture here" consider getting a support contract with someone who can help you. The topic (car) and the town (Stuttgart) suggest that you're not working for someone who can't afford it
__________________
Note: I don't use "Friend"-feature on this forum out of principle. Ah. And by the way: I'm not on Facebook either. So don't be offended if I don't accept your invitation/friend request
gschaider is offline   Reply With Quote

Old   July 29, 2013, 13:37
Default
  #3
New Member
 
Stefan Hildenbrand
Join Date: Aug 2009
Location: Stuttgart, Germany
Posts: 20
Rep Power: 7
shildenbrand is on a distinguished road
Quote:
Originally Posted by gschaider View Post
Sorry. Without even a sketch it's hard to understand what you're saying.

And in case of "I am not allowed to post of picture here" consider getting a support contract with someone who can help you. The topic (car) and the town (Stuttgart) suggest that you're not working for someone who can't afford it
Ok you're right. Without a sketch - hard to understand. So I tried this sketch (cut through the configuration):IMG_0310.jpg

And by the way - no- I am not working for the company you are assuming, neither directly nor doing project work for them.
shildenbrand is offline   Reply With Quote

Old   July 30, 2013, 04:19
Default
  #4
Senior Member
 
akidess's Avatar
 
Anton Kidess
Join Date: May 2009
Location: Delft, Netherlands
Posts: 912
Rep Power: 16
akidess will become famous soon enough
Much better. Is the outer basin essential, or could you just drain into "infinity" instead?
__________________
*On twitter @akidTwit
*Spend as much time formulating your questions as you expect people to spend on their answer.
*Check out the scientific computing exchange http://scicomp.stackexchange.com
akidess is offline   Reply With Quote

Old   July 30, 2013, 04:33
Default
  #5
New Member
 
Stefan Hildenbrand
Join Date: Aug 2009
Location: Stuttgart, Germany
Posts: 20
Rep Power: 7
shildenbrand is on a distinguished road
Hi Anton,


for the first period of the process (which is the draining part) it is not necessary, for the second period (when the last drops come out) I need at least the bottom side.

Stefan
shildenbrand is offline   Reply With Quote

Old   July 30, 2013, 04:40
Default
  #6
Senior Member
 
akidess's Avatar
 
Anton Kidess
Join Date: May 2009
Location: Delft, Netherlands
Posts: 912
Rep Power: 16
akidess will become famous soon enough
An idea might be to just have the door sill alone and initialize it, and then use mapfields to a domain containing the outer box.
__________________
*On twitter @akidTwit
*Spend as much time formulating your questions as you expect people to spend on their answer.
*Check out the scientific computing exchange http://scicomp.stackexchange.com
akidess is offline   Reply With Quote

Old   July 30, 2013, 04:59
Default
  #7
New Member
 
Stefan Hildenbrand
Join Date: Aug 2009
Location: Stuttgart, Germany
Posts: 20
Rep Power: 7
shildenbrand is on a distinguished road
Quote:
Originally Posted by akidess View Post
An idea might be to just have the door sill alone and initialize it, and then use mapfields to a domain containing the outer box.
I thought about that, too, but the problem is that the door sill has many holes and gaps. If I tried to mesh it separately, I'd have to close all holes.
I think about it again - thanks for your idea!
shildenbrand is offline   Reply With Quote

Old   July 30, 2013, 06:56
Default
  #8
Assistant Moderator
 
Bernhard Gschaider
Join Date: Mar 2009
Posts: 3,803
Rep Power: 38
gschaider will become famous soon enoughgschaider will become famous soon enough
Quote:
Originally Posted by shildenbrand View Post
I thought about that, too, but the problem is that the door sill has many holes and gaps. If I tried to mesh it separately, I'd have to close all holes.
I think about it again - thanks for your idea!
Is there a way to mark the "inner" region in the mesher and then write it out as a cellZone/cellSet? Because then setting the liquid only inside would be as simple as "(zone(inner) && pos().z<maxZ && pos().z>minZ) ? 1 : 0"
__________________
Note: I don't use "Friend"-feature on this forum out of principle. Ah. And by the way: I'm not on Facebook either. So don't be offended if I don't accept your invitation/friend request
gschaider is offline   Reply With Quote

Old   July 30, 2013, 10:09
Default
  #9
New Member
 
Stefan Hildenbrand
Join Date: Aug 2009
Location: Stuttgart, Germany
Posts: 20
Rep Power: 7
shildenbrand is on a distinguished road
I have tried the following:

Wrap the door sill only with large meshsize (-> holes get closed), use convex hull (-> really big holes get closed). This gave me a coarse closed volume, which I could use in snappyHexMesh.

Next steps:
- funkySetFields (as before, only looking at the z-axis)
- mapFields to my "old" fine case

When looking now at my initial solution at t=0 I am quiet happy with the result.
However, my computation now diverges quickly (which did not happen before - my previous initialization were two approximated boxes with different height).

@Bernhard: I have found no way to generate a cellset of the inside of the door sill - so far I tried only opensource-software to prepare the surfaces and SHM to generate a volume mesh.


Thanks for you answers, I have now a better initial condition - although a different new problem
shildenbrand is offline   Reply With Quote

Old   July 30, 2013, 13:01
Default
  #10
Assistant Moderator
 
Bernhard Gschaider
Join Date: Mar 2009
Posts: 3,803
Rep Power: 38
gschaider will become famous soon enoughgschaider will become famous soon enough
Quote:
Originally Posted by shildenbrand View Post
@Bernhard: I have found no way to generate a cellset of the inside of the door sill - so far I tried only opensource-software to prepare the surfaces and SHM to generate a volume mesh.
SHM has the ability to generate cellZones. See the snappyHeater (or so) tutorial
__________________
Note: I don't use "Friend"-feature on this forum out of principle. Ah. And by the way: I'm not on Facebook either. So don't be offended if I don't accept your invitation/friend request
gschaider is offline   Reply With Quote

Old   August 2, 2013, 05:08
Default
  #11
New Member
 
Stefan Hildenbrand
Join Date: Aug 2009
Location: Stuttgart, Germany
Posts: 20
Rep Power: 7
shildenbrand is on a distinguished road
Hi all,

just a quick update:

I've played around with cellSets (separately or within SHM) but none of this seems to work in my case. The (as of today) best way to generate an initial condition for alphawater seems to be:
- generate a closed version of the door sill
- funkySetFields with pos().z < <some value>
- go to the "big" case
- mapFields <from closed version>
- funkySetBoundaryField with expression "internalField(alphawater)"

--> the liquid looks pretty good afterwards, unfortunately I get convergence problems due to the mapping (I assume)

Stefan

P.S. Is there any way to include the boundary condition "constantAlphaContactAngle" within swak4Foam?
shildenbrand is offline   Reply With Quote

Old   August 2, 2013, 06:24
Default
  #12
Assistant Moderator
 
Bernhard Gschaider
Join Date: Mar 2009
Posts: 3,803
Rep Power: 38
gschaider will become famous soon enoughgschaider will become famous soon enough
Quote:
Originally Posted by shildenbrand View Post
P.S. Is there any way to include the boundary condition "constantAlphaContactAngle" within swak4Foam?
What do you mean with "include"?
__________________
Note: I don't use "Friend"-feature on this forum out of principle. Ah. And by the way: I'm not on Facebook either. So don't be offended if I don't accept your invitation/friend request
gschaider is offline   Reply With Quote

Old   August 2, 2013, 07:38
Default
  #13
New Member
 
Stefan Hildenbrand
Join Date: Aug 2009
Location: Stuttgart, Germany
Posts: 20
Rep Power: 7
shildenbrand is on a distinguished road
Quote:
Originally Posted by gschaider View Post
What do you mean with "include"?
If I use funkySetFields on 0/alphawater, where I use the BC (for example)

Schweller_Schweller_links
{
type constantAlphaContactAngle;
theta0 50;
limit gradient;
value uniform 0;
}

funkySetFields complains about the used type, gives me a overview about the correct ones (e.g. zeroGradient and some 50 others) and stops. I have to change the type to zeroGradient, use funkySetFields, and switch back to constantAlphaContactAngle.

By "include" I mean: How and where do I include $FOAM_SRC/transportModels/twoPhaseProperties/lnInclude such that funkySetFields does not complain anymore?
shildenbrand is offline   Reply With Quote

Old   August 3, 2013, 05:34
Default
  #14
Assistant Moderator
 
Bernhard Gschaider
Join Date: Mar 2009
Posts: 3,803
Rep Power: 38
gschaider will become famous soon enoughgschaider will become famous soon enough
Quote:
Originally Posted by shildenbrand View Post
By "include" I mean: How and where do I include $FOAM_SRC/transportModels/twoPhaseProperties/lnInclude such that funkySetFields does not complain anymore?
Thanks for the clarification. But you see the point? It was clear to you because you were confronted with the problem before you wrote that posting. But everyone else is lacking that context. You wouldn't be happy with the answer "That is OK. The boundary condition is not in the runtime-selection table" instead of the one below (and there ARE people who'd confronted with that answer would say "Ah. Of course. Libs in controlDict. Only have to find the right library")

Quote:
Originally Posted by shildenbrand View Post
If I use funkySetFields on 0/alphawater, where I use the BC (for example)

Schweller_Schweller_links
{
type constantAlphaContactAngle;
theta0 50;
limit gradient;
value uniform 0;
}

funkySetFields complains about the used type, gives me a overview about the correct ones (e.g. zeroGradient and some 50 others) and stops. I have to change the type to zeroGradient, use funkySetFields, and switch back to constantAlphaContactAngle.
The "problem" is that constantAlphaContactAngle is a "special" boundary condition that is not present in the finiteVolume library. And FSF only links against this (would be hard to link it against every library that holds some boundary conditions). The good new is that you can "inject" additional BCs if they are in a library: just add that library to the libs-list in the controlDict then the runtime-selection-table will know about it and FSF can read and write this BC. The name of the library that has this boundary condition starts with "libtwoPhase" (which it is differs between OF-versions. Have a look in $FOAM_LIBBIN and experiment)
nimasam likes this.
__________________
Note: I don't use "Friend"-feature on this forum out of principle. Ah. And by the way: I'm not on Facebook either. So don't be offended if I don't accept your invitation/friend request
gschaider is offline   Reply With Quote

Old   August 5, 2013, 05:55
Default
  #15
New Member
 
Stefan Hildenbrand
Join Date: Aug 2009
Location: Stuttgart, Germany
Posts: 20
Rep Power: 7
shildenbrand is on a distinguished road
Thank you very much for your help - it actually was "libtwoPhaseProperties.so" I had to declare in my controlDict.
You're right about the description of the problem (in text or by a sketch). I'll try my very best in the future.
shildenbrand is offline   Reply With Quote

Old   August 6, 2013, 03:44
Default
  #16
New Member
 
Stefan Hildenbrand
Join Date: Aug 2009
Location: Stuttgart, Germany
Posts: 20
Rep Power: 7
shildenbrand is on a distinguished road
Hi all,

my final comment for future reference. The way I get my computation converging is slightly different as posted above. It is

- generate a closed version of the door sill
- funkySetFields with pos().z < <some value>
- go to the "big" case
- mapFields <from closed version>
- generate cellZone of cells near walls of door sill
faceSet f0 new patchToFace <Patches>
cellSet c0 new faceToCell f0 any
cellZoneSet c0Zone new setToCellZone c0
- use funkySetFields with condition "(zone(c0Zone))" to set alpha=0 there
- run case

Apparently my divergence problems came from alpha=1 at some outlet holes in the door sill.

Thanks again for the help.
Stefan
shildenbrand is offline   Reply With Quote

Reply

Tags
compressibleinterfoam, funkysetfield, setfields

Thread Tools
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 On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
How to write k and epsilon before the abnormal end xiuying OpenFOAM Running, Solving & CFD 8 August 27, 2013 15:33
pisoFoam - unstable pressure residual Industrial_CFD OpenFOAM Running, Solving & CFD 21 February 24, 2013 15:39
Extrusion with OpenFoam problem No. Iterations 0 Lord Kelvin OpenFOAM 6 April 12, 2011 11:24
Velocity blows up suddenly after 30,000+ iterations lordvon OpenFOAM 14 October 27, 2010 11:13
Computational time sunnysun OpenFOAM Running, Solving & CFD 5 March 16, 2009 04:32


All times are GMT -4. The time now is 06:15.