CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM (http://www.cfd-online.com/Forums/openfoam/)
-   -   Is Immersed Boundary Method already available in OF 2.1.x or 2.2.x (http://www.cfd-online.com/Forums/openfoam/124472-immersed-boundary-method-already-available-2-1-x-2-2-x.html)

keepfit October 5, 2013 20:46

Is Immersed Boundary Method already available in OF 2.1.x or 2.2.x
 
Hi Foamers,

I just wonder, is IB (Immersed Boundary) method already integrated in OF 2.2.x (or OF 2.1.x), or just available with 1.6-ext version?

Also I am interested in VOF + IB method, can someone recommend related journal papers on the multiphase flow (VOF or similar) with IB method?

Thanks a lot.


Best regards,

David

cfdivan October 6, 2013 09:41

Hi David,

AFAIK the IBM method is not available on "standard" OF versions.

This post could help u out - meshless

Regards,

keepfit October 6, 2013 16:24

Quote:

Originally Posted by cfdivan (Post 455288)
Hi David,

AFAIK the IBM method is not available on "standard" OF versions.

This post could help u out - meshless

Regards,

Can we use the source code of IB method, and compile within OF2.x.x. Since I dont want to install extra 1.6-ext.

Phicau October 7, 2013 04:58

Hi,

of course it is possible. It may take longer than implementing it from scratch, though, as there are thousands of changes to be made. OpenFOAM has changed a lot from version 1.6 to 2.x : some methods have disappeared and have been substituted by others, some of them are now called using other syntax, and so on.

Good luck!

Pablo

cfdivan October 11, 2013 18:06

Hi keepfit,

It will be nice if you succeed compiling an working version of IBM for 2.x. I already tried, but then I just realized that I was not able to sort out all the problems.

I am looking forward for your success story :) !

Regards,

keepfit October 24, 2013 13:01

Fictitious domain method in OF
 
Hi Foamers,

Instead of implementing IB method in OF 2.1.x/2.2.x, a so called fictitious domain (FD) method is used to present solid phase in fluid domain.

Compared to the original IB method in OF-1.6-ext that inside the immersed boundary the p/U/alpha1 field is void and the BC is fixed, the fictitious domain (FD) method treats the immersed body as fluid, plus a rigid constraint (D(U)=0) to force the "fluid" inside IB boundary behaviour as rigid body.

This new solver currently works fine in general (needs improvement), here is some rough results:

A solid ball drops from air into water

1. Initial status

http://31.media.tumblr.com/0c9d5927c...saho1_1280.png

2. ball dropping into water

http://25.media.tumblr.com/c07c01121...saho2_1280.png

3. air-water interface after ball dropped into water

http://25.media.tumblr.com/341011eb8...saho3_1280.png

4. Fictitious domain U field (ball is slowed down in water due to the drag force)

http://24.media.tumblr.com/a98ddc20d...ho4_r1_500.png


Cheers,

David

cfdivan October 24, 2013 15:29

Hi keepfit,

Pretty interesting approach! Have you tried to model external flows with this methodology...flow around a profile or sphere. It would be interesting to have a close view of the mesh at the wall/fluid interface.

Regards,

mechy October 25, 2013 02:40

Hi
I can not find IBM in OF-1.6-ext
can you give me its position in application folder ?

Regards

Phicau October 25, 2013 03:35

Hey David,

it certainly looks very good, well done! Just a question, since you are dealing with a 2-phase flow, can you confirm that the mass (VOF) is conserved with this implementation?

keepfit October 25, 2013 15:37

Quote:

Originally Posted by Phicau (Post 458905)
Hey David,

it certainly looks very good, well done! Just a question, since you are dealing with a 2-phase flow, can you confirm that the mass (VOF) is conserved with this implementation?

well, since the solid phase is treated as liquid (alpha1 is forced to 1), the alphaEqn is solved on the whole domain, the mass conservation can be satisfied in general.

However, a little "liquid" is observed leaking from within the IB boundary (only happens when solid phase in gas phase) . The reason is that the constraint is applied onto the solid phase after UEqn and pEqn is solved. It means that during the sub time-stepping (solving p and U), there is no CONSTRAINT on the solid phase. To avoid this problem the constraint might be applied in every sub time-steps, but it's computational expensive.

If someone has good advice, please feel free contact me.

keepfit October 25, 2013 15:42

Quote:

Originally Posted by mechy (Post 458900)
Hi
I can not find IBM in OF-1.6-ext
can you give me its position in application folder ?

Regards

there you are:

slides:

Source codes of IB method in OF-1.6-ext

Note that the IB method in OF-1.6-ext uses IBMask (0 - inside IB boundary, 1 - fluid domain) to calculate various fields (e.g. p. U), hence physical behaviour inside the immersed body is not solved.

wyldckat April 5, 2014 15:09

Greetings to all!

Thank you David for posting the direct links to those files!

I've finally come around to git'ifying this source code and creating a wiki page for the Immersed Boundary toolkit:
Anyone and everyone, please feel free to edit the wiki page and fork the repository! ;)

Best regards,
Bruno

keepfit April 11, 2014 07:36

good job, wyldckat!

Quote:

Originally Posted by wyldckat (Post 484045)
Greetings to all!

Thank you David for posting the direct links to those files!

I've finally come around to git'ifying this source code and creating a wiki page for the Immersed Boundary toolkit:
Anyone and everyone, please feel free to edit the wiki page and fork the repository! ;)

Best regards,
Bruno


miladrakhsha September 4, 2014 00:40

Hi all,
I just wanted to ask if there has been any update in regard to implementation of IBM
in official openFoam or at least in OpenFOAM 3.0-ext.
I am working with the one that has been implemented in OpemFOAM 1.6-ext but I am not satisfied with the solver and stability of the solver at all.
The solver terminates when bounding k, or omega warning appears. And, I realized that bounded schemes are not available in of1.6-ext.
Any comments would be greatly appreciated.

-----------
Moderator note: This topic is discussed further here: http://www.cfd-online.com/Forums/ope...m-3-1-ext.html


All times are GMT -4. The time now is 07:20.