[ANSYS Meshing] Bubble modelling to simulate shape oscillations

akash · January 30, 2013, 13:16

Hello,

I'm trying to simulate a single bubble shape oscillation and I needed help in modelling it.

I created the grid using Gambit. The problem I'm facing is that the bubble interface is not being exported.

I have a made a square domain. At the center of the domain I've created a edge who's shape is that of the bubble. I meshed the whole square domain and then tried to split the mesh along the bubble interface. I'm getting some errors saying that the split is unsuccessful.

When I tried by creating two faces but meshed only the square domain, the bubble interface doesn't get exported because the other face is not meshed. I do not want to define two separate faces which are non-intersecting and mesh them, as that will not result in a structured mesh.

What I need is that the whole domain must be meshed as shown in the attached figure http://imgur.com/0lFY5jD. Then split the mesh along the bubble boundary so that it allows me to specify two different zones.

I'm fine if the grid is generated in any other software (Hypermesh or Pointwise or ICEM). At the end of the day, I just need Fluent to identify the two zones explicitly alongwith the bubble interface. Also note that the initial shape of the bubble is complex (not a conic).

Since the problem is 2D I'm attaching the .dwg and .tin files
http://www.sharecad.org/CADView.aspx...3-7d0c1b40c205

http://sdrv.ms/WfwFB0

Please Help!

-mAx- · January 31, 2013, 02:25

Quote:

Originally Posted by akash

What I need is that the whole domain must be meshed as shown in the attached figure http://imgur.com/0lFY5jD. Then split the mesh along the bubble boundary so that it allows me to specify two different zones.

For the 2d-picture, in Gambit before you mesh both domains, you need to split the square with circle. Is gives you 2 connected domains: the disk and the square with an hollow.
Now if you want to work with interfaces, you need to disconnect both surfaces.
The quickest way (my point of view) is to copy the disk anywhere, then delete the disk at its original place. Now define both interfaces, and move the disk (copy) at its original place.
Both surfaces are disconnected, and you can mesh

akash · January 31, 2013, 10:07

The problem with splitting the square domain along the curve is, I'll not be able to get a structured grid. I needed a structured grid (all the cells must be rectangles).

-mAx- · January 31, 2013, 10:38

everything is possible

Here is on of many ways to get full hexa mesh
It is not nice, but I did it very quick
Sans titre1.png Sans titre.jpg

akash · January 31, 2013, 13:16

That's a good idea! But I also needed to grid within the circular region, to simulate air.

-mAx- · February 1, 2013, 02:32

Sans titre.jpg

January 30, 2013, 13:16	Bubble modelling to simulate shape oscillations	#1
akash New Member Akash Join Date: Jan 2013 Location: India Posts: 20 Rep Power: 13	Hello, I'm trying to simulate a single bubble shape oscillation and I needed help in modelling it. I created the grid using Gambit. The problem I'm facing is that the bubble interface is not being exported. I have a made a square domain. At the center of the domain I've created a edge who's shape is that of the bubble. I meshed the whole square domain and then tried to split the mesh along the bubble interface. I'm getting some errors saying that the split is unsuccessful. When I tried by creating two faces but meshed only the square domain, the bubble interface doesn't get exported because the other face is not meshed. I do not want to define two separate faces which are non-intersecting and mesh them, as that will not result in a structured mesh. What I need is that the whole domain must be meshed as shown in the attached figure http://imgur.com/0lFY5jD. Then split the mesh along the bubble boundary so that it allows me to specify two different zones. I'm fine if the grid is generated in any other software (Hypermesh or Pointwise or ICEM). At the end of the day, I just need Fluent to identify the two zones explicitly alongwith the bubble interface. Also note that the initial shape of the bubble is complex (not a conic). Since the problem is 2D I'm attaching the .dwg and .tin files http://www.sharecad.org/CADView.aspx...3-7d0c1b40c205 http://sdrv.ms/WfwFB0 Please Help!

January 31, 2013, 10:38		#4
-mAx- Super Moderator Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 3,297 Rep Power: 41	everything is possible Here is on of many ways to get full hexa mesh It is not nice, but I did it very quick Sans titre1.png Sans titre.jpg __________________ In memory of my friend Hervé: CFD engineer & freerider

February 1, 2013, 02:32		#6
-mAx- Super Moderator Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 3,297 Rep Power: 41	Sans titre.jpg akash likes this. __________________ In memory of my friend Hervé: CFD engineer & freerider

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Bubble Column Modelling	dlaw	FLUENT	24	October 31, 2017 07:01
two phase modelling in bubble column	srikkanth	FLUENT	1	March 27, 2017 10:21
Modelling a bubble flow ina microchannel	Aritra Sur	FLUENT	0	April 25, 2008 12:19
modelling slug flow with pressure inlet	Azman	FLUENT	0	May 29, 2006 07:45
Modelling of bubble merging submerge on water	Carlos	Main CFD Forum	0	October 26, 2005 10:30

January 31, 2013, 10:07		#3
akash New Member Akash Join Date: Jan 2013 Location: India Posts: 20 Rep Power: 13	The problem with splitting the square domain along the curve is, I'll not be able to get a structured grid. I needed a structured grid (all the cells must be rectangles).

January 31, 2013, 13:16		#5
akash New Member Akash Join Date: Jan 2013 Location: India Posts: 20 Rep Power: 13	That's a good idea! But I also needed to grid within the circular region, to simulate air.