GGI: rotating object completely within a static object

Tobb · October 6, 2010, 09:01

Hello together!

I got a problem with a GGI case.

I want to simulate a paddle mixer. My first attempt was based on the mixerGgi tutorial, of course with a 3D mesh. I used an inner cylinder with the rotating parts surrounded by an outer static cylinder with the same height. This worked quite well.

Now I want the inner cylinder to be completely surrounded by the outer one. Therefore I cutted out a small part of the inner cylinder at the top as well as at the bottom and copied them to the outer cylinder. Merging the two meshes and setting up the case made no difficulties.
But after running the simulation I noticed that instead of the inner cylinder, now the outer cylinder was the rotating one. And in addition to that, the cells at the ggi interface seem to be permanently connected. So after a short time, the cells become very skew and of course my results are strange and completely worthless.

I did not change anything in the dynamicMeshDict file and I don't get the reason for that problem.

What is even more confusing to me is the fact, that it works if I just cut out and paste the upper part of the inner cylinder in the outer one. If I do that for the lower part, I get the same problems.

I would be glad if someone could give me a hint. Thank you in advance!
If it helps I can try to post some pictures later on.

Tobb

Ohbuchi · October 15, 2010, 01:09

Hi,

If you doesn't define "movingCells" in CellZones, mixerGgiFvMesh will define cell region nearest to origin as movingCells automatically.
So, you should define cellZones by hand.

Tobb · October 15, 2010, 06:29

Thank you very much!
I tried it out and it works fine. I just moved down the mesh, so that the center is near the origin. (Now I also understand, why cutting out the upper part didn't work whereas cutting out the lower part did not produce any problems.The bottom of the tank is at z=0, so without the lower static part, the desired moving zone was closer to the origin than the static one.)
Sometimes it is really annoying that a beginner like me can regularly spend hours on problems that basically can be solved in a minute...

I will have a look on defining the moving cells in "cellRegion" by hand later, but at the moment, I get along with just moving the mesh in the "right" position.

October 6, 2010, 09:01	GGI: rotating object completely within a static object	#1
Tobb New Member Join Date: Aug 2010 Posts: 6 Rep Power: 15	Hello together! I got a problem with a GGI case. I want to simulate a paddle mixer. My first attempt was based on the mixerGgi tutorial, of course with a 3D mesh. I used an inner cylinder with the rotating parts surrounded by an outer static cylinder with the same height. This worked quite well. Now I want the inner cylinder to be completely surrounded by the outer one. Therefore I cutted out a small part of the inner cylinder at the top as well as at the bottom and copied them to the outer cylinder. Merging the two meshes and setting up the case made no difficulties. But after running the simulation I noticed that instead of the inner cylinder, now the outer cylinder was the rotating one. And in addition to that, the cells at the ggi interface seem to be permanently connected. So after a short time, the cells become very skew and of course my results are strange and completely worthless. I did not change anything in the dynamicMeshDict file and I don't get the reason for that problem. What is even more confusing to me is the fact, that it works if I just cut out and paste the upper part of the inner cylinder in the outer one. If I do that for the lower part, I get the same problems. I would be glad if someone could give me a hint. Thank you in advance! If it helps I can try to post some pictures later on. Tobb

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
error: uninitialized local variable 't' used	MASOUD	Fluent UDF and Scheme Programming	5	October 17, 2016 04:24
coplanar GGI: static tube and rotating tube	ernsa	OpenFOAM Running, Solving & CFD	1	October 10, 2012 07:46
defining a term for a domain using DEFINE_ADJUST	MASOUD	Fluent UDF and Scheme Programming	1	September 24, 2010 05:08
Compilation error OF1.5-dev on Suse10.3	darenyang	OpenFOAM Installation	0	April 29, 2009 04:55
[blockMesh] BlockMeshmergePatchPairs	hjasak	OpenFOAM Meshing & Mesh Conversion	11	August 15, 2008 07:36

October 15, 2010, 01:09		#2
Ohbuchi Member Masashi Ohbuchi Join Date: Oct 2009 Posts: 74 Rep Power: 16	Hi, If you doesn't define "movingCells" in CellZones, mixerGgiFvMesh will define cell region nearest to origin as movingCells automatically. So, you should define cellZones by hand.

October 15, 2010, 06:29		#3
Tobb New Member Join Date: Aug 2010 Posts: 6 Rep Power: 15	Thank you very much! I tried it out and it works fine. I just moved down the mesh, so that the center is near the origin. (Now I also understand, why cutting out the upper part didn't work whereas cutting out the lower part did not produce any problems.The bottom of the tank is at z=0, so without the lower static part, the desired moving zone was closer to the origin than the static one.) Sometimes it is really annoying that a beginner like me can regularly spend hours on problems that basically can be solved in a minute... I will have a look on defining the moving cells in "cellRegion" by hand later, but at the moment, I get along with just moving the mesh in the "right" position.