CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   rotating propeller in stationary baffled vessel (http://www.cfd-online.com/Forums/fluent/80894-rotating-propeller-stationary-baffled-vessel.html)

wlt_1985 October 10, 2010 12:21

rotating propeller in stationary baffled vessel
 
I would like to model a rotating propeller inside a stationary baffled vessel with water as the fluid inside. I had a lot of trials. Unfortunately, in all my model, the rotating propeller always rotating with the baffled vessel. I do not know which part I am making wrong. Does anyone how to model the correct model in this case? Please help me. Thanks.

-mAx- October 11, 2010 02:51

The easiest way would be to treat your problem with sliding mesh.
Basically you need 2 regions which are not connected: rotor and stator, and then define interfaces

wlt_1985 October 11, 2010 03:03

Quote:

Originally Posted by -mAx- (Post 278614)
The easiest way would be to treat your problem with sliding mesh.
Basically you need 2 regions which are not connected: rotor and stator, and then define interfaces

Thanks for help. I tried to model with two regions, one is the solid propeller and another is water fluid. When trying to define the interface between this two region, message has been shown as below:

Warning: materials in neighbor cell threads (3 and 4) of interior zone 14 are of different types (aluminum and water-liquid). This problem MUST be fixed before solving!

I totally do not have idea in solving it.

Please help.

Thanks

-mAx- October 11, 2010 03:09

not like that.
Substract your solid region (you should have empty area instead of solid impeller)
Then if your geometry enables it, draw (split) a circle around your rotor.
You have now 2 regions, the disk which is your rotor, and the outer domain which is your stator.
As I said you need to disconnect your 2 regions, for applying interfaces.
Check online help chapter 11.2 Sliding Mesh Theory

wlt_1985 October 11, 2010 04:23

Quote:

Originally Posted by -mAx- (Post 278620)
not like that.
Substract your solid region (you should have empty area instead of solid impeller)
Then if your geometry enables it, draw (split) a circle around your rotor.
You have now 2 regions, the disk which is your rotor, and the outer domain which is your stator.
As I said you need to disconnect your 2 regions, for applying interfaces.
Check online help chapter 11.2 Sliding Mesh Theory

Thanks for your help. If I would like to model the motion behavior of the fluid (initially which is stationary) which is affected by the motion and shape of the solid stirrer inside the stationary baffled vessel, the solution mentioned above still valid for it? Or any modification should be applied?

Really thanks to your help.

-mAx- October 11, 2010 04:30

what I mentionned above is just the geometry / mesh side (no physics).
How they move, can you explain later in fluent (not in gambit)

wlt_1985 October 11, 2010 05:24

Quote:

Originally Posted by -mAx- (Post 278632)
what I mentionned above is just the geometry / mesh side (no physics).
How they move, can you explain later in fluent (not in gambit)

Thanks. One more question, if I split my rotor (impeller) by using a geometry which has the same shape but larger size to my rotor (impeller) instead of a circle, does it work for my sliding mesh model?

-mAx- October 11, 2010 05:37

no.
The split should be a circle.
For undestanding, try to imagine a split like a star. If you rotate your star with 10, you cannot superpose rotor and stator regions.

wlt_1985 October 11, 2010 05:48

Quote:

Originally Posted by -mAx- (Post 278649)
no.
The split should be a circle.
For undestanding, try to imagine a split like a star. If you rotate your star with 10, you cannot superpose rotor and stator regions.

In order to split my rotor (impeller) by using a geometry which has the same shape but larger size to my rotor (impeller) instead of a circle, does it mean that I need to use dynamic mesh?

-mAx- October 11, 2010 05:51

if you cannot create a circle-split, then yes you will have to handle with dynamic mesh.
Can you quick post a sketch of your rotor-stator?

wlt_1985 October 11, 2010 06:11

1 Attachment(s)
Here is my initial gepmetry

wlt_1985 October 11, 2010 07:00

1 Attachment(s)
Here is my initial geometry to be studied. What should I do in order to have best model of the fluid motion behavior which induced by the solid impeller? Thanks a lot.

-mAx- October 11, 2010 07:24

your picture is quite small...
As far as I saw, you can draw a circle (cylinder-surface in 3d) surrounding your impeller.
You can also attach your geometry (*.dbs) if you want

wlt_1985 October 11, 2010 08:20

I would like to attach my dbs file. Unfortunately, it was around 1M and error during attaching. I tried to model as solution provided before and it was work. Thanks for your help. The cylinder should be just created surrounding the impeller or it should be huge enough until close to the stationary baffles?

wlt_1985 October 11, 2010 08:23

If I would like to model it with dynamic mesh, what should I do to the model setting?

-mAx- October 12, 2010 01:01

so I received your file.
Do followings:
*delete volume2
*create cylinder (height 0.2 / radius 1=radius 2=0.04 / Axis Location Positive Z)
*split volume1 with the cylinder (volume4)
*delete volume5 (solid) --> should be already deleted (volume2) but it's early this morning, don't want to investigate :o)
here we are, you have 2 regions: rotor and stator.
*copy volume1 with translation (0.1 0 0) (it generates volume 5), delete volume1
*move copied volume5 with translation (-0.1 0 0)
* The 2 volumes are now disconnected.
Apply interfaces (one for each volume)

wlt_1985 October 12, 2010 01:23

Quote:

Originally Posted by -mAx- (Post 278745)
so I received your file.
Do followings:
*delete volume2
*create cylinder (height 0.2 / radius 1=radius 2=0.04 / Axis Location Positive Z)
*split volume1 with the cylinder (volume4)
*delete volume5 (solid) --> should be already deleted (volume2) but it's early this morning, don't want to investigate :o)
here we are, you have 2 regions: rotor and stator.
*copy volume1 with translation (0.1 0 0) (it generates volume 5), delete volume1
*move copied volume5 with translation (-0.1 0 0)
* The 2 volumes are now disconnected.
Apply interfaces (one for each volume)

Thanks for help. Max. :)

cfd_confuse October 13, 2010 13:45

Hi Max, for the above mixing tank problem, may i know what is the ideal number of cells for both the moving zone and the fluid zone, in a 2.5m tall cylindrical tank?? Also, may i know how to go about setting the boundary condition in fluent?

-mAx- October 14, 2010 01:23

There is no rule. Just check that you have enough cells where you think there is pressure drops etc...

cfd_confuse October 14, 2010 08:14

Hi Max, what about the boundary condition to set in fluent? i am not sure about the boundary condition to be use for the mixingtank


All times are GMT -4. The time now is 08:03.