CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Dynamometer, 2D Axisymmetric! (http://www.cfd-online.com/Forums/fluent/102798-dynamometer-2d-axisymmetric.html)

Dynamometer, 2D Axisymmetric!

2 Attachment(s)
Hi All;

I am working on a dynamometer (you dont need to know what it is!!!), and for simplicity I changed its geometry as in Fig.1, as you can see water is inside a cylinder which contains a rotor spinnnig with angullar velocity of 5500 rpm. I set my Computational Domain like in Fig. 2. Note that there is no any inlet and outlet in this analysis to decrease computational difficulty.

Through my case setup in Fluent I did following:

1- Pressure Based- 2d Axisymmetric swirl (I dont know why I cant use just 2D Axisymmetric ??)- Absolute Velocity Formulation

3-Viscouse: I used Laminar, k-e and spalart-...

2- Material water; I assumed constant density

3- BC:
Fluid zone as single rotating frame (its rotation axis is x-axis as stated in Fluent Docs) , angular velocity of 5500 rpm

Rotor: rotating moving wall with relative velocity of zero (However I think I could set it as stationary wall as well. right????)

wall: rotating wall with absolute velocity of zero.

4- First Order Discritization Scheme for starting iteration and then shift 2 2nd order. For pressure I selected PRETSO! though.

5- operating pressure: about 3 bars, since I know from experiment its pressure is in that range.

6- initialization: just changed pressure from zero to 2 bar.

My results: Although solution converges, result is crazy. for example the inside pressure varies for -3 bar to 100 bar!

I also checked my Y+ value. it was alright.

I would appreciate if you reply anything you may think useful regarding my problem!

Regards.

 xrs333 June 3, 2012 04:45

Nihao, Khayyamian,
I think i can explain your puzzle. Your model has no flow inlet and outlet, and the flow is imcompressible, so the pressure in the model is floating. To overcome this, define the operating pressure at a position in the domain.

I will check out what you mentioned, it takes some days though. I will notify then.

But I asked another question, what if used 2D Axisymmetric without Swirl but wall still moving?

 delaneyluke June 8, 2012 06:41

2D axisymmetric without swirl with not capture the effect of the rotating wall on the fluid and hence your results will look the same as when your wall does not rotate.

Regards
Luke

 vahid.najafi July 27, 2012 03:06

1 Attachment(s)
Im produse 2Daxisymetric.
but after run , my results is not one with 3D???and incorrect!!!!
Have i correct boundary type?

axis:empty (empty or symmetryPlane is correct?????? )
wall:wall
inlet: patch
outlet: patch
frontAndBackPlanes:empty
frontAndBackPlanes_pos:wedge
frontAndBackPlanes_neg:wedge

 flotus1 July 27, 2012 04:05

The case you have is not axisymmetric. The main velocity component is the tangential component, which is not even computed in an axisymmetric case.
The right way to model your case is with a periodic boundary condition and a 3D geometry or as axisymmetric swirl.

 vahid.najafi July 27, 2012 07:36

an other question

Hi dear Alex S.
please learn me more for this idea!!!

 flotus1 July 27, 2012 08:26

You will learn much more about it by reading the section "axisymmetric swirl" in the Fluent manual.

 vahid.najafi July 27, 2012 09:29

an other question

2 Attachment(s)
I worked with openFoam!!!why you tell me fluent???please tell me Details about it.

 flotus1 July 27, 2012 10:24

Because this is the FLUENT subsection of the forum.
I know nothing about open foam.

 vahid.najafi July 27, 2012 10:29

ohhhh

ohhhhh!!!!!Im very sorryyyyyy!!!!
thanks.