CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Coanda ejectors (

Rajesh October 9, 2005 21:46

Coanda ejectors
Hi all

I am trying to simulate the flow in a coanda ejector (a device which is used in aerospace applications..and coanda effect is the pehenomenon by which a fluid follows a curved wall without separation). In the coanda ejector, the high pressure gas at 6 bar is flowig along a curved wall due to coanda effect and it induces a secondary flow due to its motion. The total ejector is like a divergent nozzle. I have tried with both RKE and K-Omega model ...ofcourse with lot of mesh geometries. ..and I believe that the mesh quality is quite good. But all the models show almost same convergence history as, initially convergence (Oscillating always)is much faster and with the number of iterations increasing, the convergence becomes slower and even stops.Why is it happening? is this the problem of my mesh or the solution model?

Can any body help me?

Thanks in advance


Matt October 10, 2005 02:07

Re: Coanda ejectors
Hi Rajesh,

I think it's a bit of both to be honest. I would suggest you remesh your model and most importantly the solution setup may be wrong.

What turbulence model are you using and what sort of solution setup have you got, most importantly what is the mesh like?



Rajesh October 10, 2005 02:15

Re: Coanda ejectors
Dear Matt, Thank you for your immediate response to my problem. My problem is 2D axi-symmetric with structured quadrilateral cells. I have used KE (RNG) and K-Omega (tarnsitional) models for solving the problem. Both models give the same convergence history. (the total number of nodes is 31000). i.e., initially the solution is converging well and after some 20000 iterartions (in both the models) the residual history shows some oscillations and after that it is not at all converging for a long time and finally it diverges. This is what's happening. I think I haven't missed anything. If you want a more detailed discription, I will send you my geometry and convergence history. Thanx a lot


bitak September 11, 2009 07:31

I work on project about subsonic ejector like you.
My solve is :air - couple - axisymetric - k-e rng - pressure inlet and outlet ideal gas - my presure inlet are 150 ,250 ,350 ,450 Kpa and out let is atmosphere.
unfortunately my solve is divergence!!!!!!!!!
really I need to your help.
I change k-e to k-omega and pressure inlet to mass flow but devegence again.I know i have a shoke and so reverse flow in fluent.

tstorm September 11, 2009 11:49

I've done quite a bit of work with circulation control wings, which use the coanda effect to increase lift. Right away I found that k-e and k-w models are not useful models for flows like these. Both based on the Boussinesq hypothesis, which makes the Reynolds stresses isotropic.
One solution is to use the RSM model, but the coupling of Reynolds stress equations usually makes convergence difficult. My solution has been to use the nonlinear v2-f model by Pettersson-Reif using UDFs and UDSs - no easy task. You could also use a nonlinear modification of the k-e model, which I've seen but haven't used.

amira December 12, 2010 11:02

what did u use for simulating ejector
i am working on simulating an ejector..
i read many papers but i didn't find informations about which model they used, wether it was multiphase, VOF, mixuture, Euler ??? transport equations ???
and when i set a pression at the pressure outlet condition, it divergs... i need really to have informations about this

All times are GMT -4. The time now is 11:07.