CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Problems with convergence with an easy system (http://www.cfd-online.com/Forums/main/65458-problems-convergence-easy-system.html)

 franzdrs June 15, 2009 18:17

Problems with convergence with an easy system

Hi, I am trying to use FLUENT to simulate a quite simple system: it is a dryer for herbs and things like that. It is simply a rectangular box as plenum chamber, 3.6m long, 1.2m wide, and 0.7m high. On top of the plenum chamber comes the product to be dried, which I set to be a porous media (I got the Porous media coefficients from deltaP vs air speed data). The length and width of the porous media is as the dryer, 3.6m x 1.2 m, and the height is 0.25m. On top of the porous media there is simply a little more wall on the sides and there is the Pressure outlet. The air inlet is an horizontal rectangular duct with cross area 0.4m x 0.3m, and 0.7m length, attached to the center of one of the 1.2m x 0.7m walls.
The air inlet is a velocity inlet with a velocity of 3.6m/s in the normal direction.
What I want to check is the air velocity distribution in the outlet. The objective is to find a dryer configuration that optimizes distribution in order to get a fairly uniform drying. And this is my most simple and basic configuration.

I am a total beginner with FLUENT and not really familiar with the whole fluid mechanics issue, but I think I have followed all the relevant suggestions given in the Users Manual to get a convergence. However I dont get convergence (according to the default covergence criteria of FLUENT). The residuals for continuity go below 1e-3 (down to 1e-4 maybe) but the residuals for the velocity vectors dont, especially the y-velocity (the y-axis being aligned with the length of the dryer) which remains at around 1e-2 and sometimes goes down to 7e-3 or so, to go up again in later iterations. By the way, the residuals oscillate significantly, especially for the y-velocity. I was expecting a constant reduction of the residuals.
I have done up to 1000 iterations, and the oscillations continue without any clear tendency downwards. Being such a simple system I thought it would be pretty easy to reach convergence and now I am quite desperate. Can anyone help me to find what is wrong?
I have started (and stayed so far) in laminar flow (although I am pretty sure the conditions are turbulent, at least in the inlet duct and the plenum), because that is one of the strategies I have read. I have also tried disabling the porous media model and not even with this I get convergence.
I used the first-order scheme for momentum, I used the Standard and PRESTO! pressure interpolation schemes (although I dont know if they influence convergence). I have let the default Under-relaxation factors.
The EquiangleSkew of my mesh was very good, I used only hexahedral cells. I think the resolution is ok too.
Although I didnt reach convergence I visualized the solution using vectors and contours and they are not what I expected. I dont know if it is because it is far from a correct solution or because what I expect is way off. In any case I think the residuals I detailed above are pretty bad, arent they? Please, if somebody could read this whole thing, give me a little help on this.

Last thing, could it be that the vectors and contours I visualized were not what I expected simply because the system is really in turbulent flow and I have so far only used the laminar model?

 All times are GMT -4. The time now is 10:27.