Hi all,

I want your help in clearing my doubts regarding the residue when using Euler model for a unsteady gas-solid two phase flow inside a riser column CFB. One can consider it as a verticle pipe too.

Now, I set up all the relevant boundary conditions for Eulerian multiphase model using Granular solid. Before I calculate solids, i tried to get the results only for Air(primary flow)by deactivating volume fraction and gran. temp. equations. I set 10^-6 as my convergence criteria for continuety eqn.

The solution met convergence criteria for each time step and finally flow became steady with many number of time steps.

Here I activated VOLUME FRACTION AND GRANULAR TEMP. equations. I set very low, 0.0005 sec time step.

Now the problem here is,

- Continuety eqn does coverge only upto 0.l or maxm 0.01.

- I tried to iterate as many as 500 iterations for a time step, though it was not a good idea.

- I think that making time step smaller would give more convergence. But this may take very very long time that one may not wish to.

QUESTION: Can anyone suggest what could be the steps to follow to improve convergence?

Thank you very for your time and help

-US

If you're using the complete differential equation for granular temperature implemented in FLUENT 6.2, the only possibility you have is to use a low time step.

Gidaspow generally adopts a time step between 10^-5 and 10^-6. Arastoopour used 5x10^-4.

To improve convergence, the only tip I know to work is to set the URF of pressure to 0.6 and the URF of momentum to 0.4.

P.S. Use at least a second order upwind scheme for convective terms. The first order upwind is too diffusive.

Best regards,

ap