Hello! I am working on my thesis regarding VOF modeling of flow in open channels. The issue is that by specifying transient state conditions for the multi-phase flow, the solution is unable to converge even for large number of iterations. What should be the optimum value of time step size and number. Also, what should be the convergence criteria for this case. Regards
Little detail given about your application and what you like to accomplish. So difficult to give any advice.
Are you attempting to run your case in steady-state or is it already a transient simulation. If the latter, switch to pseudo-transient method with coupled solver, if not already done so. You can run Fluent with automatic timestep setting, switching verbosity to 1. In this case Fluent tells you, what automatic timestep it has determined and is using. Try to systematically reducing the timestep from this automatic timestep selection by subsequent division by factor of 2 and see, what influence this has on convergence.
In addition you might want to calculate the Courant number based on the maximum occurring velocity in your domain and the governing cell size in this location. If this Courant number is rather high for the used timestep size (high means something between 10 and 100), you might want to reduce the timestep to get Courant number closer to unity. Or you might want to make your mesh more isotropic in this location.
Regards,
Th. Frank.
Without seeing details as Thomas Frank said hard to give you solid answer
- If your channel Froude number is low your convergence is easier to obtain (even in Fr
I think the divergence may be happening because the number of COURRANT is greater than 1, or even the coefficients of sub-relaxation you can use 0.5
- Badges
- Science topic