I am simulating buoyancy-driven flow using some turbulence model. Continuity residual scale I have set to 10^(-3). During the simulation, I have got a steady-state mass flow rate (constant), but residual of continuity is still near about 10 and 10^(-1) sometimes, but I got non-varying mass flow rate at some location. Is it correct solution?
Thanks in advance
You may often meet the convergence problem for continuity. While if the continuity can not be converged to the residual rule you set, you might compare the mass flow rate in FLUX. If the error is very small, personally I think the results make sense. Or, Please check your mesh and models.
In my opinion the problem is not converged.
Try to reduce the residual of continuity equation by improving the mesh quality (by fluent) and changing the model ( i.e. coupled -pseudo transient- High order relaxation)
Buoyancy driven applications are frequently not steady-state, but transient in nature. This might be a reason for the not reducing residuals. The consequence would be a transient simulation in Fluent until a quasi-steady state (e.g. periodically oscillating flow) is reached. Time-averaging can be applied to such a solution afterward for a characteristic and representative time frame.
Regards,
Dr. Th. Frank.
Earlier I solved this problem using horizontal heater, in that case solution is well converged in steady state problem. While in the case of the vertical heater, convergence is the problem with the exact same mesh. Thomas Frank Hang Yi Mohammad Reza Saffarian
Geometry can affect the results and convergence
Also i recommend to try using double precision. Some time this can help you for a better convergence.
- Badges
- Science topic