[ufo20]

Hello Fluent users,
I am analyzing the hydrodynamic flow in a wastewater treatment reactor with Re = 450. The laminar model was chosen for the simulation. A polyhedra-based mesh with approximately 1.3M cells was generated in fluent meshing. The mesh has an average and maximum skewness of 0.12 and 0.5, respectively.
I use the coupled pressure-based solver and second-order upwind discretization schemes. To initialize the simulation, I use hybrid initialization. However, with this setup, I have not been able to get a converged solution, the velocity residuals do not drop below 1e-3, and the continuity residual rises above 0.
I am attaching images of the reactor, the baffles, the water inlet, outlet and the residuals.
I am looking forward for someone's advice.

Cheers Juan
reactor.jpg

Residuals.png

[ashkan.ghafari92]

Hi,
There are some tips you can try :
1- use first-order upwind discretization at the beginning of the simulation.
2- reduce the courant number or urf in the solution control section.
3- use transient solver.

[LoGaL]

Hi juan,

The mesh is good, so no problem about that. I would only put some more elements in the pipe inlet and extend it a fair bit (6-8 Diameter lengths). You need to have developed flow entering the reactor, place the inlet far away from where interesting stuff happens. This may also help with convergence

Apart from that, let me understand better: is there something spinning inside the reactor? Like a sort of propeller? If yes, no wonder the steady simulation doesn't converge