[freekeer]

Hi,

I am working on a combustor, and for the moment I simply try to model the airflow inside of it.
I designed 1/6th of it to use its periodicity.
I have generated several grids with various geometries of the devices, and for some reason, I always get a "Turbulent viscosity ratio has been limited" error on some of them when I switch from a 1st order converged solution, to a 2nd order (Presto for the Pressure, 2nd order for the other variables).

The problem always comes from the inlet of my device (I have specified a mass-flow inlet).
Here is a picture of the turbulent viscosity ratio of the inlet, for a converged 1st order :



and after a few iterations with the 2nd order :



I don't know how to get rid of these viscosity bubbles.
I tried to set some yplus modification, but I am not sure I am doing it correctly, and it doesn't have any effect.

Here is the detail of my simulation :
Green-gauss cell based implicit steady solver
k-e standard model for the viscosity, with non-equilibrium wall function.
I set the air as an ideal gas.
The BOCO of the outlet is a negative pressure-outlet.

What should I do to have a correct viscosity near my wall ? (I tried to adapt the model with a normal distance from the boundary, to refine the grid near the wall, but the error keeps appearing a few iterations after).

Thanks!

[Piotrek]

Dear jean,

I suggest you keep on iterating. The problem should dissipate. I expect your "bubbles" are caused by a sudden temperature change due to combustion. You can check if in the locations of high viscosity you have combustion taking place.

[freekeer]

Thanks for your fast reply.
The problem is that at the moment, I am only modelling the airflow, with no fuel added and no combustion.
I tried to keep on iterating, but these area usually spread and make the simulation crash :/

[Piotrek]

in that case go to
solution - monitor - limits
(the path might not be exact)
and change the viscosity limit.