[Andrea159357]

Hello, i'm stuck with a problem.
I'm running a simulation of a ch4 gas in a pipeline with a t-junction.
Downstream of the junction, turbulence viscosity ratio is exeding 1e5 in a lot of cells. I tryed with a lower value of velocity and the simulation did converge but with higher value did not.
After this test i thought the problem was strong gradient downstream of t-junction so i refined the grid in that area, though the residuals did not reach 1e-6, they are stuck to 1e-4.
I even run a test with different numerical schemes. With first order upwind for all the variables the simulation converged, but with second oreder did not.
Having a group of cells with turbulence viscosity ratio up to 1e5 means that production of turbulence kinetic energy is way too large.

What am i missing? Am i doing something wrong?
How can i improve my simulation?

thanks for your help