[Lonewolf_Mcquade]

Hi!


Im trying to simulate a supersonic hot methane jet (Mach number close to 5) in co-flow with low velocity air (0.5 m/s).



The computational domain is cylindrical with the methane gas inlet in center bottom and air flow in the rest of the bottom boundary (around the methane jet).


The sides of the domain is symmetry and top is pressure outlet.



Im using the coupled solver and ramping the CFL from a low value to 500 using the expert driver and continuity convergence accelerator along with Grid Sequencing. Tried both 2nd and 3rd order pseudo-time schemes.


The mesh consists of approx 4 million cells with finest region near the inlet.


My main problem is that the injected methane gas does not seem to propagate through the domain so in a global mass balance I only have mass inflow but no outflow out of the domain.


The turbulence model I use is k-e Realizible (RANS).


For "Coupled Inviscid flux" I use AUSM+ FVS.



One thought I have is if it could be the low velocity air coflow that could cause problem but I dont see how in a steady simulation the leaked gas mass flow is not eventually exiting the domain?


Any help would be very appreciated, thank you!


Edit: Maybe should add im modelling as non-reacting, it is the mixing im interested in.

[flotus1]

eventually exiting the domain?

Are you 100% certain that you let the simulation run long enough?
Did you check some contour plots of iso-surfaces of the methane concentration to keep track of the progress?