I am attempting to validate fluent for a scramjet. Currently, I am trying to produce results for a fuel (hydrogen) and air mixing case (no reactions). I am using the coupled explicit unsteady solver (have also used implicit with similar results). My problem is that after roughly 1000 iterations, reversed flow develops. I have tried running the code longer to see if it stabilizes, but it will eventually either reverse through the entire model, or the solution will completely diverge and fluent will stop. Any suggestions?

Thanks, Scott

What are your outlet boundary conditions?

I do not have mcuh experience with the coupled explicit, but with the coupled implicit the following usually applies:

- CFL has to be increased very slowly. It usually happens that you reach reverse flow but you have to be patient. Sometimes you have to run 2000-3000 iterations with CFL=1

- Monitor qunatities to ensure that you reach a stable solution before you change CFL

- Try 1order discretization schemes first and ske if you still have problems