I am trying to write a CFD solver for a convection dominated flow. I decided to discretise the advective fluxes with an upwind approach and to use an explicit approach for the time discretisation. However, while the code works in 1d, I have problems with the second dimension. I assume the problem is that the flow is more or less oriented along the first dimension (I am simulating the flow in a duct).

Any practical suggestions on how to solve this problem? Or any open source CFD solver to suggest? (P.S. the problem is compressible ...)

Thanks in advance ...

No suggestions? Do you know any open source CFD soler where I can have a look to the solution algorithms? The books I have read so far solve the problem in 1D and are not very clear about the 2D approach ...