I'm trying to simulate compressible Poiseuille flow using a program I have written that implements the MacCormack algorithm and I'm having problems getting an accurate velocity profile. It's a pressure driven periodic flow, so the velocities at the inlet match the outlet and the pressure at the inlet is force*rho. It also assumes that the flow is isothermal, so I don't have to worry about temperature variations.

The reason it has to be compressible is that I'm comparing it to a molecular dynamics flow of the same type. So, I'm trying to reproduce the same conditions in both simulations. It also has to be periodic due to the same condition in the molecular dynamics simulation. Does anyone have any insight into why my MacCormack program isn't working under these conditions? If you have any questions, feel free to ask. Thanks so much.

Tell us more about your Grid, Re, Mach, time-stepping algorithm. And which kind of problem arises when you run your programm and what means "not accurate profile" ?