supersonic flow past flat plate

shuo

Hi

I am simulating it by solving the conservation form of the Navier-Stokes equations using MacCormick's method. I am solving it at Ma 4, Re 700. The velocity profile looked reasonable (compared with vanDriest) but the pressure and density profile had sharp oscillations across the boundary layer(BL). The temperatue profile was smooth. Is there any way I can minimise the oscillations(artificial viscosity?).

ALso for the initial conditions, I had to use a assume a linear BL velocity profile over an equivalent subsonic Blasius flow. If I assumed uniform velocity everywhere except at the plate, the solution blows up immediately. Is it because of the huge velocity difference at the surface nodes? For hypersonic flows Ma > 6 the solution always blow up.

Cheers Shuo