[luca_g]

I would like to compute incompressible inviscid flows (i.e. Euler equations) using OpenFOAM.
I've tried using icoFoam with nu=0, which in principle should be right since pressure correction schemes are also valid for the inviscid flow equations.
However, the solution (with symmetry or slip bc at the solid wall, they actually are the same) with any limited scheme (e.g. gammaV 1.0) is far too dissipative and shows some problems close to the wall. For example, applied to a circular cylinder case it resemble a low-re flow, converging to a solution with a large closed recirculation bubble downstream the cylinder.

I've also checked the rhoSonicFoam and sonicFoam solver, which also address inviscid flow (the first one is only for inviscid flow).
Strangely, I've run the forward step cases included in the tutorial with both solvers and obtained two completely different results from two codes. The one using sonicFoam looks better although there is a large recirculation on top of the obstacle.
In the tutorial cases the wall bc is specified as adiabaticWall with velocity equal to zero, which doesn't look consistent with an inviscid flow.
However, a run with sonicFoam specifying slip bc at the obstacle wall does not change the solution except in the first layer of cells adiacent to the solid wall: thus, even in the case of sonicFoam as was for icoFoam the solution seems too dissipative.

Does anybody have suggestions/comments about these points ?

Thanks,

Luca