rhoSimpleFoam blows up with second order schemes
 

Hi Foamers,

I attach the log filoe of my case. Without any apparent reason, rhoSimpleFoam blows up after a few iterations (residuals was going good before this).

I am using second order schemes (Gauss linear corrected for laplacians, Gauss linear for gradients and Gauss SFCD for convective terms). I attach also the checkMesh log.

I am simulating pressure driven flow through a complicated channel with two inlets and an outlet. I obtain good convergence with Gauss upwind and Gauss linearUpwind in div(phi,U) and div (phi,e), but I would like to switch to higher order schemes...Reynolds number is about 600.

Does anybody know what's happening? I suppose that if a second order scheme doesn't converge I can see this in residual plot, but I don't understand why it explodes!

Thanks in advance!