[luc89ita]

Hi, I'm simulating the flow through a centrifugal compressor stage with a LES approach. I inizialized the simulation with a normal RANS (k-eps) calculation to make the residuals stabilized and so I started the large eddy simulation using a timestep size of 10e-5 sec. After some timestep, reversed flows occur on pressure outlet surface and they increase from a timestep to the next one. Plotting radial velocity on pressure outlet, I observed that they are positioned on boundary layer. My boundary conditions are pressure inlet at inlet and pressure outlet at outlet. Anyone can help me?