stability threshold for simple LBM hydrodynamic system


I am working with the palabos sample vortex shedding of a cylinder in a 2D channel.

When I want to increase the Re > 10 it crashes due to density instabilities. This happens because the omega value reaches values bigger than 1.74.

If I increase the velocity to compensate the value of viscosity and therefore the value of omega, I can reach values of Re=55, by increasing the Umax from 0.02 to 0.18, which corresponds to the maximum velocity permited in the incompressible flow (more or less). In these conditions the omega has values of 1.62, not 1.74. Why is not possible to increase the Re if I haven’t reached still the value of 1.74 of omega?? The mesh is always 50y and 250x.

is there available information of the omega threshold versus Re number or anything that relates the stability with the Re with LBM?