continuity convergence in buoyancy driven flows
 

Hi,

I'm trying to solve a mixed convection flow. All the residuals are dropping except the continuity which is oscillating and decreasing very very slowly. I'm using an incompressible ideal gas for the density with a RNG k-epsilon model. I increased the density of the mesh around the cylinder (representing the human body) and the total number of elements is around 1,700,000. I am monitoring the velocity variations near the human body and it seems they stabilized and the variations are of the order of 10e-5. The number of iterations reached so far is 2700. Should I consider the solution as converged? Why are the continuity residuals oscillating around 1e-3?

Thank you