[isit12]

I am trying to calculate coefficient of performance for a vertical axis wind turbine. The turbine blade is 90 meter long. While running simulation in fluent, initial residuals are converging well. After a while, the warning appears saying 'turbulent viscosity limited to viscosity ratio of 1.000000e+05 in XXX cells'. I checked the areas where the turbulent viscosity ratios are very high. Initially it is around the interface area where I am using sliding mesh interface. Then the number of cells keeps on increasing, mostly outside of inner rotating domain. I found on the internet that it could be because the turbulent eddies are not being properly resolved as a result of large mesh size or non-uniformity of mesh. I checked the non-uniformity of mesh around the interface and refined it. The problem was still there. I also tried reducing the time step, increasing or decreasing relaxation parameters, altering turbulence model but couldn't get out of the problem. I am getting the results from the simulation, but I am sure the results are non-physical. What could be the issue causing high turbulent viscosity ratio. Is it because of two interfaces having different distribution of mesh or the shape of interface?

If mesh is still the problem how could I make it less problematic? Right now, mesh skewness, orthogonal quality are good. The Y+ in the blade is below 2. The boundary layers are well resolved. But don't know about the wake because of the uncertainty of flow around the vertical axis wind turbine . I have to limit my cell count below 2.5 million because of computation limitations. What can I do to solve the issue?