Please tell me how to achieve the convergence of the solution ... In Newton's method I set 50 iterations and the finest grid, but the solution still does not converge
First, you can use a low time range (for speed up to 30000 rpm). It is not needed to calculate hundreds of rotations during the simulation. Can be limited for expl. from 0 to 50 ms. Your project BCs are good, only in moving mesh some empty setting. If nothing changes in your project, we can get some results. I did not find in your project convergence problems!