O calculate the aerodynamic performance of a helicopter in hovering e ight is a problem of great practical importance as well as the theoretical complexity. Theoretically, a solution of the fullNavier‐Stokes equations with appropriate turbulence modeling and body-conforming grid is sufe cient for a good description of all ofthephysicsinvolved.Butunlikethee owe eldaroundae xedwing, the trailing vortex wake of a rotary wing rotates with the rotor and is shed to a far distance below the rotor plane by its self-induced velocity. The helical vortex sheet interacts strongly with the lifting surfaces, but this process is hard to be simulated unless using a quite clustered grid, which generally requires very large computational resource. Srinivasan et al., 1 using about one million grid number to solve the thin-layer Navier ‐Stokes equations, calculated the whole e owe eld including the induced effects of the wake and the interaction of tip vortices with successive blades, but they also found their captured vortex structure was overdiffused because of the coarse grid used. The current methods for calculating rotor performance usually solve the potential, Euler, or Navier ‐Stokes equations coupled with anexternalfree-orrigid-wakemodelbasedonthe liftlineorliftsurfacetheory. 2i 6 Butitis clearthatthese governing equations arehard to match with the linear trailing wake modeling in a physically consistent manner. Further, those approaches also require fairly large computer resource from solving two coupled models simultaneously. Agarwal andDeese 7 calculatedthe aerodynamic loadsby solving the thin-layer Navier ‐Stokes equations, and the rotor-wake effects were modeled with a correction applied to the geometric angle of attack along the blades. This correction was obtained by computing the local induced downwash by the rotor wake with a free-wake analysis program. In fact, this method just established a weaker link between the rotor and its wake and avoided the complex boundary handling and the solution of coupled equations. Therefore, the grid number used is not huge, and the accuracy of calculated results is satisfactory. In the present paper we essentially borrowed the approach from Ref. 7, but extended it to the solution of a complete Navier ‐Stokes equation, rather than a thin-layer one. In addition, an improved method is suggested to obtain a proper correction of the local angle of attack of the blades. This is constructed by the comparison between the results with and without rotor wake modeling, in addition to a heuristic consideration of the coupling rotor-wake and threedimensional blade-tip effects that is expressed by a semi-empirical