Influence of the velocity distribution at the inlet boundary on the CFD prediction of local velocity and pressure fields around a hydrofoil

Abstract

A study of the influence of the inlet boundary condition on simulation results is presented. Experimental measurements and simulations of the flow over a hydrofoil in a closed circuit cavitation tunnel were performed. Using the LDA method the velocity profile, 365 mm upstream of the hydrofoil, was measured. PIV method combined with the LIF technique was used for the determination of the velocity field in the vicinity of the hydrofoil. Static pressure at 38 positions on the surface of the hydrofoil was measured. The CFD code Fluent was used for the simulations. A RANS approach with the k– ε RNG turbulence model was applied. Two simulations of the fluid flow were performed – one with a uniform (ideal) velocity profile at the inlet of the computational domain and the other with the experimentally determined velocity profile. The simulation with the ideal boundary condition produced rather poor results with a considerable discrepancy to the experimental measurements. On the other hand the simulation produced more accurate results when the real (measured) boundary condition was used. The influence of the boundary condition could be seen in x- and y-velocity component in the vicinity of the hydrofoil, in c p distribution on the surface of the hydrofoil and in the evolution of the velocity profiles downstream of the inlet. It was concluded that it is essential to either use the real (measured) velocity profiles for the inlet boundary condition or to assure that the real upstream flow is almost ideally uniform, if an accurate simulation result is required.