Abstract

The full potential model, though having known its major development through the 1970's into the 1980's, is currently used in a wide range of applications and it continues to be a valuable CFD tool proving certain advantages. The current approach, using an unstructured tetrahedral mesh and finite volume method, tries to take advantage of integrating a fast initializer (a linear potential Prandtl-Glauert corrected panel model) whose results are computationally very affordable while giving a good start for the actual full potential solver in terms of lift. As a result, the implementation of the Kutta condition will be of significantly lower computational cost, whereas it is reportedly known for damaging convergence. Also, the outer boundary of the computational domain can be set closer to the studied lifting body, as it will be aware of the lifting flow inside. Hence, a smaller computational domain can be used. Another full potential issue addressed here is the slow convergence of transonic solutions and a way of circumventing this problem has been proposed. Overall, the present full potential model implementation is intended as an accurate subsonic tool, with the capability to converge transonic solutions, while being at least one order of magnitude faster than an Euler solution on a similar mesh