Numerical study of far-field non-local boundary conditions for transonic flow around wing
A numerical study of non-local far-field boundary conditions for transonic flow problems around the ONERA-M6 wing is done. It is shown that the conditions permit using sufficiently small computational domains providing the accuracy of the lift coefficient Cy within 0.5% (the average radius is about 2 of the wing span); this is two-three times better comparing with the commonly used characteristic-based boundary conditions. A further refinement of the model and a higher computational efficiency of the non-local boundary conditions method (like e.g. obtained in two-dimensional case) are possible after improvement of the difference scheme of integrating the Euler equations in the computational domain, first of all owing to increase the first order of accuracy to the second one.
Mathematical problems and theory of numerical methods