Problem building 3D Panel Method

Daniel

Hi!

I'm building a matlab 3D panel method; one with constant source/doublet panels (using Dirichlet BC) other with constant doublet panels using Neumann BC). The Neumann code has an error of to high pressure coef. on places where it shold be lowest (the sphere should have CPmin -1.25, but mine is around 0.5). This means that the speed is actualy decreasing over top or lower sides of an airfoil!!! The collocation points are placed in center of weight of panel and final velocities are calculated there as well. Calculation of doublet strenghts I consider correct since their strengths make some logic. Dirichlet code plotes improper results as well. Maybe the problem could be in DeltaL expression for tangential velocities calculations in which I had no correction? (please refer to Katz, Plotkin Page357). There is another problem concerning both codes, that is sensitivity to grid! I get best results (I know it is risky to call it better when the code is flawed on actual wing when there is no tip fill. But anyhow, there are huge gradients on wing tip trailing edge.

Please let us talk about this subject for a while. I couldn't find ANYWHERE a 3D panel code so this could be my graduate work and therefore free for use.

The only reference I'm using and I could find is Katz, Plotkin: Low-Speed Aerodynamics

Thank you very much! Daniel