[jancellor]

Could anyone give me some pointers on how to write a panel method that can be monolithically coupled with a structural model I have already written?


Background: The lifting surface in the structural model is a sail-like membrane (a 2D surface in 3D space) which consists of a triangular mesh and displacements and pressures specified for each node. I can already solve this system for a given pressure field by supplying an extra n equations (or rather a residual function with n outputs that must be zero'd). I would like these extra n equations to come from a panel method instead. I have already written a vortex lattice type of panel method. But I believe the fact there is one unknown and one collocation point per panel, not per node, is a problem. Would you agree? And can I instead do something based on at least 1st order doublet distributions?



My primitive thoughts so far: Specify the doublet strength at each node. For 1st order doublet panels, the velocity is still infinite at the nodes(?) so I can't simply have collocation points at the nodes. Can I do a weighted residual integral type thing instead? Or use 2nd order doublet panels - Plotkin & Katz refer to a NASA paper by Johnson but it's too complicated for me to follow. Does this sound like the right approach? Any pointers?


Thanks