# Panel methods for use in aeroelastic model

 Register Blogs Members List Search Today's Posts Mark Forums Read

 September 24, 2009, 12:49 Panel methods for use in aeroelastic model #1 New Member   Sara Join Date: Sep 2009 Posts: 3 Rep Power: 16 I'm currently researching an aeroelastic model to help tailor composite lay-ups. As part of this we need an aerodynamic model to provide the loads on an aerofoil for given geometry, then calculate the deflection, then iterate with the new geometry. Due to the number of iterations and as these will be initial sizing calculations I think a panel method will be best, rather than a full 3D mesh. I was wondering if anyone could advise on the major differences between the vortex lattice method and the panel method. Obviously they both have the limitations due to inviscid modelling, but what are the major differences in assumptions? Am I right in thinking that VLM cannot account for thickness effects, but trhe panel method can? Are ther any advantages VLM has over the panel method? The two programs I'm looking at in more detail are PANAIR and Tornado, does anyone have any advice on relative advantages/disadvantages Any comments would be gratefully received Many thanks Sara

 September 24, 2009, 18:02 #2 Senior Member     Paolo Lampitella Join Date: Mar 2009 Location: Italy Posts: 2,163 Blog Entries: 29 Rep Power: 39 Hi Sara, On the theoretical side both the methods are developed from the same assumptions and, in some cases, they are completely equivalent (VLM are just a kind of panel method). Moreover, it is not strictly true that VL methods cannot handle thickness (they just need some attention to avoid uncorrect behaviors or unsolvability of the equations). However, on the practical side, you're right...the panel methods (as meant by you) can handle thickness and, in general, allow for more flexibility while the VLM can only handle open surfaces on which some circulation exists (that is no actual code exists in which a VLM is modified to handle thickness). The difference between the two methods relies in the singularity distribution and boundary condition used to solve the same integral equation for the velocity potential (this is much more a math aspect...see Katz & Plotkin - "Low Speed Aerodynamics" for a very good reference to start with). In my knowledge, there is no advantage in the use of a VLM as in the programming, which is far more easy with respect to a 3D panel code. Also, if you have to enter the code and modify it then a VLM should be much more easy to handle. Being more specific, i.e., with respect to PANAIR and TORNADO, the first one comes from a joint effort of NASA and BOEING and is a really complex, flexible code developed in FORTRAN 77 (several papers should be available at http://ntrs.nasa.gov/search.jsp) but there is (almost) no way to modify it without a very big effort. In contrast, Tornado is a less flexible VL code developed in MATLAB during a M.Sc. thesis work, not really different from what someone else, like you and me, could have done in his aero courses (except that Tornado works and has a lot of functionality continuously added). It goes without saying that you could learn to use it in less than a day and start coupling it with your structural solver in less than two weeks. Maybe Tornado is the best choice, at least to start with. But if you want more functionality (like unsteadiness, boundary layer coupling, compressibility) and you're not scared by the effort than definitely go on PANAIR. Tornado is freely available while PANAIR should be not (i don't know exactly).

 September 25, 2009, 04:24 #3 New Member   Sara Join Date: Sep 2009 Posts: 3 Rep Power: 16 Thank you, that's a really useful reply Definitely given me some good points to think about, I just need to decide which is more suitable for this application!

 September 25, 2009, 05:04 #4 New Member   Sara Join Date: Sep 2009 Posts: 3 Rep Power: 16 Just to clarify... even though I won't be able to change the code in PANAIR, as it's a FORTRAN program I should be able to automate the input/output process, right?