On Wednesday 03 December 2003 23:31, Andy Ross wrote: > Lee Elliott wrote: > > A couple of things about modelling sea-planes in FG though - a) unless > > you start in the air, you have to start on a runway, and b) with > > YASim, at least, you can't define the fuselage properly because part > > of it has to start below the surface and you get a collision at > > start-up. Dunno how JSBSim and UIUC handle this. > > I'm not following you. While it's true that YASim has no support for > sea plane float modelling (buoyance and extra drag), it's not clear to > me that it's doing the wrong thing on a runway. I mean, if a real > plane started with its tail in the ground... :) Seriously, though: on > hard surfaces, there ought to be something you can do with normal gear > objects to get correct behavior. Let me know if you're having any > trouble. > > And actually, seaplane float handling really wouldn't be that hard -- > the float just acts like a skid with a softer spring and slightly > different drag behavior (need to include a v^2 term, which friction > doesn't have). > > The harder part is teaching the scenery system how to tell water from > land, and writing an API to forward that information to the to FDM. > This could be rolled into the task of handling: > > Brandon Craig Rhodes wrote: > > Wow, and I suppose simulating sea-plane landings means we have to > > model ocean swells. Once we have them, though, they will provide the > > bonus that carrier decks can start pitching and yawing realistically. > > The existing ground interface only handles elevation, which means that > it doesn't understand non-level surfaces. A fancier per-gear > interface needs to be written, which would include indicators for > "normal vector", "velocity" (carriers move!), and I guess "surface > type". > > Andy > Hi Andy,
When I was trying to do a YASim fdm for the SR45 I used three fuselage definitions for the hull - the upper and lower decks, and the planing hull. This was for aerodynamic effect. It was the planing hull fuselage object that was, by definition, partly below the surface as it extends below the water line, which is the effective u/c position. Removing the planing hull fuselage object fixed the problem but at the expense of the full aerodynamic effect of the hull. I agree that a pretty good attempt at getting the u/c characteristics could be achieved with what's already in YASim, at least by some who knew what they were doing:) so that's not really a problem. Ther fuselage problem actually made some sense to me - the fdm has to be aware of the a/c contact points, to detect crashes, and must take it from stuff like the fuselage definitions. if I then define a fuselage object that extends below the u/c then I shouldn't be surprised that the fuselage contacts the ground before the u/c - crash condition. LeeE _______________________________________________ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
