Helicopter simulation
2005-10-27 19:27:31.758641+00 by Dan Lyke 4 comments
I've built the helicopter controls and started flying the FlightGear Bo105 helicopter. Gotten pretty good at it, too, my hovering's a little shakey, and I'm having trouble holding forward flight altitude at speed at better than +/- 50 feet or so, but within the limits of my hovering I can do rudder turns, and fly sideways, and I'm starting to feel like I have some control.
At least in that flight model. And the FlightGear helicopter flight model is missing some critical bits, there's no downwash calculation, so there's no ground effect (and, less importantly but still an issue, no vortex ring state (VRS) issues), there's no engine model so you're blessed with 100% rotor speed all the time, but this also means that there's no autorotation, and the helicopter is powerful enough that I've hovered straight up 2,000 feet, which means that a whole set of issues with climb rate and such don't exist.
So I downloaded the X-Plane demo (6 minutes, then all the joystick inputs go to 0, with amusing results...), fired it up with their model of the Bell 206, tried to reset the controls and...
The problem here is that I've never been at the controls of a real helicopter. It's been many years, but I've felt the buffet of a stall in several different light fixed wing aircraft, I've felt rotation and the wheels getting light during take-off, I know what an approach and flare should feel like, I've held an altitude and direction with rudder turns and with the ailerons. And, frankly, most of the airplane simulators I've ever played with have been, to the limits of their model, relatively true to that fixed wing experience.
But every helicopter simulation I see starts with some basic similarities, and then is radically different from all of the other ones. I've done a lot of reading, I've been a passenger in a real helicopter once, but when the thing leaps off the ground with a very minimal collective input, and the cyclic lag doesn't seem to mesh with the numbers on the helicopters I've read, I start to wonder what's going on.
Everyone raves about X-Plane's fidelity to aerodynamics, and I can accept that it's my technique or something about how my controls are trimmed out or whatever, but parts of the experience are so completely different from those of FlightGear that I've realized I need to find an authority.
One of the reasons I stopped playing video games was that at some point in the last few levels of Tony Hawk's Pro Skater 2 I realized that I'd gotten all of the general purpose dexterity with a controller I was ever going to get, and the only thing that continued play gained me was getting better at Tony Hawk's Pro Skater 2. And I was one hidden area away from having seen everything in the game.
There is a lot more that I can learn about helicopter flight from any simulator that's applicable to the real world, but at some point many of the more difficult reflexive skills, like maintaining a hover, or being able to make the axis of a rudder turn around the mast, or the tail, or the nose, feel like they start to become more an exercise in learning a simulator than in learning flight.
And what I haven't found is anywhere that a real world pilot compares the helicopter flight model of, say, Microsoft Flight Simulator to that of X-Plane and FlightGear. I've seen several pilots talk about tuning the "reality settings" of MSFS to be fairly realistic (and I think most of those are using a third party model, like the Dodosim RealStart 206), I've seen comments which indicate that the X-Plane aerodynamics simulation is more true-to-life...
So, anyone got a suggestion about how to proceed? I don't, right now, have the budget to go drop a couple of grand on real world flight training so that I can then evaluate each of these simulators. The Hovercontrol.com folks have a community built around Microsoft Flight Simulator, which I've wanted to stay away from for reasons of future platform (I'd hoped to not maintain a Windows XP box for very much longer), but there are real pilots over there who seem happy with that. I could add basic ground effect calculations to FlightGear fairly simply, although I'm not up to more complex rotor downwash aerodynamics yet.
And, frankly, I'm seeing enough code during the day that if I'm going to be building stuff to chase this as a hobby I'd rather spend that time working on something physical, like further improving my flight controls.