"Skyray 1" Personal Single Jet
Flight simulators generally simulate the flight model of an aircraft by reference to look-up tables, which consist of numeric values the author of the aircraft had to insert. So the degree of realism achieved depends on how complex these look-up tables are, and how correct their figures are. Needless to say, that only existing aircraft with known flight models can be simulated this way, and only known and well examined flight conditions. All but one flight simulators I have ever heard of work this way, weither it is a common PC version for private entertainment or a multi Millionen Dollar installation for professional pilot training.
This one exception is x-plane®. By using the aircraft's geometry, as it is exposed to a virtual atmosphere, to compute the flight model, the author cannot go wrong by writing incorrect numbers. Although there are a few values which have to be calculated or estimated, the main workload is done by the sim, taking into account all the forces caused by wind, turbulences, prop- and jetwash, gravity etc. This computation is done many times per second, giving the pilot a most realistic feeling of how those forces are working on his plane. It handles even abnormal flight situations in a manner which are far too complex to be covered by a look-up table, or even unexplored. So when I built the C-5B Galaxy, the Aero Commander or an Airbus, I never had a need to search for roll rates and and similar parameters to get a flight model right, I just had to get the geometry right.
Certainly I do not totally trust this system. I always am varifying my results by comparing them to known performance data of the real thing and by letting a real pilot test my model (which is not a very precise method). Whenever I detected a discrepancy, I always had to find that something in the geometry was not right, and x-plane® worked correctly by showing me the consequences of this error. As far as the flight model is concerned, an aircraft, correctly modelled for this simulator, will give the virtual pilot a more precise and realistic impression of how the real aircraft flies than any other existing system.
More than that, x-plane® is quite helpful in evaluating basic concepts for new aircraft models. It's unique principle let's you test the flight characteristics of aircraft, which do not exist in reality. There is no way in doing so by using look-up tables. Several real aircraft companies are using x-plane® for this purpose, reportedly even Boeing, but only few of them officially state this, like Stavatti, Carter Copter and some others. - On the other side, it may be disastrous for a small aircraft company not to use it. I tested a rough XP model of the Eclipse 500 in x-plane®, about one year before the real aircraft's first test flight. I could easily find out, that the aircraft had too much drag for two JJ22 engines with 770pds of thrust each. It could take-off and fly, but only with almost full power of both engines, and it did never reach the predicted speed and altitude. That's exactly what the real test pilot found out 12 monts later, but Eclipse Aviation put it into other words: The engine provided not sufficient power for their aircraft. - Well, this certainly was correct, but I understood their goal was to build an aircraft just for this particular engine, which provided it's thrust as promised... Well, they had to redesign their aircraft in order to fit for another type of engine, which has 900pds of thrust, but is much heavier, a lot more expensive and less fuel efficient. Their business model seemed to crash and a vast number of subscribers cancelled their orders and reclaimed their downpayment. Fortunately the Exclipse 500 adventure turned out to be successful meanwhile. And now Eclipse develops the Concept Jet, a single jet with the basic configuration of my Skyray 1, which I built in x-plane® in early 2002. I wonder if they are using x-Plane® this time...?