Hey Folks: I hope you don't mind me jumping in here on this but I do have an opinion on this issue.
The problem with using flight simulators to “teach” unusual attitude recoveries, is that simulators are merely computers, doing what they’ve been programmed to do. The thing that makes a B737 simulator fly like a B737, and a B747 simulator fly like a B747, and not vice versa, is the programming of the appropriate simulator computers with data obtained from the B737 or B747, respectively. The data that is available is typically just a bit more than the maneuvers performed for and during aircraft certification. AND, the big factor is that ALL of this gathered data is within, and just barely beyond, the normal operational envelope of the aircraft. Yes, there are some additional data that are acquired that are only for flight simulation purposes, but its not a lot and it certainly doesn’t go beyond the flight test validated envelope.
When someone takes a simulator outside of the flight test validated envelope, while the simulator may still roll or loop, there is absolutely no guarantee that the airplane will do anything like what the simulator does. Some serious queries have been made, on more than one occasion, to the experts in the field, regarding the ability to get the data and the cost, the reliability, the repeatability of that data. Interestingly, they ALL, everyone of them, has said the following. The ability to get the data would be limited. It would be very costly. It would likely be dangerous in more than just some cases. The accuracy of the data could be easily challenged (this is mostly because repeating the maneuver several times is likely to yield wildly different results). We discussed the possibility of extrapolating the data… Certainly this could be done, but to what accuracy? Where would the data curve go if it were to be extrapolated beyond the flight-tested results? More to the left, more to the right, would the curve wrap tighter or relax? Extrapolation would be directly dependent on the suppositions used. What suppositions would we depend on? To whom could we turn? What pilots have experience in these uncharted areas in the airplane we’re attempting to simulate? Who would have the experience to make such suppositions?
Please understand; I am NOT advocating not using simulators to address the recovery from unusual attitudes. To the contrary! What I am saying is that how we use these very valuable training, testing, and checking tools is critical and we must prevent our dependence on them beyond their capabilities. For example, showing a pilot a 120 degree bank and 10 degree nose down, and then explaining a technique to roll the airplane all the way around – to “take advantage of the rolling energy” already established – (yes, I’ve heard just such stories!) would be ill advised, at the very best. This is true if for no other reason than the fact that no one knows how the airplane would react if such a stunt were attempted. However, showing a pilot what 120 degrees of bank and 10 degrees nose down attitude looks like outside the front window in comparison with the flight instruments is not only possible – it is quite valuable. We’re not talking about how to recover, we’re looking at what information is available to the crew and how they can best use that information. The discussion could go on from there about the shortest distance to roll to get the airplane back to a “wings level” condition. You could discuss whether or not you should pull on the controls during the roll? You could discuss what kind of “g’s” are likely to be generated in doing so? How much aileron should you use in this circumstance? I don’t know. And neither do you. However, you could discuss what to do about “the ball?” Is it important to know what to do with it? Or is it irrelevant?
Simulators are really very good tools. Lets use them to the very limits of their ability and not one bit more than that.
Thanks for the opportunity to "sound off."
AirRabbit