The only additional comment I might offer in addition to all the preceding comments (in respectful opposition to the advocacy of training fully developed stalls in a simulator), is that as the stall is approached, even with very careful, flight idle power, one knot per second deceleration, in calm air, with pitch trim stopped a relatively safe point, one very, very important aspect is the side slip. A VERY minor difference in this single parameter will more than likely to provide a rather dramatic difference in handling characteristics as the stall is approached and entered. However, given the tolerances for matching flight test data, the simulator, once programmed, will do it exactly the same way, every time – and the pilot at the controls of the simulator can be relatively wide of the mark with all the above parameters, and get essentially the same response. This is likely to cause an incorrect understanding of what to expect if similar circumstances are encountered in flight. Of course, the regulator can demand a narrowing of the tolerances allowed – but the cost of such narrowing goes up at least exponentially. And, even if this cost were justified, the tolerances were narrowed to almost zero, and the simulator was programmed to respond exactly as programmed, the only time the simulator would respond like the airplane (and it would respond more like the airplane than before) would be when ALL of the parameters of the flight test airplane were met (to within those now tightened tolerances) in the simulator. Just a bit off in any of the parameters and the simulator will do one of two things: revert back to the generic programming (not validated with flight test data) or get “stumped” and not know what to do – sort of like “division by zero.” Neither result is appropriate and the training presented to the pilot would be, at best, inaccurate. Simulation is a formidable and powerful training tool. It took the simulation industry decades to get the aviation world to recognize and trust the simulator. Well, that effort has succeeded and may have succeeded to an over zealous level. Whatever is learned in a simulator is likely to be transferred to the aircraft – even if what is learned is not correct.
Two “war stories” to drive this point a bit:
1. In the early operation of the MD-11, a flight crew was enroute from DFW to MIA. Over TPA the crew experienced an engine fire warning on the number 3 engine. All attempts to determine if the warning was real or false resulted in the belief that it was a real, honest-to-goodness engine fire. The crew, having declared an emergency and initiated a descent for landing at TPA, attempted to extinguish the fire. They attempted to discharge all the extinguishers they could, but the extinguishers “malfunctioned.” After landing and evacuation, the officials determined that there was no fire, merely a false warning, but, and this a BIG but, when they attempted to discharge the extinguishers, they worked perfectly. The reason? The flight crew (both pilots) attempted to exercise the procedures learned in their simulator. But the simulator was not checked in this area and the force required to discharge the bottles in the simulator was about one-fifth of what was actually required. The crew gave up, believing that they had done all they could. The significance of the story is that even when two different pilots believe they are in immediate danger, what they learned in a simulator takes over. Unfortunately, there are a disturbingly large number of similar situations where inappropriate simulator training led an unsuspecting crew to the brink of danger.
2. A flight crew was attempting to slow, descend, and prepare to enter holding at 15,000 feet. The autopilot was engaged but the autothrottles were not. As one would expect with modern, FMS equipped aircraft, the altitude and holding fix were achieved at the same time. The crew was engrossed in the typical “prepare for arrival” stuff. The autopilot leveled the descent and initiated a right bank to enter holding. As you would expect, without a corresponding increase in power at that time, the airspeed decayed quickly. It was a smooth holding pattern entry. However, as the aircraft was rolling toward the maximum commanded bank angle, the stick shaker activated. The captain took immediate control, grabbed the yoke, disengaged the autopilot, and “feeling” just like the stall entry he had just completed earlier in that week in simulator training, he did as his instructor had encouraged him. He advanced the throttles and used right rudder to help bring the nose down to allow more easy acceleration out of the impending stall. Unfortunately, the rudder was a bit much, and the airplane continued to bank, very quickly, and exceeded 110 degrees of right bank. Several passengers were injured and at least one flight attendant was taken to the hospital with a broken arm (or clavicle?). Why?? The simulator was an older machine and one of the characteristics was when you exceeded a preset angle of bank, the program stopped. You could take your hands and feet off the controls and it would stay at that pitch/bank attitude (“division by zero”). The instruction was to “pressure the down rudder to gradually bring the nose down to the horizon while adding power, accelerate and roll to wings level.” This is what the crew was attempting to do. But it didn’t work. Very powerful training tools, those simulators!
The application here is that if we “teach” something, virtually anything, in the now revered, simulator, when the flight crew sees, hears, or feels something similar in the airplane, they are likely to DO what they did in the simulator. My respectful opposition to Old Smokey’s position is minor, but, I believe, significant. I believe we should show pilots what UAs look like, both outside the window AND on the gauges. From this information a lot of very valuable discussion can, and should, take place. But I’d prefer to stop short of “teaching” in a simulator, the techniques to use in an airplane in such instances. I would MUCH rather use the simulator to teach the characteristics of the airplane at airspeeds down around the stick shaker. We do it in small airplanes, why not larger ones? “Slow flight.” I’d like to see pilots learn to recognize what his/her airplane feels like at these low airspeeds; see how much control deflection is necessary to achieve turns. See how much rudder displacement it takes to stay coordinated. See how much power it takes to climb at that airspeed. I’d like to see a combination of all of the above – in the simulator.
My apologies for getting long-winded here – but this is a serious issue that deserves a lot of discussion. And I would welcome that.
AirRabbit
