Continuing to build upon the late-1990s legacy of knowledge, intensive efforts in the past four years have propelled airplane upset prevention and recovery training (UPRT) from the milieu of a few subject matter experts to finalizing international standards and guidance for commercial air transport (ASW, 7/13).
What’s new is increased experience among airlines that — after working closely with other stakeholders to stem the risk of loss of control–in flight (LOC–I) — have become voluntary early adopters of UPRT, some receiving glowing responses from pilots who have completed this training.
Presenters and attendees filled in details of these developments, cited a few points of controversy and highlighted next steps on their agendas during the World Aviation Training Conference and Tradeshow (WATS 2013) in April in Orlando, Florida, U.S.
“Stall training [has] been required, for a private pilot license through type ratings, forever,” said Paul Kolisch, a captain at Pinnacle Airlines. “Nonetheless, we continue to lose airplanes in the commercial fleet, and most [such accidents] are following stall events and the loss of control–in flight. [Pilots] didn’t need to get upside down or even close to it. But the airplane wasn’t flying. And the pilots didn’t recognize it. … We have to use our imaginations and be open to new possibilities that really, virtually, violate our traditional stall training.”
Despite consensus recommendations of industry specialists behind the U.S. Federal Aviation Administration’s (FAA’s) August 2012 publication of Advisory Circular 120-109, “Stall and Stick Pusher Training,” Kolisch said he still encounters skepticism and distrust about UPRT-related changes. “There are still people out there who say ‘They don’t apply to us,’” he told the conference. “Well, wings apply to you. And if they stop flying, they’re a problem for you.”
A significant discrepancy has endured, he said, between the 5,000-ft to 10,000-ft altitudes traditionally used in approach-to-stall training for airline pilots and the altitudes where actual stalls occurred in recent LOC–I accidents. Moreover, the traditional training had failed to emphasize that typical pilots instinctively react to a startle/surprise affecting their flight path with immediate control input to increase pitch. UPRT instructors address this response, telling simulator students, “That’s what you’re going to do; now here’s how you recover,” Kolisch said.
“In current practices, approach to stall is a scripted maneuver; it’s limited to non-realistic scenarios and it’s typically hand-flown,” he said. “I call it choreography.” Updates to regulatory standards, official guidance and practical test standards overcome such weaknesses. Therefore, stall prevention, recognition and recovery can be accomplished today before an anomaly deteriorates to a violent, possibly unrecoverable, airplane upset, he said.
“The first thing you do is get the nose down,” Kolisch said, paraphrasing the key message adopted by various stall and UPRT working groups. “If you don’t get the nose down, the wings aren’t flying long enough to level them. … [Training requirements also] should not mandate a predetermined value for altitude loss, nor mandate attaining an altitude during recovery.”
Pinnacle Airlines is an example of the airlines that have opted to implement the latest best practices. “No one comes out of our training without going through high-altitude and low-altitude stall training — most of it starting on the autopilot [to be] realistic,” Kolisch said. One point of potential confusion that must be overcome in low-altitude stall training, he said, involves flight crews hearing “PULL UP” alerts from the terrain awareness and warning system at the same time that stall recovery requires them to reduce angle-of-attack to reattach airflow to the wings so that responding to the alerts becomes possible. While showing a video of an airline crew’s simulator session in this scenario, he said, “The GPWS [ground-proximity warning system] was telling them ‘PULL UP.’ Pulling up kills you.”
Stick Pusher Training
For operators of stick pusher–equipped airplanes, UPRT elements should be implemented to avoid negative training, Kolisch said. Without this training, the typical response of a pilot who encounters the shaker or other indicators of stick pusher activation (firing) is to pull the control column.
Simulator instructors tasked with inducing surprises can tell you, “Don’t pull, don’t pull, don’t pull — and you’ll pull,” he said. “But if you practice it a few times, then you’ll release it. … Release, put the nose down, and you’ll recover.”
Lou Németh, chief safety officer and a captain, CAE, cited a potential source of confusion. “You’re sitting in the cockpit of a stick pusher–equipped airplane, and the preflight procedures require you to hang on to the stick pusher and fight through the stick pressure,” he said. “So every preflight, you’re sitting there holding onto this thing, and you’re doing absolutely the opposite of what you should do if the stick pusher fires in flight. … We want to demonstrate the stick pusher, and we want to see the pilot demonstrate proficiency in respecting the stick pusher when it fires in flight.” Németh was chairman of the Stick Pusher and Adverse Weather Aviation Rulemaking Committee and Loss of Control Avoidance and Recovery Training, a committee of global civil aviation authorities; and co-chairman of the International Committee for Aviation Training in Extended Envelopes Training Committee.
In February 2009, the Colgan Air Flight 3407 crash in the United States (ASW, 3/10, and 5/12) began to raise LOC–I to the highest tier of the safety agenda of the Air Line Pilots Association, International (ALPA), eclipsing airline pilot selection, licensing and mentoring, recalled Frank Cheeseman, human factors and training group chairman for ALPA and an Airbus A320 captain for United Airlines.
“We need to give pilots the tools to survive: low-altitude, medium-altitude and, in most of our operations, high-altitude [UPRT],” he said. “It’s a pass-fail exercise. It’s a train-to-proficiency exercise. It’s a survival exercise.”
The airline industry absolutely must “avoid negative training, but at the same time, we shouldn’t be afraid to use our simulators because they’re not exactly perfect,” Cheeseman said. ALPA recently has joined follow-on activities, including a search for ways to enhance the “difficult skill of pilot monitoring” and redouble its effectiveness.
ALPA urges the U.S. airline industry not to attempt to circumvent full UPRT — specifically, the training requirements included in U.S. law1 — by not exposing pilots to full aerodynamic stalls and recoveries in approved flight simulation training devices (FSTDs). “We don’t understand why this is a big discussion in the United States,” he said.
Airbus and Boeing representatives at the conference reiterated their positions that the stall avoidance and recovery aspect of UPRT is valid and important for all airline pilots, regardless of the protective automation of fly-by-wire airplanes. Such training is necessary in part to prepare flight crews for a lower level of protection — for example, changing from normal law to alternate law2 on Airbus aircraft.
“We do the stall exercises up to the point where it is valuable,” said Jacques Drappier, a captain and senior adviser training, Airbus. “We’re not at the point that we go into a post-stall situation, which was the case of the Air France [Flight 447 LOC–I, ASW, 8/12]. There you are in a totally different regime. And I don’t think anybody is ready to go into that regime at this point in time. … I think the limitation is ‘What are the capabilities of training?’ not ‘What are the airplane capabilities?’”
ALPA’s Cheeseman said, however, “There are some airlines that are engaging [in simulator] training in a flight-protected airplane in a full-stall situation. The initial critiques from [ALPA-represented] pilots that have gone through that training have been extremely positive … a tremendous confidence-builder. I happen to be one of them.”
Other Pilot Reports
As part of its advanced qualification program for pilots, UPS Air Cargo has designed programs to teach low-altitude, low-speed stalls and high-altitude, low-speed stalls in FSTDs, said attendee Jeff Ryan, a UPS captain, Boeing 767 check airman and simulator instructor.
“Having gone through it, I will tell you that it was extremely eye-opening,” he said. “First, [practicing recovery from] a low-altitude, low-speed stall, initially dropping the nose, the airplane regaining flying speed, and then doing a high-altitude, low-speed stall with the autopilot engaged, [reinforced] how much time it actually took once you dropped the nose. … The airplane would break the stall, but your tendency to recover way too early and get a secondary stall was extremely impressive [in] that it took patience to get the airplane … roughly back to about 200 kt. Anything between the 130-kt to 140-kt stall — and even [at] 170 kt — you get back to a secondary stall. So the training was fantastic.”
- The Airline Safety and FAA Extension Act of 2010 affected many aspects of airline pilot licensing and training in the United States in the wake of the Colgan Air Flight 3407 crash near Buffalo, New York.
- The Airbus system logic called normal law provides a number of automatic protections against exceeding flight envelope parameters. Manual selection or unexpected system reversion to alternate law logic requires different flight procedures.
Preview of UPRT Enhancements to Instructor Operating Stations
Imminent flight simulation training device (FSTD) enhancements for teaching pilots upset recovery in large commercial jets include several currently under review by the U.S. Federal Aviation Administration (FAA) and already described in FAA interim guidance. An example of the technology being finalized within the flight simulation industry provides upgraded feedback to instructors/evaluators via four new displays intended for instructor operating stations (IOS) and debriefing rooms,1 says Lou Németh, chief safety officer and captain, CAE.
During one session at the World Aviation Training Conference and Tradeshow (WATS 2013), he presented designers’ concepts of what this IOS feedback may look like and how it will work. So far, Németh said, one purpose for giving the airline industry previews of these displays has been to “dispel the myth … that simulators are not an effective tool for upset prevention and recovery training [UPRT].”
The four new IOS displays are tools that an instructor will use to provide feedback to the student during and after six standardized UPRT maneuver scenarios. The interim FAA guidance specifies that this feedback must indicate the fidelity of the simulation, the magnitude of student control inputs, and the aircraft operational limits that could affect the success of one or more maneuvers.
The FAA’s three minimum requirements for any FSTD used for this purpose (Figure 1) are a simulator validation envelope (the alpha/beta crossplot2 on the IOS or equivalent alternate method); IOS display of flight control inputs while the student is performing maneuvers (especially any inputs that otherwise cannot be assessed by the instructor, such as rudder pedal displacement and control forces); and a method to inform the instructor of the relationship between the maneuver and aircraft operational limits. The latter typically would be a V-n diagram3 showing how these limitations may affect the maneuver. Ideally, the FAA suggests, operators should install capability to record and play back all these dynamic parameters.
“Our graphs should give the instructor enough information to show whether or not the airplane was recovered within the airplane structural limits and the limitations as they’re defined by the Airplane Upset Recovery Training Aid [ASW, 7/13, p. 30],” Németh said. “The instructor has to look at all four of these” alongside data from flight control positions and flight instrument indications.
CAE predicts that typically these IOS upgrade tools will replay and freeze moments or frame sequences during debriefing animations of flight crew maneuvers/aircraft performance, and help simulator students understand the safe/unsafe outcomes.
“When you are doing UPRT, there’s so much going on over a short period of time that you need to be able to stop and look at it almost frame-by-frame to understand and show pilot performance and properly debrief the student,” Németh said. “Recorded data may be replayed because it’s so rich, and it happens so fast. The instructor can stop at the IOS and retrace the dots across the screen to see what happened, to see where the flight controls were in every part of the maneuver on a second-by-second scroll.”
For example, capturing the alpha-beta plot readily depicts the aerodynamic status of the airplane inside and/or outside the normal (green) aerodynamic envelope. “I maintain that a good upset recovery will stay inside that green envelope,” Németh said.
Similarly, the V-n diagram shows that as a pilot’s maneuver increases aerodynamic load in relation to the known structural limits, the stall speed increases along the diagram’s coefficient of lift/drag curve. Plotting maneuver data as green, yellow and red dots traced over the V-n diagram helps the instructor and student to visualize and grasp the complex interaction of aerodynamic parameters.
On one of the new IOS displays, the control positions and primary flight display are supplemented by data for speed-trim status, trim indicator, speed brake status, rudder pedal deflection (percentage and the amount of force applied), g-meter (a device indicating aerodynamic load relative to standard acceleration of gravity [g]), autopilot on/off status, autothrottle status, and landing gear up/down status.
The differences among safe, unsafe and unsurvivable performance quickly become apparent when these feedback data immediately are available. The CAE presentation focused on two pilots’ recovery performance during one of six standardized UPRT simulator scenarios, the nose-up maneuver.
In one slide, the UPRT-trained pilot’s “maneuver started at just slightly under 2 g,” Németh said. “The nose pitches up, the pilot unloads, moves the airplane toward the center of the envelope and then starts a dive recovery until the end of the recording. Each dot represents one second. The margin between where the pilot was and the positive-g stall is very important; this shows a good recovery. This shows that the pilot understood the relationship between g and stall, unloaded the airplane and moved the airplane away from the danger zones.”
He compared the preceding safe performance on this maneuver with that of a similar type-rated airline pilot who had not completed UPRT. “The green dots at about 1.8 g are the start of the maneuver,” Németh said, showing the corresponding slide (Figure 2). “The airplane is pitched up in the nose-high recovery maneuver and the unload is not as significant as the good recovery example. This pilot is in and out of the stall during the recovery. We can represent the difference — or the margin of safety — between where this pilot was and the safe recovery of the airplane as per the previous slide. There is no margin of safety, and he is in and out of the stick shaker in the recovery.”
Simply counting dots on the display of the alpha-beta plot tells the story. The unsafe-maneuver pilot was in a stall for 15 seconds. “I know that because I can see that he was on the stall line on the V-n diagram, and he was in this red region on the alpha-beta plot,” he said. Moreover, this pilot’s unsafe effort even exceeded the simulator validation envelope — that is, the known range of fidelity to the actual airplane — of the FSTD’s approved aerodynamic database.
- FAA National Simulator Program (NSP). “FSTD Evaluation Recommendations for Upset Recovery Training Maneuvers.” Flight Simulation Training Device Qualification Guidance 11-05.
- An alpha-beta plot graphs the wing angle-of-attack in degrees (alpha) on the vertical axis and the airplane sideslip in degrees (beta) on the horizontal axis. Flight envelope boundaries are overlaid based on flight test data, wind tunnel data and data from engineers’ extrapolations. In Figure 1 and Figure 2, the irregular green-outlined shape is the approved aerodynamic envelope of the simulator incorporating flight test data under license from the airframe manufacturer.
- A V-n diagram shows the normal load factor and airspeed limits of a specific aircraft type.