Southwest Airlines, Boeing’s biggest customer for the troubled 737 MAX, said this week that the airplane maker’s documentation incorrectly claimed that its aircraft had operable angle-of-attack disagree warning lights. But Boeing informed Southwest that the AoA function was actually inoperative only after the Lion Air crash in Indonesia last October.

Even as Boeing finishes software revisions on its 737 MAX airplanes and Boeing CEO Dennis Muilenburg insisted the original design met certification standards and was safe, Southwest’s statement suggests the airline didn’t get what it paid for. In a statement, Southwest said: “Upon delivery (prior to the Lion Air event), the AOA Disagree Lights were depicted to us by Boeing as operable on all MAX aircraft, regardless of the selection of optional AOA Indicators on the Primary Flight Display (PFD). The manual documentation presented by Boeing at Southwest’s MAX entry into service indicated the AOA Disagree Light functioned on the aircraft, similar to the Lights on our NG series. After the Lion Air event, Boeing notified us that the AOA Disagree Lights were inoperable without the optional AOA Indicators on the MAX aircraft. At that time, Southwest installed the AOA Indicators on the PFD, resulting in the activation of the AOA Disagree lights - both items now serve as an additional crosscheck on all MAX aircraft.”

In its own statement, Boeing said the company “included the disagree alert as a standard feature on the MAX, although this alert has not been considered a safety feature on airplanes and is not necessary for the safe operation of the airplane. Boeing did not intentionally or otherwise deactivate the disagree alert on its MAX airplanes.”

But Boeing added that “the disagree alert was not operable on all airplanes because the feature was not activated as intended.The disagree alert was tied or linked into the angle of attack indicator, which is an optional feature on the MAX. Unless an airline opted for the angle of attack indicator, the disagree alert was not operable.”

The disagree lights are of interest because the investigations into two 737 MAX hull losses—Lion Air in Indonesia and Ethiopian Airlines near Addis Ababa last month—focus on the MAX’s MCAS stability augmentation system. MCAS—for Maneuvering Characteristics Augmentation System—is active when the aircraft is hand flown at high angles of attack with flaps and slats up. It rolls in nose-down stabilizer trim to lower the aircraft angle of attack.

MCAS derives data from a single AoA sensor and faulty sensors are implicated in both crashes. If the aircraft had had operable AoA disagree lights, the pilots might have known MCAS activation was being caused by faulty data. In fact, according to the summarized CVR data from the Ethiopian crash, the pilots did mention faulty AoA indications.

Shortly after the Lion Air crash, Southwest added a software update that displays the actual AoA on the cockpit primary flight displays. But, as the airline’s statement suggests, Boeing documentation led Southwest to believe the AoA disagree warning would function whether the PFDs had the angle values displayed or not.

The revised MCAS software, according to an Aviation Week report, uses data from both AoA sensors and includes filters to detect anomalies that would indicate one or both sensors are unreliable. The flight computer would then inhibit MCAS.

Although media outlets—including AVweb—have described MCAS as a stall-protection feature, Boeing now insists that it never was. MCAS was designed to address stability issues at the very corner of the flight envelope with slats and flaps retracted and at light weights with full aft center of gravity. MCAS is inhibited when the airplane is flown on autopilot.

Boeing is finishing testing on the revised software, but no firm date has been set for returning more than 350 grounded MAX aircraft to service.

Addressing shareholders in Chicago Monday, Boeing CEO Dennis Muilenburg spoke to reporters about the 737 MAX’s troubled MCAS (Maneuvering Characteristics Augmentation System). “The MCAS system as originally designed met our design and safety analysis criteria,” he said.

When asked if the MCAS was the cause of both Lion Air and Ethiopian Airlines accidents, Muilenburg sidestepped a direct answer, saying, “There were multiple contributing factors, there are factors we can control. And in this case that common link in the MCAS and its activation. We’re going to break that link and prevent accidents like this from happening again.” The accidents were the result of a “chain of events,” he said, “there is no singular item, it’s a chain of events.”

"It's not correct to attribute [the accidents] to any single item," he said. "We know there are some improvements we can make to the MCAS and we will make those improvements. But the reason this industry is safe is that we never stop on making safety improvements. We never claim we have reached the end point. We are continuously, across all of our airplane programs, improving safety every day. We always look for opportunities to improve."

As part of his update to the press, Muilenburg said that nearly all of the 50 airlines that fly the 737 MAX have “experienced the software updates themselves” in simulators. Total test flying so far is up to 246 hours over 146 flights, according to Boeing. As reported earlier, Boeing hopes to have the airliner back in the air in July.

While admitting that Boeing has been speaking to the airlines economically impacted by the grounding of the MAX, he said, "The first focus here is safely getting the Max up and flying. And then we'll address the follow-on issues.”

Muilenburg himself has faced challenges in the wake of the two fatal accidents and subsequent grounding. Shareholders had attempted an ouster of the CEO, but less than half of the investors voted for his removal. Boeing’s income was down to $2.8 billion from $3.1 billion in the first quarter of 2018 largely on the downturn of deliveries of 737 aircraft.

Our concept of drones is that they land—sometimes with grace, sometimes not—on a more or less flat surface when their mission is over. But researchers at Yale are working on a drone that can land and “perch” like a bat to save energy during portions of its mission.

Where some studies have tested drones that can partially land on a stable surface, allowing some of the electric motors to be switched off, this latest study equips drones with scissor-like landing gear that can grasp a pole or even a tree branch to allow the aircraft to hang like a bat. In this configuration, the propulsion motors can be turned off completely. With the drone presumably parked above the ground, it’s closer to returning on station with a little more battery capacity in the tanks.

Designed with a variety of landing-gear configurations, the “bat drones” would be set up for the expected mission terrain. In fact, the idea of giving drones a rest during missions to conserve energy isn’t new. As battery technology slowly advances along with the inevitable increases in motor and controller efficiency, there’s still a sizable gap between performance and endurance. This study hopes to increase drones’ endurance so that any given battery size can carry more payload, or a given capacity can fly farther and longer.

Airbus has been flying the ACJ (Airbus Corporate Jets) 319neo for only a couple of days, but has already set an endurance record. Thanks to additional fuel tanks compared to the airline version of the A319, the ACJ-neo was able to make a 16-hour, 10-minute test flight.

As previously reported, the 319neo completed a short first flight last week. A combination of more economical engines (the aircraft can be fitted with either CFM Leap-1A or Pratt & Whitney PW1121G1 engines) and a lot more fuel capacity (from five additional center tanks) over the airline-spec A319 gives the aircraft up to 6800-NM range, something it’s capable of with eight well-heeled passengers aboard. The total tankage amounts to 37,400 liters, which still sounds like a lot when converted to 9880 U.S. gallons.

Airbus claims to have 200 jets in corporate service.

Let’s start by dispensing with the obvious: “Loss of control in flight” is a lousy explanation, and not much better as a description. Eventually we’ll come up with something better, which hopefully will reflect the myriad ways pilots can let aircraft get away from them. Spatial disorientation in IMC is as different from a moose stall as wake turbulence is from sloppily flown S-turns on final. At best, the ICAO’s accident taxonomy—adopted by the FAA and NTSB, presumably in the name of “harmonization”—provides snapshots of how accident sequences end with negligible insight into what triggered them or how they developed. As a safety strategy, “Don’t lose control” is about as useful as “Don’t let the engine quit.”

One way to make something useful from a misleadingly generic label lumping together disparate elements is to separate them back out into more meaningful components. The hazards that lead to LOC-I can be usefully partitioned along two dimensions: those resulting from pilot inputs vs. external factors, and the altitude available to recover.

It’s Not You, It’s Me

Pilot-induced upsets begin with stalls. Regular readers already know that accidents resulting from unintended stalls are almost always initiated below 1000 feet AGL. This shouldn’t come as a surprise; angle of attack routinely edges nearer its critical value while getting on or off the ground, and altitude equals options for recovery. That said, stalls while climbing to cruising altitude are not unknown, particularly in airplanes controlled by autopilots set to maintain a constant rate of climb. The newest avionics suites are a little smarter, but older setups (both analog and digital) happily keep pulling the nose up until full break. Using the autopilot’s constant-airspeed climb feature is more prudent, even if that means telling ATC you can no longer manage 500 feet per minute.

How much altitude constitutes “altitude” depends on both airplane and pilot. A sharp 172 jockey ought to be able to recover from a coordinated straight-ahead stall in as little as 200 feet, but that won’t work in an amateur-built Lancair. Once initiated, few light airplanes recover from a spin in less than a thousand feet. Many require more, which is one reason 80 percent of stalls during the turn from base to final are fatal. If your technique is sharp and your reaction immediate, you might survive a spin entered at 2000 feet AGL—but both your laundry and the upholstery will need some attention afterward.

Accelerated stalls during steep turns and sudden pull-ups are survivable given enough room, but typically require more of it than the wings-level, low-airspeed variety. The pull-up in particular generates a sudden break often accompanied by a wing drop, chewing up more of the precious space between spinner and terra firma. Then, of course, there’s the wingtip strike or runway excursion during a crosswind landing—a subject that’s inspired more than one article of its own.

Boom Times

If the pilot didn’t unsettle the aircraft, the culprit must be the air outside. Turbulence severe enough to upend an airplane can arise from at least four distinct sources: thunderstorms, larger aircraft, mechanical interference from topography or buildings, and the unpredictable phantom of clear-air turbulence (CAT).

CAT is primarily a high-altitude phenomenon, unexpected until it’s reported via PIREP. Fortunately, it’s relatively rare. That combination makes it both important and possible for pilots who operate in the flight levels to learn how to get the situation back in hand.

Mechanical turbulence is usually worst down low, although mountain waves can propagate into the flight levels. Anticipation is a good starting point: If strong winds are perpendicular to fixed obstacles, whether mountains or hangar rows, disruption can be expected. If avoiding the area or delaying the flight aren’t options, at least try to make the transit at the highest possible altitude to soften the blows and buy time in case of the worst. Cross ridgelines at 45 degrees rather than straight-on to preserve the option of turning away before it gets too bad. And it’s not just rotors, which you may never have seen outside textbooks. Any unusual-looking clouds on a mountain’s lee side are reasons to avoid the area. At pattern altitude and below, swirling winds strong enough to challenge your crosswind technique are reason to go around, then go somewhere else.

Wake-turbulence encounters are concentrated in the same places aircraft are—around airports—but altitude doesn’t grant immunity. In 2013, a Piper Arrow crossing Lake Michigan was broken up in flight by the wingtip vortices of a McDonnell-Douglas MD-80 on approach to Milwaukee. The Arrow was a mile and a half behind the airliner, crossing perpendicular to its track 39 seconds after the jet had passed. And larger airplanes aren’t the only turbulence hazard. On a single day—March 8, 2014—three light airplanes in different parts of the country were upset by rotor wash from heavy military helicopters operating at civilian airfields. The rolling moments imparted by the ’copters’ outwash were enough to overcome the airplanes’ control authority at full deflection. Recovery was impossible at pattern altitude, making awareness and avoidance the only practical strategy.

Thunderstorms are also best avoided, but some combination of misunderstanding, ignorance and bravado still results in a handful of penetrations every year. Once things start getting rough, the best approach (beyond prayer, if so inclined) is to minimize loads on the airframe by throttling back to well below maneuvering speed and trying to keep the wings level without too much worry about altitude or heading. Turns increase the load factor, so try to minimize time spent in the cell by motoring straight through—especially if the airplane and the storm are moving in opposite directions. Most accidents resulting from thunderstorm encounters are in-flight breakups, so minimizing mechanical stress and getting out as quickly as possible are the top two priorities.

Well, That Didn’t Work

If anticipation and avoidance worked perfectly, there’d be no need to practice emergency procedures. But things happen, and when they do there’s no substitute for preparation. If you’d had to make your first crosswind landing flying solo, relying only on whatever you’d picked up from a textbook, how well would it have gone? Likewise, if you can’t count on never finding yourself pitched 30 degrees nose-up while rolling through inverted, you’ll want it to happen the first time in an aerobatic trainer under expert supervision.

Systematic upset prevention and recovery training (UPRT) serves three basic purposes. The prevention part sensitizes a pilot to early indications of unfavorable trends, improving the chance of correcting excursions before they become extreme. The combination of classroom and in-flight analysis of potential upset scenarios develops the ability to analyze the nature of the excursion (nose high or low? upright or inverted?), determine the correct response and execute those steps in real time. Finally— and perhaps most importantly to the light GA community—experiencing the dizzying speed and altitude loss of any departure from controlled flight imparts a visceral sense of how low is too low to take risks, reinforcing your primary CFI’s admonition to limit bank angles and be mindful of coordination while close to the ground. It also drives home the degree to which the startle response and accompanying adrenaline rush can impede the ability to think.

What’s ‘Unusual?’

While the terms are often confused, UPRT is not the “unusual attitude” recovery familiar from instrument training and proficiency checks. While the latter includes any significant divergence from the intended flight path, the FAA has defined an “upset” as any unintended pitch excursions beyond 25 degrees nose-up or 10 degrees nose-down, or unintended rolls of more than 45 degrees bank. It also includes less extreme attitudes “at airspeeds inappropriate for the conditions.” Unfortunately, the agency’s principal guidance—Advisory Circular 120-111, “Upset Prevention and Recovery Training”—is targeted almost entirely at scheduled airline operations, with little relevance to the aircraft or resources available to most GA pilots.

For practical purposes, we’d equate “upset” with “obscenity”: You’ll know it when you see it. A typical training profile begins gently, with some normal maneuvering—climbs, descents, turns—to let the student get a feeling for the aircraft. (This alone can be worth the price of admission to someone who’s never flown anything more responsive than a Cherokee.) Next is likely to be a stall sequence, beginning straight-ahead and power-off, and progressing through deep stalls—stick full back, trying to keep the nose straight with rudder in a “falling leaf” descent—and on to accelerated stalls during steep turns. An introduction to inverted flight might come next, followed by spin entry and recovery. By that time, the inner ears of anyone without prior aerobatic experience are likely to have had enough.

Follow-up flights introduce typical excursion profiles: nose-high and overbanked (past 90 degrees) to simulate a wake-turbulence upset during climb, and nose-low and skidding as in the base-to-final turn. A common philosophy guides recovery from both (see “Yeah, But I Came Here for the Acronyms” below). More than one session is typically needed for the situations to become familiar and recoveries effectively prompt—accelerating the student’s perceptions enough to break high-speed sequences down in real time and fine-tune their responses.

You won’t come out of it invulnerable to upsets— but you’ll benefit from both some experience recovering when possible and a sharpened understanding of when it’s not.

Don’t Do It Yourself

It goes without saying that upset recovery is something you don’t want to learn by trial and error. The inevitable inexpert recovery attempts can easily exceed your Skylane’s load limits, so aerobatic certification is a necessity.

There’s no shortage of commercial providers offering upset prevention and recovery training comprising a combination of classroom and cockpit time. Many are geared primarily toward airline, corporate and Part 135 operators flying turboprops or jets. Some are simulator-based, which is a tradeoff. Sims still can’t reproduce the kinesthetic sensations of, say, inverted spins, but they’re cheaper and safer than the airplane for many things. Training flights in former military trainers or high-end aerobatic piston models, however, translate more readily to typical owner-flown aircraft.

The chief drawbacks are time and cost: two or three days and several thousand dollars in tuition for a full course, plus travel, meals and lodging. But it does save the hassle of trying to set up a less-expensive, informal alternative locally. Patty Wagstaff—pictured above, who not coincidentally offers a one-day, two-flight Upset Prevention and Recovery Training course—is on record as saying that upset recovery “is just aerobatics;” to one well-versed in that discipline, no attitudes are really unusual. Working with a sympathetic aerobatic instructor, you can devise a curriculum that suits your needs and schedule while controlling the rate of cash burn.

Yeah, But I Came Here For The Acronyms

There’s no knowing how often the acronym PARE—Power, Ailerons, Rudder, Elevator—has helped guide recoveries from inadvertent spins, but it’s better than nothing. (Better still is to have practiced actual spin recoveries.) A more general upset-recovery checklist doesn’t lend itself to neat acronym mnemonics. The best we’ve been able to come up with is USTOP’R:

• UNLOAD the airframe. Even inverted, it’s usually a gentle push on the stick, yoke or column to reduce excessive G-forces.
• STOP any rolling or yawing motion with appropriate control inputs.
• THROTTLE as appropriate—add power if you’re nosehigh, reduce it if nose-low.
• ORIENT the aircraft’s lift vector to avoid negative Gs during the rest of the recovery, typically by leveling the wings.
• PULL the nose back up to level flight.
• Once everything’s back under control, RECONFIGURE for normal flight. This includes power setting, trim and any necessary reset of gear, flaps and/or spoilers.

Without hands-on practice, the chances remembering and executing these steps in an actual upset are...slim. Saying them on the ground and performing them in the cockpit are two very different things. But memorizing them is at least a start.

I Can See Clearly Now

Remember the first time you went under the hood as a student pilot? Or maybe your instructor had you try to fly with your eyes closed. If those experiences didn’t convince you of the impossibility of staying shiny-side-up without either visual references or practice interpreting the flight instruments, you’re probably hopeless (and possibly dead). Even so, attempts to skirt around, slide under, or climb or descend through inconvenient clouds continue to kill several dozen pilots and passengers each year. Once the nose gets pointed downward, airspeed builds quickly. If the ceilings are low, there probably won’t be time to recover after falling out the bottom. Even with adequate room, there’s a risk of pulling the machine apart if control inputs amplified by panic overload the airframe.

David Jack Kenny has been a statistician twice as long as he’s been a pilot—and rarely gets upset. He’s a fixed-wing ATP with commercial privileges for helicopters.

This article originally appeared in the November 2018 issue of Aviation Safety magazine.

For more great content like this, subscribe to Aviation Safety!

AVweb’s General Aviation Accident Bulletin is taken from the pages of our sister publication, Aviation Safety magazine. All the reports listed here are preliminary and include only initial factual findings about crashes. You can learn more about the final probable cause on the NTSB’s website at www.ntsb.gov. Final reports appear about a year after the accident, although some take longer. Find out more about Aviation Safety at www.aviationsafetymagazine.com.

January 25, 2019, Mt. Hood, Ore.

Rockwell Commander 112

The airplane collided with terrain at 1559 Pacific time while maneuvering around the peak of Mount Hood. The solo private pilot was fatally injured and the airplane was destroyed. Visual conditions prevailed.

Radar data indicate a target approaching Mount Hood at 1521 from the north at about 10,000 feet MSL. The target then flew a counterclockwise, six-mile-wide orbit around the 11,239-foot peak. The target got closer to the peak as the orbit continued, until it reached its highest altitude of 11,900 ft about mile north of the summit. The target continued to track around the peak until it reached the southern side, when it rapidly descended. The last recorded radar target was at 9600 feet, about 400 feet northwest of the crash site.

Nearby airports reported relatively light surface winds while area upper air soundings indicated wind speeds reached about 45 knots out of the north at elevations between 10,000 and 15,000 feet.

January 26, 2019, Lexington, KY

Beechcraft S35 Bonanza

At about 1540 Eastern time, the airplane was substantially damaged during a forced landing while on final approach. The airline transport pilot and his passenger were not injured. Visual conditions prevailed.

According to the pilot, the airplane was “high” on the approach and he “slipped” it until descending to his desired approach angle. At about 1000 feet AGL, the pilot added power to arrest the descent but the engine did not respond. Remedial actions were unsuccessful at restoring power, and the airplane lacked the altitude necessary to glide to the runway. During the off-field landing, the airplane struck several fences which divided the property and substantially damaged the left wing, fuselage and empennage. The airplane came to rest 1.4 miles from the approach end of the runway. Thirty-seven gallons of fuel were removed from the fuel tanks during recovery operations.

January 27, 2019, Fort Worth, Texas

Beechcraft A36 Bonanza

The airplane lost engine power during a practice instrument approach and was force-landed in a field at 1634 Central time. The airline transport pilot sustained minor injuries; the passenger was seriously injured. The airplane sustained substantial damage to the forward portion of the fuselage. Visual conditions were reported at the airport about the time of the accident.

Examination revealed clear, bright fuel free of contaminants was aboard the airplane. The fuel gauges indicated slightly more than � in the left fuel tank; the right tank was empty. The fuel selector was positioned on the left tank. According to the pilot, both tanks were � full at departure. When the engine lost power, he switched “to the other tank” and attempted to restart the engine, but to no avail.

January 28, 2019, Oceanside, Calif.

Piper PA-28-151 Warrior

At about 2052 Pacific time, the airplane collided with a hillside during initial climb. The commercial pilot was seriously injured, and a pilot-rated passenger was fatally injured. The airplane was substantially damaged. Instrument conditions prevailed; no flight plan had been filed.

A witness saw the airplane take off but lost sight of it when it flew behind a tree line. However, he then heard a loud impact that he likened to a car crash. He was not able to see the hillside, which was covered in a low fog layer. He reported the crash, but authorities were unable to locate the site until the next morning at about 0715. The airplane had come to rest just below the ridgeline of a 210-foot hill.

January 28, 2019, Prospect, Ore.

Wittman 8-W Tailwind Experimental

The airplane was substantially damaged at about 1750 Pacific time when it impacted terrain following a partial loss of engine power and subsequent forced landing. The solo pilot was seriously injured. Visual conditions prevailed.

The pilot had purchased the airplane the previous week, had assembled it the day prior to the accident, and had worked on the engine just prior to the flight. Shortly after taking off to the south and at an altitude of about 200 feet, the engine experienced a partial loss of power. The pilot, unable to maintain altitude, made a forced landing about two miles south of the departure airport. During the landing sequence, the airplane collided with a stand of trees.

January 29, 2019, Grand Prairie, Texas

Cessna 172S Skyhawk SP x 2

At about 1330 Central time, the two Cessnas collided in midair about six nm from their base airport. Both airplanes sustained substantial damage to one wing; one airplane also had a damaged fuselage. The flight instructor and student pilot aboard each airplane were uninjured. Visual conditions prevailed.

The flight instructor aboard one Cessna recalled the other airplane in his left peripheral vision immediately before the collision. He did not have time to react. The other flight instructor also reported he did not have time to avoid the collision, estimating impact occurred within one second of observing the conflict. Both airplanes landed without further damage.

This article originally appeared in the April 2019 issue of Aviation Safety magazine.

For more great content like this, subscribe to Aviation Safety!