Part one: The Fact and Fiction of Rough Air
Part One: The Fact and Fiction of Rough Air
Turbulence: spiller of coffee, jostler of luggage, filler of barf bags, rattler of nerves. But is it a crasher of planes?
Judging by the reactions of many airline passengers, one would assume so. I'd been a commercial pilot for the better part of ten years, a job that requires its share of impromptu coaching sessions with white- knucklers, and figured I had a pretty good grasp of the fearful flyer mindset. I didn't. Not until I began writing for this magazine, and fielding questions from the public, did I realize how upsetting, if you'll grant the pun, turbulence is for tens of thousands of travelers.
Intuitively this makes sense. Everybody who steps on a plane is on some level uneasy, and there's not a more poignant reminder of flying's innate precariousness, and all its potential complications, than a good walloping at 37,000 feet. It's easy to picture the airplane as a helpless dinghy caught unawares in a stormy sea. Boats are occasionally swamped, capsized, or dashed into reefs by swells, are they not? Everything about it seems dangerous.
Except that, in all but the rarest circumstances, itÕs not. And frankly that boat-airplane analogy isn't a very good one: airplanes are much less susceptible to deadly upset than boats, and turbulence itself is quite different from a roiling sea. For all intents and purposes, a plane cannot be flipped upside-down, thrown into a tailspin, or otherwise flung from the sky by even the mightiest gust or air pocket. Inherent in the design of airliners is a trait known to pilots as "positive stability." Should the aircraft be shoved from its position in space, its nature is to return there, on its own and with no drastic input from the crew. Conditions might be annoying and uncomfortable, but the plane is not going to crash. Turbulence is an aggravating nuisance for everybody on the plane, including the crew. But it's also, for lack of a better term, normal -- as naturally occurring as clouds, precipitation, or a summer--day breeze.
When a flight changes altitude in search of smoother conditions, this is by and large a comfort issue. The captain isn't worried about the wings falling off, he's trying to keep his customers as content and relaxed as possible. Pilots are sometimes approached by passengers after landing who remark about the roughness of a flight. "Man, you must have had your hands full with that one." Yet the crew will have little or no recollection of it having been bumpy at all. Not because they're jaded or cocky (much as that might also be the case) but because they understand the realities of atmospheric instability, and don't misinterpret those rocks, knocks, and jigs. Flyers tend to overestimate the effects of turbulence by orders of magnitude. "We dropped like 500 feet in two seconds!" If I've heard that once, I've heard it a thousand times. In truth, a jetliner's altitude is rarely displaced by more than about 50 feet; its bank (turn) and pitch (nose up/down) will seldom change more than a few degrees.
During a wind-whipped approach, the frightened passenger is liable to imagine the pilots in a sweaty lather: the captain barking orders as the ship lists from one side to another; hands tight on the wheel. Nothing could be farther from the truth. The crew is not wrestling with the beast so much as merely riding things out. Most of the time, pilots will sit back and allow the plane to buck and buffet rather than attempt to recover every lost foot or degree of heading. Indeed many autopilot systems have a special 'turbulence' mode. Rather than increase the number of corrective inputs, it does the opposite, desensitizing the system.
Much worse than turbulence itself could be an overzealous reaction to it -- as demonstrated with catastrophic results five years ago, after an American Airlines A300 was hit by the wake from a 747 ahead. The first officer responded with a needlessly violent deflection of the rudder, causing the plane's tail to fracture. The ensuing crash killed 265 people -- the second worst aviation disaster ever on U.S. soil. Design flaws and a possibly pre-existing stress crack may have played a role in the A300 crash, and wake turbulence, to be discussed in greater detail below, is an altogether different phenomenon than the naturally occurring kind. It's unfair to say that turbulence, in and of itself, brought down flight 587.
Just how rare is such a downing? Around the globe each day, about five million people take to the air aboard 35,000 commercial departures. Yet over past half-century, the number of airliners downed by turbulence can literally be counted on one hand, and almost always there were extenuating circumstances.
Not that you'd know it listening to the media. Last summer, after a Sibir Airlines Airbus A310 overran a runway at Irkutsk killing 124, news stories, including a widely disseminated item from the Associated Press, spoke of "turbulence" as a "potential cause." That one had me sputtering and dashing off another petulant email to the AP. It's possible that blustery weather, together with crew error, led to an unstable approach and the subsequent overrun, but the implication was that rough air itself had somehow slammed the Airbus to the ground or swept it from the runway.
So that I'm not accused of sugar-coating, I freely concede that powerful turbulence has, on numerous occasions, resulted in damage -- sometimes serious damage -- or injury. With respect to the latter, these are typically people who fell or were thrown about because they werenÕt belted in as requested. Incident archives speak of twisted ankles, broken bones, and even a fatality or two. But airplanes are engineered to take a remarkable amount of punishment, and are required to meet stress limit criteria for both positive and negative G-loads. You'll routinely see an airplane's wings flex during flight. This is to better absorb and dissipate any bumps. (I once watched a fascinating video showing the wing of a 777 in a test chamber, intentionally bent to an almost unbelievable angle before finally its skeleton failed.) The level of turbulence required to dislodge an engine or bend a spar is something even the most frequent flyer won't experience in a lifetime of traveling.
Pilots work to avoid unstable air when they can, taking their cues from weather patterns and reports from other aircraft. Some meteorological indicators are more reliable than others. Those burbling, cotton-ball cumulus clouds, particularly the towering variety that occur in conjunction with thunderstorms, are almost always a lumpy encounter. (Low-level stratus clouds, on the other hand, tend to provide smooth passage.) Flights over mountain ranges and through certain frontal boundaries will also get the cabin bells dinging. Boundaries along the jet stream are notoriously turbulent, as are various other climactic markers, many of them invisible to the eye (and to on-board radar), but decipherable from the weather data available to pilots and dispatchers before and during flight.
Unfortunately, predicting the exact location, strength, and duration of turbulence is an imperfect science at best. Partly for this reason, I'm known to provide annoying, noncommittal answers when asked how best to avoid it. "Is it better to fly at night then during the day?" Maybe. "Should I avoid routes that traverse the Rockies or the Alps?" Hard to say. "Are small planes more susceptible than larger ones?" It depends. "They're calling for rain and gusty winds tomorrow, will it be rough?" Probably, but who knows. "Where should I sit, in the front of the plane or in the back?"
Ah, now that one is easier to handle. While it doesn't make a whole lot of difference, the smoothest place to sit is over the wings, closest to the plane's centers of lift and gravity. The roughest area is usually the far aft. Imagine a see-saw. As it teeters and totters, the part that moves the least distance is the center, at the fulcrum. An airplane is the same way, except two fulcrums are at work. A plane rotates around its lateral axis (the nose-up, nose- down motion called "pitch") through the center of lift, also called the center of pressure. It rotates around its vertical axis (the side to side swing called "yaw") through the center of gravity. Both centers will shift slightly as fuel is burned away, and depending on wing sweep, speed, and angle of climb; however, they will always remain close to the areas where the wings and fuselage are joined. For the smoothest ride -- and the worst possible view -- this is your spot.
When pilots pass on reports to other crews, turbulence is graded on a scale from "light" to "extreme." The worst encounters entail a postflight inspection by maintenance staff. There are definitions for each degree, but in practice the grades are awarded subjectively by the crew. I've never been through an extreme -- not unusual for most pilots -- but I've had my share of moderates and one or two severes.
The most memorable of them took place in July, 1992, when I was captain on a twin-engine, 15-passenger turboprop. It happened during, of all flights, a 25-minute run from Boston to Portland, Maine. It had been a hot day, and by early evening a forest of tightly packed cumulus towers stretched across eastern New England. As towering cumulus goes, the formations were short -- about 8,000 feet at the tops -- and extremely nice to look at. As the sun fell, it became one of the strangest and most picturesque skyscapes I've ever seen -- buildups in every direction forming a giant garden of pink coral columns. They were also deceptively, startlingly violent -- little volcanoes spewing out invisible updrafts.
As we cruised toward Portland, a thousand or so feet above the cottony peaks, the slamming came on with a vengeance. We requested a climb, but not soon enough. When the worst of the pummeling hit, it was like being stuck in an upside-down avalanche. Even with a shoulder harness pulled snug, I remember holding up one hand to brace myself, afraid my head might hit the ceiling.
I asked Peter Murray if he's noticed any unusual trends in his data. Anecdotal evidence suggests that turbulence is becoming more prevalent -- possibly as a byproduct of global climate change. Pilots around the world report increased frequencies of, among other things, furious storms and alarmingly high winds. "Not really," says Murray. "Except that I never see the same thing twice." But turbulence is a symptom of the weather from which it spawns, and it stands to reason that as global warming intensifies certain patterns, experiences like that one I had over Maine will become more common.
There's nothing exceptional about the approach to runway 27R at Philadelphia International Airport. Least of all on a clear, calm afternoon like the one I remember in 1994. Ours was a long, lazy, straight-in course from the east. We'd come from Boston, our 19- passenger turboprop packed to the gills. Traffic was light, the radio mostly quiet. At five miles out we were cleared to land. The traffic we'd been following, a 757, had already cleared the runway and was taxiing toward the terminal. Our checklists were complete, and everything was perfectly normal.
At approximately 200 feet, only seconds from touchdown, with the approach light stanchions below and the fat white stripes of the threshold just ahead, came a quick and unusual nudge -- as if we'd struck a pothole. Then, less than a second later, came the rest of it. Almost instantaneously, our 16,000-pound aircraft was up on one wing, in a 45-degree right bank.
"Get it!" I called out, reaching for the wheel. It was the first officer's leg to fly, but suddenly there were four hands at the yoke, turning it to the left as far as it would go. Even with full opposite aileron -- something seldom used in normal commercial flying -- the ship kept rolling to the right.
A feeling of helplessness, of lack of control, is part and parcel of nervous flyer psychology -- the fear that comes from being at the mercy of two unseen strangers, who you hope are competent, qualified, and sober. It's an especially bad day when the pilots are experiencing the same uncertainty. There we were, hanging sideways in the sky just a few feet from death. Everything in our power was telling the plane to go left, and it insisted on going right.
How far did it go? Sixty degrees, or thereabouts. To get a sense of how drastic that is, normal banks are around 15 degrees, and will rarely exceed 20 degrees. Never before had I seen the ground from such a perspective, and it was positively terrifying. We even threw in differential power, instinctively bringing up the right engine to overcome the twist.
As suddenly as it started, the madness stopped. In less than five seconds, before either of us could utter so much as an expletive, the plane came to its senses and rolled level. We evened the asymettrical power and, just like that it, was over.
"Go-around," is what I said next, instructing the copilot to abort the landing, which at this point wasn't a landing at all, and get the hell out of there. He'd already commenced the maneuver on his own. We set target torque and began a climb; we retracted the gear; brought up the flaps; and around we went for another circuit, this time finishing off with a smooth-as-silk touchdown.
What happened -- and we fully knew it -- is that we'd been slammed by the preceding aircraft's wake.
Chances are you've heard the term "wake turbulence" before. If you can picture the cleaved roil of water that trails behind a boat or ship, you've got the right idea. With aircraft, however, wake effect is exacerbated by a pair of vortices that spin from the wingtips. At the wings' outermost extremities, the higher pressure air beneath is drawn toward the lower pressure air on top, resulting in a circular flow that trails behind the aircraft like a pronged pair of sideways tornadoes.
The vortices are normally invisible, but are occasionally revealed when passing through mist or cloud, as seen in this sensational image. They are most pronounced when a plane is heavy and slow -- that is, when the wing is working hardest to produce lift. Thus, prime time for an encounter is during approach or departure. As the vortices rotate -- at speeds that can top 300 feet per second -- they begin to diverge, and they sink. If you live near an airport, stake out a spot close to a runway and listen carefully as the planes pass overhead; you can often hear the vortices' whip-like percussions as they drift toward the ground.
Here's another masterful shot. Those ghost-like whirls show the vortex rotation. Get a wing stuck in that blender, and it's easy to visualize what might happen. The long white streams of condensation show the vortex cores. Those core streams (not to be mistaken for contrails -- the long white patterns left by planes at high altitudes) are a common sight when flying in moisture-laden air, and are sometimes mistaken for jettisoned fuel.
As a rule, bigger planes whip up bigger, more virulent wakes. And as you'd expect, smaller planes are considerably more vulnerable should they run into one. For a widebody jetliner, wake encounters are rarely serious; for those like our 19-seater, they are a known and carefully avoided hazard. This is one of the reasons certain radio call-signs in include the suffix "heavy" (as listeners to United Airlines' Channel 9 audio feature will recognize). It's a reminder for crews and controllers alike that said flight requires a wider than normal buffer zone. During approaches, non-heavies following a heavy require at least five, and sometimes as many as six miles of separation. On takeoff, they need two minutes of wait time at the end of the runway. In the United States, "heavies" are those planes whose maximum takeoff weights exceed 255,000 pounds. Outside the U.S., ICAO has its own, marginally heavier ":heavy" designation, though the term is not used over the radio.
When landing behind a heavy jet, pilots of smaller aircraft will, whenever practical, remain slightly above the standard glidepath toward the runway -- a steeper approach. This keeps you above any vortices as they sink. It's nothing drastic, and won't be noticeable to passengers, but your descent is a degree or two sharper than usual. "Half a dot high," in pilot parlance, referring to the instrument markings used to monitor glide angle during approaches. Another trick is to use the wind. Gusts and choppy air will break-up vortices, or otherwise move them to one side.
Landing in Philadelphia, we had our half a dot. In fact it may have been a full dot. We were especially attuned because we knew the preceding traffic was a Boeing 757. Technically, the twin-jet 757 isn't a heavy. A mid-sized jet, it's barely one-third the heft of a 747, 777, or A340. But what it lacks in weight it makes up for with a nasty aerodynamic quirk, producing an outsized wake rivaling or exceeding that of its larger siblings. A 1990 study by the U.S. National Oceanic and Atmospheric Administration (NOAA) pronounced the 757's vortices to be the most powerful ever recorded. Does this look like something you'd want to tangle with -- the plane like a storm unto itself? Check out the spin from that left-wing core.
In 1993, a business jet carrying the president of In-N-Out Burger, a popular fast food chain, crashed at Santa Ana, California, killing the executive and four others. They'd been following a 757. A year earlier, in Billings, Montana, a Cessna Citation jet was rolled inverted after the pilots, on a visual approach, failed to maintain adequate distance from a 757 ahead. In response to these and other incidents, FAA and ICAO rules call for increased separation limits between 757s and other aircraft. Basically, "heavy" minimums apply, meaning, in most cases, an extra mile of clearance on approach.
Not to be outdone, the new Airbus A380, the largest and heaviest (and ugliest) commercial plane to date, will require an even greater parcel of sky. For aircraft stuck behind this beastly behemoth, a recently completed, three-year study recommends up to ten miles of separation for landings, and three-minute runway hold times for takeoff. In airspace choreography, greater spacing means longer delays, and potential reductions in airport capacity. This was bad, if not unsurprising news for the A380 program, already problem- plagued and well behind schedule.
Many modern jetliners are fitted with winglets -- those small upturned fins out at the tips. These devices increase aerodynamic efficiency -- meaning, in turn, economy and range -- and one of the ways they do so is by mitigating the severity of wingtip vortices. Generally, a winglet-equipped plane will produce a more docile wake than a similarly sized plane without them. Although the 757 is no longer in production, winglets are available as a retro-fit. The package costs about $250,000 per airplane, not including downtime, and reduces fuel consumption by as much as five percent. Reductions in the severity of the plane's wake are tougher to quantify, but certainly welcome.
Back in Philadelphia that day, we thought we'd done everything right. Plying the busy Northeast corridor, avoiding wakes from 757s was routine. So, what happened?
Beats me. Perhaps the Boeing too, for reasons unknown, had come in slightly high. And it was one of those windless, dead-air kind of days, which would have allowed any vortices to linger. In any event, nobody was hurt, and not a single barf bag was removed from its pouch. It was, you could say, just one of those things.
Our experience was highly unusual in its severity, but typical in that it lasted merely a few seconds, leaving everybody unscathed. I haven't raised this topic to scare you, and of the accidents in which wakes are listed as primary cause, the vast majority have involved small, non-commercial aircraft. One exception was a crash in 1972, when a DC-9 got too close to a Lockheed L-1011. But widebody jets like the Lockheed were new at the time, and the dangers of wingtip vortices weren't fully understood.
Five years ago, wake turbulence played a role in the crash of American Airlines flight 587 in New York City, but exactly how is easily misconstrued. Moments after takeoff from JFK airport, the flight was struck by vortices from a Japan Airlines 747 ahead. The crew -- the first officer in particular -- then overreacted, repeatedly commanding full, back-and-forth deflection of the rudder, overstressing the tail and causing it to separate. The overreaction itself was traceable, in part, to design of the Airbus A300's rudder system -- engineered in such a way that pilots could inadvertently summon violent deflections with relatively light inputs. There may also have been a preexisting stress crack in the tail's composite skeleton -- the result of a powerful turbulence encounter years earlier -- though this has never been proven. A300 crews have since been retrained.
Today, pilots and air traffic controllers are well versed in the hows and how-nots of wake turbulence. Existing protocols, it would seem, are effective ones, and statistics bear this out. Over the past two decades, the world's commercial aircraft fleet has doubled, with small jets and turboprops -- the types most vulnerable to upset -- accounting for almost a third of the total. Yet there has been no corresponding uptick in wake-related accidents. That's the beauty in a discussion like this: we can marvel at the fury of those horizontal tornadoes, knowing how easy they are to avoid. Knowledge is power, for crews and for anxious passengers.
One final and humorous footnote to what happened in Philadelphia: To make the incident even more memorable than it already was, the crew of that 757 got to watch the whole thing. As we did our little dance on final approach, they had already turned clear of the runway and were taxiing inbound along the parallel taxiway. In other words, the pilots were looking directly at us. They had to have witnessed our hapless fluttering, and probably knew what caused it. Whether they felt bad, or were laughing, is something I'll never know.
End
© Patrick Smith 2005, 2006. The above text is adapted from ASK THE PILOT, and appeared previously on Salon.com