On 7 January 2017, a Bombardier Challenger 604 was cruising at 34,000 feet over the Arabian Sea, roughly 500 nautical miles from the nearest land. An Emirates A380 passed 1,000 feet overhead in the opposite direction, entirely legally.

Forty-seven seconds later the Challenger rolled through 42 degrees in a single second, pulled 1.6 g and then minus 3.2 g in the next, pitched from nine degrees nose-up to twenty nose-down, and fell roughly 8,700 feet while rotating several times about its longitudinal axis. Both inertial reference systems failed. The aircraft landed at Muscat and never flew again: the airframe had exceeded its certification design load limits.

Nothing had touched it. The A380 had left behind a pair of invisible, counter-rotating tornadoes, and they were still there three quarters of a minute later.

Where the tornado comes from

A wing generates lift by holding lower pressure above it than below. At the wingtip, nothing prevents the high-pressure air underneath from spilling around into the low-pressure region on top, and that spillage, combined with the forward motion of the aircraft, rolls into a vortex.

The FAA describes the result as two counter-rotating cylindrical vortices, with most of the energy concentrated within a few feet of each core. This is not a by-product that clever design can eliminate. It is the direct signature of lift. An aircraft producing lift is producing vortices, always.

Strength follows a simple relationship: initial circulation rises with weight and falls with wingspan and speed. Which is why the memorable FAA formulation is heavy, clean and slow. A heavy aircraft, flaps and gear retracted so the whole load is carried on the wing, flying slowly, produces the worst wake it is capable of. That describes an aircraft on departure or on final approach rather more accurately than one in cruise.

Peak tangential velocities of almost 300 feet per second have been recorded, roughly 178 knots.

Where it goes after it is made

Vortices sink. The FAA’s current Aeronautical Information Manual says only that wakes from larger aircraft descend at several hundred feet per minute, having dropped the specific numbers it once published. The FAA’s own Wake Turbulence Training Aid is more precise: a sink of roughly 300 to 500 feet per minute for about 30 seconds, after which the rate decreases and approaches zero somewhere between 500 and 900 feet below the flightpath. EASA, writing about the cruise environment, gives a typical sink rate of about 400 feet per minute and persistence of two to three minutes.

Near the ground the behaviour changes and becomes far more dangerous. Within 100 to 200 feet of the surface, vortices stop sinking and begin moving sideways at two or three knots.

This produces the single most counter-intuitive hazard in the subject. A light crosswind is worse than none.

The capitals are the FAA’s. With no wind, both vortices drift apart symmetrically. With a gentle crosswind, the upwind vortex is held stationary directly over the touchdown zone, precisely where the next aircraft intends to be. Above about six knots the wake is swept clear quickly.

The Australian Civil Aviation Safety Authority sets out the operational rules in the briefing above.

The separation rules, and why they differ

ICAO sorts aircraft into four wake categories: Super, currently comprising the A380-800 alone; Heavy at 136,000 kg and above; Medium between 7,000 and 136,000 kg; and Light at 7,000 kg or less.

The FAA uses a different scheme, and the mismatch trips people up. FAA Heavy begins at 300,000 lb, which is close enough to ICAO’s threshold, but the FAA then splits everything below into Large and Small at 41,000 lb and has no Light category at all. An 8,000 kg twin is ICAO Medium and FAA Small. The Boeing 757 sits in a category of its own in practice, classified as Large but with special separation criteria attached, because its wake proved unusually strong for its weight.

On approach, ICAO separations run from 4 nautical miles for Heavy behind Heavy out to 8 miles for Light behind Super. Departure intervals are measured in time rather than distance: two minutes behind a Heavy, three behind a Super, and the FAA is explicit that controllers may not reduce or waive those particular intervals.

Both regulators have since moved to finer-grained schemes. Europe uses RECAT-EU, six categories assigned on maximum take-off mass and wingspan jointly. The FAA went further, to a nine-category system under Consolidated Wake Turbulence, on the reasoning set out in its own order: four miles is appropriate for a 767 following a 747, and unnecessary when the 747 is following the 767.

American Airlines 587, and what it does not mean

On 12 November 2001, an Airbus A300-605R crashed at Belle Harbor, New York, killing all 260 on board and five people on the ground. It had departed behind a Japan Air Lines 747-400 and encountered its wake twice.

The accident is routinely cited as proof of what wake turbulence can do to an airliner. That reading is wrong, and it matters.

The separation was not merely legal, it was generous: the NTSB recorded that the two aircraft were separated at all times by at least 4.3 nautical miles horizontally and 3,800 feet vertically, because the local controller had added extra radar separation. NASA Langley modelling for the investigation put the wake’s age at about 100 seconds and its remaining circulation at between 63 and 80 per cent of initial strength. The NTSB record notes that the lead wake analyst described the encounter as nothing extraordinary.

A normal wake encounter behind a legally separated aircraft was the trigger. The response to it broke the aeroplane. That is a different lesson, and a more useful one.

The NTSB’s own animation of the encounter is above.

The regulator eventually wrote that lesson down, prompted by the A380 and Challenger event fourteen years later.

The part nobody had planned for

The Challenger 604 case exposed a genuine gap. Separation minima are built for the terminal area, where aircraft are heavy, slow and close together. Nobody had written rules for two aircraft passing in opposite directions in the cruise, because nobody expected a wake to survive the journey.

EASA’s subsequent guidance is blunt about the scale of it: wake can be encountered up to 25 nautical miles behind the generating aeroplane, with the most significant encounters reported within 15 miles, and no specific horizontal wake turbulence separation minima exist for en-route flight at all. Cruise wakes from the largest types can descend more than 1,000 feet, which is exactly one RVSM level.

EASA’s other recommendation is the one that runs against instinct: if the pilot reacts at the first roll motion, while still in the vortex core, the roll can be amplified by that initial input.

Two things worth unlearning

The first is winglets. They reduce induced drag and save fuel, and it is widely assumed they must therefore tame the wake. The FAA has tested this and says otherwise: studies during approach and landing show no discernible difference between aircraft with and without winglets, and the AIM describes the effect on wake generation as negligible.

The second is that this is a big-aircraft problem. It is not. What determines whether you survive an encounter is whether your wingspan and ailerons extend beyond the rotational flow field of the vortex. A large aircraft straddles the vortex and rolls slowly. A light aircraft sits entirely inside it, and can be displaced more than 30 degrees in roll with sink rates beyond 1,000 feet per minute.

On 16 December 2021 at Knoxville, a Cirrus SR22 turned base about 1.8 nautical miles behind an Allegiant A320. At roughly 1,000 feet it rolled sharply left through as much as 135 degrees with the nose 30 degrees down. The parachute was deployed with insufficient altitude. The NTSB found the probable cause to be a wake vortex encounter resulting in a roll upset at an altitude too low for recovery, with the failure of air traffic control to issue a cautionary advisory as a contributing factor.

Again, the capitals belong to the FAA. It does not shout often.

Sources: FAA Aeronautical Information Manual Ch.7 §4, AC 90-23G and the Wake Turbulence Training Aid; NTSB AAR-04/04 and ERA22LA089; EASA SIB 2017-10R1; German BFU file BFU17-0024-2X as quoted by the Flight Safety Foundation; EUROCONTROL RECAT-EU Ed. 2.0; FAA Order JO 7110.126B; Hallock and Holzäpfel, Progress in Aerospace Sciences, 2018.