How a MD11 Design Quirk led to DISASTER! | FedEx Flight 80

The morning of March 23rd, 2009 was a

relatively quiet one at Narita

International Airport outside Tokyo. The

time was just before 7:00, and the

horizon still had that pale, washed out

color of a spring dawn. From the control

tower, the field looked normal enough.

The runway lights were still glowing,

and in the air above Japan, both

passenger and cargo jets were lining up

for their morning arrivals into what was

then Tokyo's main airport.

FedEx Flight 80 was one of those

flights, who was now nearing the end of

its overnight journey. And for the two

crew on board, this flight had so far

been completely unremarkable.

As the aircraft descended down, light to

moderate turbulence continuously rocked

the aircraft, but not outside their

experience and capabilities.

Now, what was about to happen to this

flight would take only a few seconds,

but leave the aviation world stunned.

And in the aftermath, investigators

would find a story that was much bigger

than just one flight.

This is a story about design

compromises, training gaps, possibly a

bit of fatigue, and a plane with a quirk

that had already caused an accident to

this exact airline more than a decade

earlier.

FedEx Flight 80 started off as a routine

operation over from Guang Xiao, China

towards Narita, Japan, just north of

Tokyo. In the cockpit were two pilots,

both with long and successful resumes.

In the left seat set a 54year-old FedEx

veteran captain whose career had begun

in the US Marine Corps where he had

spent years flying Marine fighter

aircraft before moving into commercial

cargo. By the time of this flight, he

had accumulated more than 8,000 total

hours of which over 3,600 were flown on

the MD11. He was known among his

colleagues as a nice guy who was steady,

methodical, and someone that could be

counted on to fly the aircraft by the

book. Years of experience had given him

not just technical skills, but also the

kind of quiet confidence that comes from

doing countless approaches in

challenging conditions, including

landings on ships at sea in all types of

weather conditions.

Beside him, he had a 49year-old

experienced first officer. Like the

captain, he had come from a military

background, having flown the absolutely

massive C5 cargo jets for the US Air

Force. After leaving the service, the

first officer had then transitioned onto

commercial flying, eventually joining

FedEx. And at the time of this flight,

he had logged over 5,000 hours, but only

about 879 of those on the MD11. So

compared to the captain, he was still

relatively new on the type.

Now, I usually consider someone

experienced on type at around the 1500

hour mark. So, with 879 hours, this

first officer certainly wasn't a rookie.

But he wasn't fully experienced with the

MD11 either. And it's important to

remember that cargo operations with long

haul aircraft often means long hours

without necessarily doing many landings

or takeoffs. But given this first

officer's previous military cargo

experience, which meant that he was used

to demanding landings and the discipline

of flying complex machines at night, the

first officer would be operating as

pilot flying on this flight with the

captain as pilot monitoring. And that

was even though the weather on arrival

looked like it could be quite

challenging.

The airplane that they had been assigned

to for this flight actually had a quite

complex and interesting history.

It was registered as November 526 Fox

Echo and it had rolled off Macdonald

Douglas production line back in 1994 as

a passenger MD11 before then being

converted into a freighter around a

decade later. In between, it had also

spent time as NASA's test bed for an

experimental system designed to control

an aircraft purely through engine thrust

in case of a total flight control

failure. something that my colleague Ben

actually covered over on Mentor now,

which you should really check out after

this video. Anyway, by the time FedEx

had acquired it, the jet was a wellworn

but capable machine, one of dozens of

MD11 freighters that formed the backbone

of the company's global fleet. But the

MD11 itself was always a point of debate

among pilots. Born in the late 1980s as

a stretched modernized successor to the

DC10, it promised to launch Macdonald

Douglas into new levels of success and

allowed them to compete with the likes

of Boeing and Airbus. The MD11 had

computerized a lot of the functions of

the DCT10, allowing it to move from a

threeperson flight crew to two, whilst

also promising longer range and higher

efficiency than its predecessor. But to

achieve those numbers, Macdonald Douglas

had also made some compromises. The

engineers had fitted bigger, more

powerful engines, which gave the MD11

plenty of power, endearing it to its

pilots, but they had also reduced the

size of the tail plane to cut down on

drag and also added some slotted

surfaces, among other things. The result

of all of this was an airplane that

could fly further and faster than its

parent, but at the cost of becoming a

little bit more twitchy and somewhat

hard to handle, especially during

difficult approaches.

Now, a lot of people have attributed

these handling issues to that smaller

stabilizer that I just mentioned, but in

reality, it had more to do with how the

weight was distributed behind and in

front the center of gravity, which gave

the aircraft a bit of a barbell effect,

making it a bit harder to get a pitch

movement going. And once that pitch was

going, it also made it harder to stop,

which is worth remembering.

Anyway, pilots who flew it quickly

learned to respect these quirks, meaning

that most landings were uneventful. But

if you got behind the airplane near the

ground, things could go wrong very fast.

And at the time of this story, there had

already been several landing mishaps

with the MD11 fleet, specifically due to

large bounces.

Now, this crew of course knew this. Both

pilots had gone through bounce recovery

training in the simulator years earlier,

so they knew the drill. Don't force the

nose down. Don't overcorrect. Just hold

the nose at around 7 1/2° and if things

gets unstable, just go around. Same as

if the approach wasn't stable, go

around.

So in the cockpit of FedEx Flight 80,

there was no sign of apprehension as

this flight went on the way. Their three

engine workhorse was doing exactly what

it had done for years, safely hauling

freight on the overnight trunk routes

that kept FedEx global network running.

But the two pilots were, however,

probably looking forward to getting some

rest in Japan because, as is the case

with most air cargo operations, they had

started this flight in the middle of the

night, leaving China just before 2:00 in

the morning after only having a few

hours of rest from their previous

flight. On paper, this was a very

straightforward leg, including a normal

departure and climb out from Guangao,

followed by a three and a half-hour

cruise north over the South China Sea,

eventually crossing into Japanese

airspace for an approach and landing

just before dawn. The radios crackled

only occasionally, but air traffic

control handoffs, and the weather on

route was calm. But while the plane

itself didn't care about this early

departure, the pilot controlling it

likely did.

On the cockpit voice recorder about 45

minutes before landing, the first

officer brought up the subject that

every pilot of an overnight cog run

knows all too well. He remarked to the

captain that they were both tired, and

the captain agreed. The first officer

then added something along the lines of,

"If it gets too quiet, make some noise."

An off-handed comment showing their

simple strategy here. keep each other

engaged so that neither of them will

drift into any fatigue induced lapses,

which is how we pilots usually deal with

drowsiness.

As we have discussed before here on the

channel when I cover the fatal accident

of UPS flight 1354, fatigue can be the

dark side of cargo operations.

You see, unlike passenger airlines who

tend to schedule their flights during

daylight hours, since obviously that's

when humans prefer to travel, cargo

carriers instead operate on the back

side of the clock. Packages need to

connect through hub and spoke systems

overnight so that they can be sorted,

processed, and eventually delivered

during the following workday. And as a

result, cargo aircraft tend to leave in

the small hours, transit through dawn,

and arrive just as most people are

getting up for work or even earlier.

For the crews, this means duty periods

that deliberately cut across circadian

lows. So even when the formal rest

requirements are met, the body's

internal clock can still be a little bit

out of sync. Physiologically, fatigue

can look a lot like hypoxia, where

pilots might not realize just how

impaired they are until it's too late.

Reaction times begin to slow, scan

patterns becomes less thorough, callouts

get sloppy, and you might start missing

small cues like a slightly high sync

rate or a drift off the center line,

stuff that would normally jump out at

you. Now, some of you might interject

here and point out that these two pilots

were based in Anchorage. So, what was

night in Asia would be daytime at their

home base. But these two pilots had also

been flying daytime duties during their

stint in Asia, which would have further

complicated their circadium rhythm.

Now, I'll get back to that issue soon,

but I want to mention something else

that affects everyone who is traveling

internationally, and that's access to

data connectivity. When I started

flying, I used an old Ericson phone and

to connect that for data in other

countries on roaming would have cost an

absolute fortune. So that meant that I

had to buy local SIM cards and fiddle

with swapping them around all the time,

which was a nightmare. But thankfully,

that's now all in my wake thanks to

today's sponsor, Holofly. Holofly are

the leading global eim provider keeping

you connected in over 200 destinations

and without you having to swap any SIM

cards or worry about roaming charges.

Whether you are just leaving for a week

or are constantly traveling the world,

Holofly offers flexible data options to

suit your specific needs. Once you have

installed the eSIM, it connects

instantly, which means that you can have

it ready before you even land and you

can still use your original phone number

for calling whilst using Holofly for the

data. All of this, including the fact

that you can cancel your plan at any

time, makes Holofly perfect for pilots,

cabin crew, digital nomads, or really

anyone who lives a fast international

lifestyle spanning the globe. Right now,

Olafly is offering you, my viewers, 5%

off their general eims and a full 10%

off their monthly subscription if you

use the code mentor. Just scan this QR

code or use the link in the description

below. I highly recommend them for your

next trip. Thank you, Holafly. Now, back

to the video.

For the first couple of hours of cruise

in this flight, it was all perfectly

manageable. The autopilot was engaged,

the systems were humming, and there were

a few check-ins with controllers as the

flight crossed into Japanese airspace.

But as the sun began to rise, the

fatigue would have likely hit harder

since early morning is one of the body's

natural low points. Now, with all of

this said, none of this was unusual. It

was just the life of a cargo pilot. Both

of these pilots would have been well

used to this. They were not unusually

sleepdeprived. The report showed that

they had both received adequate sleep

and rest in the days leading up to the

flight. But I still want to point it out

because as Narita loomed ahead of them,

they were heading into the most

complicated part of the flight at the

very moment when their mental sharpness

would be most likely to be compromised.

But whether this had much of an effect

on what was about to happen, we will

sadly never really know.

Anyway, as FedEx Flight 80 descended in

towards Narita's busy morning airspace,

everything, like I said, looked

completely routine. The checklists were

completed as normal. The coming approach

was briefed, and the radios were buzzing

in the background with the calls from

loads of other arrivals. So, these two

pilots were now busy doing what they had

done countless times before, setting up

their navigational aids, calculating

landing distances, and programming their

flight management computers to land

their heavy aircraft safely after a long

night flight. As this was going on up in

the air, Narita Airport was just

starting to come alive. The runways were

already busy with morning arrivals, and

controllers in the tower were juggling a

mix of passenger flights and freighters

streaming in from across Asia. But there

was one factor that made this morning a

little bit extra tricky, and that was

the wind.

It was not only strong coming out of the

northwest at around 30 knots, it was

also gusty with variations in wind

speeds of up to 20 knots. Now, gusty

winds create a number of different

complications for us pilots. If the wind

is coming at us from the side, known as

a cross wind, the changes in wind speed

will make it harder for us to keep the

aircraft tracking down the center line,

as we constantly need to adjust our crab

angle in response to these changing

winds. Picture trying to drive a boat

towards a specific point on a riverbank

if the water speed around the boat keeps

changing.

And if the wind is not coming from the

side, but instead from the front, the

changing wind speeds will come with a

different problem. You see, the aircraft

flies in relation to the surrounding air

primarily. That's why we display our air

speed on our primary flight displays.

And in still air, we can change that

speed with the help of our engines and

or with the help of our pitch. But if

the wind is changing rapidly, that means

that our indicated air speed will as

well. and we might have to start chasing

it with the help of rapidly changing

engine inputs.

In the real world though, it's rarely

either or when it comes to these issues

as we almost always get both crosswinds

and headwind changes at the same time.

So, flying an approach with strong and

gusty winds is always a challenge, but

that's also what makes it so fun. Now,

like I mentioned in the beginning, these

winds were forecasted and would

therefore not have come as a surprise

for this crew. They also received at

least 180s report as they were preparing

for the approach, which was then

followed by several updates from air

traffic control as they kept getting

closer.

So, the initial part of the final

descent seemed to also have been

completely normal. And the aircraft soon

found itself approaching the ILS whilst

initiating its configuration. And with

the two pilots finishing up the approach

and landing checklists

at time 0643, which was about 5 minutes

before FedEx Flight 80 was due to land,

Narita tower passed along a pilot report

from another airplane which had just

landed on runway 34 left. The same

runway that these two pilots were now

approaching.

that crew had encountered plus or minus

15 knots of wind shear on final below

2,000 ft. So aboard this flight deck

that would have meant that the air speed

seen by the pilots would have likely

been quite erratic. One moment adding 10

or 15 knots, the next moment taking it

away. Now these conditions are not by

themselves dangerous if they're handled

well. Plenty of flights every day land

in similar weather conditions. And there

is a difference between changing gusty

winds and actual wind shear which is

normally defined as a sudden drop of 15

knots of air speed or more a sudden

displacement of more than one dot on the

ILS glide slope. A windshare warning or

extreme thrust lever positions held for

prolonged periods of time. If a full

windshare is identified then a windshare

escape maneuver should be executed. But

even before those criteria are met, it

can be quite hairy to fly in these

winds. As I explained before, anyway, 2

minutes later, at time 0645, the

controller then updated the pilots

again. This time, the wind came from

320° with a speed of 26 knots, gusting

38 and dropping as low as 16. That's a

spread of more than 20 knots between the

peak of the gust and the lols. And it

should be mentioned here that with these

type of strong winds, the air will move

over buildings, trees, and hills below

and form vortices which often leads to

increasing turbulence as the aircraft

descends. So the descent down the glide

slope would have been quite eventful in

this case. And at time 0646, Narita

tower finally cleared flight 80 to land.

Winds at this time was called at 320° at

29 knots, gusting 36 with minimums of

17.

The tower controller's cadence when

delivering these messages was steady and

professional, but with a clear message.

This wind was not calming down.

Inside the cockpit, the captain now

reconfirmed the runway. Runway 34 left.

Clear to land, followed by a word with

high significance here, stabilized. And

after that, he just laughed a bit. That

tiny laugh caught forever on the cockpit

voice recorder is one of those small

human details that stays with

investigators and anyone really who

studies this flight. I really want to

point out here that something like that

wasn't a sign of carelessness or

dismissal. We pilots often use humor in

moments of tension. It's a way of

acknowledging the challenge without

making it overwhelming. The cockpit is

after all a workplace and like most

other workplaces, some dark humor can

sometimes arise in very serious

situations. So this did not in any way

mean that the pilots were being

unprofessional here. Now at this time

the air speed was swinging beyond the

plus or minus 10 knot window typically

used for determining if an approach

really is stabilized.

The gust spread was wide. windshare

reports from other aircraft were fresh,

none of it catastrophic on its own, but

together they created a scenario where

even a small misjudgment could quickly

snowball. The captain had called

stabilized as they passed the landing

gate when that decision had to be made.

And with those speed variations that

they saw, they could have gone around

there. And if they had done that,

there's a possibility that none of what

was about to happen would have occurred.

But that call didn't come. likely

because both of these men believed that

they could land as plenty of planes

before them had already done that

morning. And here I want to take a

moment and say something about

hindsight.

You see, from the safety of our chairs

and sofas when watching something like

this, it's easy to say, well, why didn't

he just call for a goound? It's one of

the first conclusions people reach when

they look at approaches like this, but

from inside of the cockpit, it's not

nearly that clearcut.

Airlines teach stabilized approach

criteria as binary. Either you're stable

or you're not. And if you're not within

the criteria for altitude, speed, sync

rate, and configuration, you go around.

End of story. But in practice,

implementing that standard in very gusty

winds is far from binary.

Every gust moves numbers. Air speed

needles swing up and down. Sync rate

fluctuates. And if we were to execute a

goaround every single time a gust pushes

us a few knots outside of the stabilized

window, we'd likely never land during

some of these type of weather

conditions. And wind is often widespread

over large geographical area. So

diverting to an alternate might not make

things any better.

So in those kind of conditions, the only

workable approach is to sometimes judge

the average and look at the overall

picture. We will work together as a crew

to assess whether over time the aircraft

is on speed, on slope and under positive

control. And that type of judgment

require experience and a willingness to

accept small excursions trusting that

the averages will still meet the

standard. So if the pilots would have

rejected this approach, they could hold

because they certainly had plenty of

fuel to do so, but there was no

guarantees that the conditions would

improve for them.

Yes, the winds were high, but not beyond

published limits. Yes, the gusts were

uncomfortable, but not outside of the

operating envelope, and the aircraft

remained on glide slope and relatively

stabilized. So, this crew just elected

to continue. And it's kind of hard to

blame them.

But one thing that could have been

considered here was for the more

experienced captain to take over the

controls and fly this approach.

Today we have concepts like monitored

approaches where the first officer can

still fly the approach and the captain

then takes over for the landing under

certain conditions but that was likely

not something that FedEx had considered

at the time. Now why the captain didn't

consider flying the whole approach

himself will never be fully known but he

must have just assessed that his first

officer was more than capable of

handling it by time 06 48 and 3 seconds.

That was just seconds before touchdown.

The tower controller then made one last

call. Winds 320 to 0 degrees at 27,

gusting to 34, minimums at 18, meaning a

slight crosswind gusting nearly 20 knots

above the lols right as the aircraft was

passing overhead the threshold.

Now before we get to the landing, I want

to take this time to discuss why landing

this particular plane was so tricky.

We have touched on it a little bit

before, but to really understand what

was about to happen, we need to step

back from that windy morning in Tokyo

and take a more hard look at the

airplane itself. Because this accident

wasn't just about tricky winds and

possible fatigue. No, it was also about

the MD11 and the compromises baked into

its very design. You see, the MD11 was

Macdonald Douglas's attempt to keep up

in a world that was rapidly passing them

by. By the late 1980s, widebody air

travel was being reshaped by a new

generation of aircraft. The Boeing 767,

the Airbus A310, and soon after that,

the Airbus A330, and the Boeing 7. These

were twin engine long haul jets designed

around the latest in aerodynamics and

engine technology able to cross oceans

with lower fuel burn and less

maintenance which was clearly the

future.

Macdonald Douglas on the other hand had

the DC10 a product of the 1970s which

was reliable powerful and widely used

but by the late 1980s it was definitely

showing its age. Now, rather than

starting fresh with a clean sheet twin

like Boeing and Airbus had, the

cashstrapped manufacturer instead tried

to stretch and modernize what they

already had. And eventually they

released the MD11. Now, on paper, it

looked competitive. a longer fuselage to

carry more passengers, a redesigned wing

with winglets for better efficiency,

updated avionics to cut the cockpit crew

down from three to two, and Douglas had

actually wanted to add an even more

modern super critical wing to the new

aircraft, but had to settle with a

beefed up wing with winglets due to

budget constraints.

The MD11 was also equipped with three

big new engines, either Pratt and

Whitney PW4000s or General Electric's

CF6s,

which gave it a lot of available thrust.

And with this, Macdonald Douglas

promised the airlines the best of both

worlds. an aircraft similar to the

familiar DCT10, which they already had

supply chains and maintenance setups

for, but with longer range, more

efficient, and a fully modern cockpit.

But sadly, the airplane then turned out

to not be able to live up to that sales

pitch. Almost as soon as it entered

service, airlines found out that the

MD11 couldn't really meet its promised

range numbers. The jet simply burned

more fuel than expected and it even

lagged behind its quad engine rivals

like the Boing 747400 and the Airbus

A340.

So some longhaul missions it was

supposed to be able to do like Europe to

Asia non-stop just wasn't achievable

without payload restrictions. So the

orders soon started to dry up. And to

make matters even worse, new airtops

regulations were then also released in

1985, which meant that for the first

time ever, twin jets could now cross

huge oceans along the same flight path

as three and four engines jets had done

before, rapidly shrinking the demand for

the MD11 even further. But the real

issue for the MD11's pilots wasn't the

range or its economics. It was handling.

Macdonald Douglas had cut down the size

of the horizontal stabilizer at the tail

to reduce drag and squeeze a little bit

more of efficiency out of it. And also,

as I mentioned earlier, the weight

distribution had changed slightly,

making the aircraft pitch controls a bit

more sluggish at certain speeds and

configurations.

The engineers had done their math,

though, and they believed that the

smaller tail plane and more powerful

engines would still work because the

MD11's updated flight control systems

would help to compensate for it. You

see, the MD11 came equipped with a suite

of computerized augmentation systems

designed to smooth out what the

aerodynamics alone couldn't fix. One of

the most important of these systems was

the longitudinal stability augumentation

system or LSAS which was intended to

provide pitch attitude hold out trim and

increased pitch damping. And this system

continuously monitored the aircraft's

attitude through the flight control

computers. If for example the MD11 would

start to drift more than a couple of

degrees away from its trimmed pitch,

Elsas would automatically feed in

elevator inputs to hold the nose where

it was. And if turbulence bumped the

nose up or down, well then Elsas would

resist that motion and try to return the

airplane to its previous attitude.

So in theory, this system could give the

airplane the stability of a bigger tail

without the drag penalty. The airplane

also had a fuelbased trim system, and

Elsas could automatically command the

fuel to be transferred in or out of a

trim tank in the tail, effectively

shifting the aircraft's center of

gravity to relieve elevator loads. But

that was mainly a cruise feature. And

during the descent, all fuel was

transferred back from that tank in

towards the middle tanks, making the

aircraft more stable again. Now,

alongside this Elsa system, the MD11

also had advanced outer throttle logics

tied into a flight management system

that was claimed by Macdonald Douglas to

be state-of-the-art for its day. On

paper, all of this was meant to reduce

workload and to make the aircraft fly

like a twin in spite of its three engine

layout and unusual balance. But the

reality was that the automation had some

limits. Elsas could keep the pitch

steady under normal conditions, but it

wasn't really designed to rescue the

airplane during sharp maneuvers. It

actually only had a 30% gain below

15,000 ft and could be overpowered by

just 2 lb of force on the control

column. So, in a scenario where, for

example, the pilots were pushing forward

while Elsa was trying to dampen

movement, the pilot inputs would not be

counteracted, leaving the airplane to do

exactly what the pilots commanded.

So that's why a plane that looked fine

on paper could become something slightly

different when it was being hand flown

close to the ground. All of the features

who had been added to increased

stability at high altitudes were then

less effective and the dynamic issues

affecting most large aircraft would

still be very much there leading to

longer response times during large pitch

inputs especially at lower speeds.

And if that happened, those slightly

delayed responses could then be

interpreted by the pilot as the aircraft

needing more input, which could lead to

a much bigger input before the response

actually came.

This led to a jet that was notoriously

difficult to land well in tricky

conditions. It demanded precision all

the way down through the flare and

touchdown, and small errors could become

big ones very quickly.

Also, where a Boeing 767 or an Airbus

A330 of similar weight might cross the

threshold at 135 to 145 knots, an MD11

often needed 155 to 165 knots to do the

same thing. And with gusty winds, the

wind additive needed could then run the

speeds even higher than that, meaning a

longer runway needed for landing. Now,

the aircraft needed those higher speeds

for its maneuverability. And if those

speeds would bleed off early for

whatever reason, the sluggish effect in

pitch control could become even more

pronounced.

A side note here was also that the outer

throttle was used throughout the landing

in most airlines on the MD11 and it had

been programmed to fit the touchdown

requirements for an outland, meaning a

landing using both the autopilot and the

outer throttle. To do this, testing had

concluded that the outer throttle should

start reducing the thrust back towards

idle for the landing as early as 50 ft

above the runway. But that would also

lead to a large speed reduction during

the actual flare, which would happen a

little bit later. And if the landing was

flown then without the autopilot

engaged, the auto throttle would still

start reducing the thrust back to idle

at 50 ft, leading to lower speeds that

would again complicate things for the

pilot. Actually, the only airline who

never suffered any serious landing

mishaps with the MD11 was American

Airlines. And it is entirely possible

that their procedure of always

disconnecting the outer throttle and

land fully manually was the reason

behind this.

In any case, the pilots described that

the effect of all of this was flying an

airplane that always seemed to be a

little bit ahead of you. On approach, it

carried more momentum than crews were

used to, which gave them perceived

lesser pitch authority to finesse it

down. If the pilot flared too late or

too early and tried to reverse it, it

could feel like the aircraft wasn't

responding enough, leading to large

overcorrections. And when things went

wrong, well, then they tended to go

wrong in the same way with a bounce.

So, the MD11 quickly earned a reputation

for violent bounce landings. If the

airplane hit hard and then rebounded,

the pilots could quickly fall behind the

aircraft and make pitch inputs that

would just make the situation worse. And

several MD11s have been destroyed in

exactly this manner. A China Airlines

cargo MD11 at Hong Kong back in 1999 had

broken apart after it had come down

hard, bounced, and then flipped upside

down. And the same for a Lufansa cargo

MD11 in Riyad in 2010.

Each time the pattern had repeated a

firm touchdown, a bounce, a nose down

input, and then structural overloads as

the aircraft slammed back into the

runway with way more force than the

surprisingly robust landing gear

structure could handle. All in all, the

MD11 have suffered more than a dozen

major accidents and incidents, many of

them during landing. So by the mid

2000s, it had one of the worst hall loss

rates of any modern jetliner.

Now again, FedEx knew this history. In

fact, they had already lost an MD11 to a

botched landing, FedEx Flight 14 back on

July the 31st, 1997, which happened

during a landing at New York Airport in

New Jersey, United States. On that

approach, the jet had touched down long

and the captain, worried about running

out of runway, had tried to get the

airplane down onto the ground with far

too much energy still on board. The

aircraft bounced only 5 ft or so, but

the nose gear was then pushed down

because the crew, who was sitting far

ahead of the aircraft's center of

gravity, thought that the main gear was

actually on the runway.

This had led the captain to pushing

forward on his yoke to get the nose

down. And that forward push allowed the

main gear to slam into the runway with

318%

more force than the limit design load,

causing it to collapse, the aircraft to

flip over and immediately catch fire.

Now, luckily, both pilots had survived

that accident, but the airplane was

completely destroyed.

The pilots on board FedEx Flight 80

obviously also knew about this accident

and so did like we mentioned before

FedEx training department.

Both of these pilots had therefore

received bounce recovery training in the

simulator and they knew the rules. Keep

the pitch angle at 7 1/2° add power and

if things got unstable go around. But

critically this training was performed

only once and was not given as part of

their recurrent training. The training

was also focused more on preventing tail

strikes than anything else. As the

ground spoilers on the MD11 could cause

a pitch up momentum during a bounce,

something that Elsas was later augmented

to help with a system known as the PAP

or pitch attitude protection.

Anyway, back to the flight. With the

wind still swirling across Narita, FedEx

Flight 80 now continued into the last

moments of its approach to runway 34

left. From the tower's vantage point,

the M11 looked stable enough, but to the

pilots in the cockpit, it was likely

still unhampful.

As they passed the 200 ft, the airplane

was still moving at about 178 knots,

which was way faster than the briefed

VREF of 164 knots. But by 100 ft, the

winds quickly shifted, meaning that the

speed now reduced down to 154 knots,

well below the minimum. And as that

happened, the auto throttle didn't have

time to react to it before going into

its pre-programmed mode at 50 ft.

And remember what I said about the

MD11's characteristics if the speed

became too low.

Yeah, this is where its mean character

could really show itself. Unlike a 767

or an Airbus A330 which tend to forgive

small errors in the flare, the MD11

demanded precision. And on top of that,

the cockpit, which sat far ahead of the

main gear, made it harder for the first

officer to judge his height visually,

something that would soon become an even

bigger problem. Now, the captain acting

as pilot monitoring made no intervention

at this point as the speed was

decreasing. He had already called the

airplane stabilized a minute earlier and

now both of the pilots seemed committed

to bringing this aircraft down to the

runway. The radar altimeter call outs

now started rolling down in quick

succession. 100 50 40

and at 30 ft the flare should have

started with a smooth pull on the York

to arrest the centrate. But the first

officer now was a little bit late. maybe

focused on keeping the picture right and

managing the swing of the airspeed

needle. He waited until 20 ft to start

his flare. And when he did, he likely

realized that he was a little bit late

and pulled a bit more than normal. But

the aircraft, due to its inertia, didn't

respond immediately. So, he pulled more,

leading to the nose suddenly moving up

more than he wanted, causing him to

reverse his inputs and push his yoke

forward.

And these slightly out of sync pitch

inputs were what really started the

whole accident sequence.

The aircraft now responded to the pitch

down command, increasing the vertical

speed just as they passed over the

runway. And that meant that the radar

altitude call outs now all came rapidly

with less than a second apart. A classic

forboarding of an impending hard

landing.

With no added thrust to cushion the sink

rate, the airplane now quickly moved

downwards and the airplane soon struck

the runway hard on all three sets of its

main landing gear. The flight data

recorder measured 1.63 gs of force as

430,000 lbs of jet, cargo, and fuel met

the concrete. As we mentioned earlier,

the nose sits so far forward of the

center of gravity on the MD11 that when

the main wheels hit the ground, the

cockpit could often feel strangely

disconnected.

Many MD11 pilots have described not

realizing that they bounced until it was

too late, so the airplane could feel

like it had settled down on the runway

when in reality the main gear had

already bounced back up into the air

again. And that is almost certainly what

happened here.

After the first touchdown, the airplane

bounced back up in the air slightly, but

the first officer may have believed that

they were just rolling out normally. So,

he was now waiting for the nose wheel to

start to come down.

And it was at this moment, right after

the first bounce, that the first

officer's reaction to the situation

created the next critical step in the

failure chain. You see, his control

inputs showed that instead of holding

the nose steady at 7 1/2° or

disconnecting the outer throttle to add

thrust in order to cushion the bounce,

he now instead again pushed forward on

the yoke.

The only logical explanation for doing

that would be that he really thought

that the aircraft was on the ground, so

he wanted to get the nose to start

drotating.

Now, when it didn't respond again due to

inertia and because the aircraft's

bounced trajectory, he added even more

forward pitch, which with a delay soon

turned the nose of the aircraft down

towards the concrete below.

So, at the time 0648 and 22 seconds, the

MD11 struck the runway a second time,

even harder than before, and with a

fragile nose gear now impacting first,

leaving some debris behind. The main

gears came after this time impacting

with a force 2.2 times their design

limit, bottoming out the struts and

transferring the huge load directly into

the fuselage.

This impact reverberated throughout the

fuselage at a low chest thumping

frequency. Investigators later estimated

the vibration at around three hertz with

the entire airplane flexing and

resonating like a struck drum. But up in

the cockpit, almost 100 ft ahead of the

main gear, the sensation probably wasn't

quite as clear. The ground spoilers,

which had started to deploy up

automatically after the first touchdown,

now snapped closed again as the airplane

lifted back into the air. And the

reverses that had just begun to spool

up, were cancelled, too, as the weight

on wheel sensors suddenly detected no

load on the struts. Now, the touchdown

of the nose and the impact of the main

gear launched the airplane back up much

higher than on the first bounce about 16

ft off the runway. And for a brief

moment, this situation could still have

been saved if either of the pilots had

recognized the danger, added power, and

held the nose up, or even executed a

goound.

That would have caused the MD11 to halt

the perpoicing motion. It might still

have touched down as part of the

goaround, but eventually it would have

started to climb away, damaged perhaps,

but largely intact.

But again, no power came and no call for

a goaround or my controls from the

captain. Instead, the nose started to

soon drop again. The MD11 had pitched up

briefly to around 7° before it started

moving nose down again. And at that

altitude, especially in relation to the

wide 60 m runway, the bounce likely

again felt smaller than it actually was

in the cockpit. And the first officer's

inputs were therefore again not in sync

with the aircraft.

So at the time of 648 and 27 seconds,

the third impact came, which was the

hardest of them all.

The Gmeter spiked past 3Gs, and since

the wings were now unloaded due to the

first officer's forward pitch, the

ensuing force on the landing gear went

well beyond its structural limits. This

was the breaking point. On that third

slam into the concrete, the left main

landing gear suddenly failed. But it

didn't fail cleanly.

Macdonald Douglas had designed it to

shear away if it was struck by something

horizontally while rolling down the

runway, but vertically it had no such

safe failure modes. By the time of this

accident, planes were being built,

requiring to have gear that could fail

safely in both axis. But the MD11's

construction predated those rules, so it

had no vertical shear design.

This meant that instead of detaching the

gear now translated all of that enormous

force directly into the wing spar which

is a main component of the wing

structure.

That spar promptly failed due to

structural overload causing the left

wing to bend and rupture the wing tank.

Nearly 50,000 lbs of jet fuel

immediately spilled out into the sparks

flying from the now scraping metal on

the left side and that caused a raging

fire which quickly spread along the wing

and behind the careening aircraft.

With the left wing compromised, the

right wing still generating lift and the

left gear dragging underneath. The MD11

now quickly started jawing to the left

whil also violently rolling over its

detached left wing. This meant that in

just a few seconds, the airplane was

almost fully inverted. And at times 0648

and 28 seconds, the cockpit voice

recorder captured the chaos. A deafening

bang. The master warning blaring and the

captain's final words. Fire. Oh. The

ground proximity warning system provided

a final bank angle bank angle warning

before the recording abruptly ended.

By time 06 48 and 29 seconds, the

aircraft was then fully upside down,

sliding along the runway, already

engulfed in flames. The inverted fuse

lodge was skidded until friction and

structural collapse brought it to a halt

beside the runway. And the fire then

spread rapidly, fueled by a torrent of

jet fuel spilling out from every

ruptured tank.

from the tower. The controllers had

witnessed this horrific scene and they

actually activated the airport fire

services even before the airplane had

come to a complete stop. The fire trucks

were rolling within seconds. But the

blaze was so intense that there was no

way for them to reach the cockpit in

time. By the time the firefighters

finally got close, nearly 40 minutes

later, both of the pilots were sadly

gone.

Now, it is important to remember here

that this whole sequence that I've just

told you took less than 10 seconds.

That's from the time that the aircraft

crossed the runway threshold until it

was upside down on fire in the grass.

And that's actually less time than it

took me to say that last sentence.

After this accident and others with

similar signatures, the operators of the

MD11 adjusted. More emphasis was put on

high sync rate prevention, clearer gore

on triggers in gusty conditions and

recurrent practice for the exact

scenario that had led to the accident of

FedEx Flight 80.

Now I often talk about mental models for

decision making and problem solving like

Pio for example. But those models can

only be used when there is time to

evaluate the situation, gather

information and then discuss options

with your team. This crash did not lend

itself to any of that. In a situation

like the one that these pilots found

themselves in, the only thing that they

had to rely on was their judgment,

training, and experience, which sadly

fell short. And that is part of what

makes this accident so frustrating and

also so understandable. Now, I have said

it before and I'll say it again.

Aviation is as safe as it is today

because we add healthy margins to all

potential risk factors. So if there's

one big lesson to take away from this,

it is to always know when you start to

eat into your margins and use that

awareness to stop a failure chain from

ever developing.

The final report stated that the direct

causes of the accident were the large

nose down elevator inputs made by the

first officer before both the first and

the second touchdown. This led to the

first bounce and also the large force

and no gear touchdown in the second one

and his overall large pitch inputs

without added thrust throughout this

sequence eventually led to the outcome.

But report also cited the winds, the

fluctuating air speed and the pilot

monitoring's failure to intervene as

contributing factors here. The design of

the aircraft itself was not mentioned as

a crucial factor since its

characteristics were known and I want to

send a special thanks to Captain Shem

Malquist here for contributing

invaluable insights into the handling of

the MD11. Without him, this accident

would have been much harder to explain.

Now, if you like this video, I would

love for you to have a look through my

backbook where I think that you will

find a lot of other interesting videos.

there. You will also see that the

quality of the channel has changed a lot

from the very first documentaries that

we did to where we are today. And that's

only possible due to the excellent team

that I have behind me. Ari has helped me

with the script writing for this

episode. And either Dom or Mo have then

made the editing that you're seeing with

the help of our team of dedicated

simulator pilots, Andrew, Nick, and

Samir who drew some truly truly

fantastic work. Now, I'm telling you

this because none of this would be

possible without the fantastic financial

support that we're receiving from our

patreons who also get to watch these

videos before anyone else and come with

important feedback. If you also want to

join this crew, well then click on this

QR code or go to

patreon.com/join/mentorpilot.

I would love to see you in our next Zoom

Hangout.

My name is Peter Hornfeld and you're

watching Mentor Pilot. Have an

absolutely fantastic day and I'll see

you next time. Bye-bye.