Original subtitles

I remember being amazed how quickly the accident took place. They were flying,

and the next thing you know, the airplane's lost.

Midwest Express Flight 105 crashes seconds after taking off from Milwaukee,

Wisconsin. All 31 people on board are killed.

And what do you say to the public about the crashes that we've been having?

1985 becomes one of the deadliest years in civil aviation.

It's up to investigators to put the public at ease.

This is how the plane hit the ground.

When they examine the engines, they find a component prone to failure.

Well, when we heard that this part had failed 45 times before, it's a big deal.

But the flight data turns the investigation on its head.

So the ruptured red engine didn't bring down this plane?

It did not.

There are 27 passengers boarding Midwest Express Airlines Flight 105, a cross

-country commuter flight from Milwaukee to Atlanta.

Many are employees of the manufacturer Kimberly Clark.

Midwest Express Airlines used the slogan, best care in the air, and they

really meant it because they had as close to an all-business aircraft as you

could have.

Treats for when we're airborne.

And they also baked fresh cookies on board and made the place smell amazing

when you walked on.

Mid-X 105 requesting IFR clearance to Atlanta.

Both pilots on this flight are captains with the airline.

They've already flown two flights together today.

Mid-X 105, clear to Atlanta.

Climb and maintain 5,000.

Climb and maintain 5,000. Mid-X 105.

For this leg of the flight, Danny Martin will be the acting captain and Bill

Weiss, the first officer.

We had two captains, and they literally could switch seats because of that. And

they did do that over the course of two days, flipping out who was the pilot

flying and who was the pilot not flying.

Our alternate would be DCA if we can't make Atlanta.

There are thunderstorms in Atlanta, so the crew has taken on additional fuel in

case they have to divert to another airport Take off weight is seven seven

one two two pounds set stab trim 2.2

Just past three in the afternoon the

pilots start the engines Starting number one

The DC-9 used for today's flight is powered by two Pratt & Whitney engines

mounted on the fuselage at the rear of the plane.

Because the engines were in the back, they were close together, which made for

a little bit easier flying in emergency situations.

Second of all, for the passengers, it's quieter.

Can I put that up here for you?

Thank you.

The plane is loaded and ready to depart on schedule.

Airspeed bug 133 set and 1.9 on the EPR.

Okay.

Ladies and gentlemen, we are number one for departure, so we should be airborne

within one minute.

Flight attendants, please be seated.

Midax 105, cleared for takeoff.

Midax 105.

Here we go.

It's pulling up.

This was a very normal morning for a very normal flight.

It was a gorgeous blue sky, September day.

There was a little bit of gusty winds, but it was nothing for Milwaukee.

Power normal.

100. The pilots need to reach a speed of 127 knots for takeoff.

V1.

Rotate.

At 3.21 in the afternoon, Flight 105 begins its journey.

The flight plan calls for the plane to climb to an altitude of 33,000 feet for

the two-hour trip to Atlanta.

Gear up.

Just 450 feet above the ground.

The hell was that?

The pilots lose power in one of the engines.

It's a critical situation because if you don't do something about it immediately,

you won't have time to do anything about it.

Altitude is your friend.

Mid-X-105.

Turn left heading 175.

What do we got here, Bill?

Here.

Mid-X-105, Roger.

We got an emergency here.

Phoenix 105, roger.

The plane begins rolling to the right and dropping.

Oh, crap!

The DC-9 is stalling.

Yeah, your head's-

Less than a minute after takeoff, Flight 105 is diving towards the ground at more

than 170 miles an hour.

The plane hits the ground less than 1

,700 feet from the runway.

A DC-9 has crashed southwest of runway 19R.

All 31 people on board have been killed.

The crash of Midwest Express Flight 105 is the third major accident in the

United States this year, and the eighth accident worldwide.

With more than 1,200 people killed, this is becoming one of the deadliest years

in the history of civil aviation.

Passengers are unnerved.

Everyone says, my God, another one. What do you say to the public about the

crashes that we've been having? Well, insofar as the National Transportation

Safety Board is concerned, we are sparing no effort to determine the cause

of each one of them, and we will make the necessary recommendations to keep a

repeat of any one of them from happening.

Can you show me where it started a bank?

Investigators from the NTSB begin by interviewing witnesses of Midwest

Express Flight 105's fatal dive.

They either came forward or we tracked them down. We ended up making out a

questionnaire with 40 or 50 questions on it to kind of cover all the bases.

Okay, got it.

While the witness accounts differ in detail, they all describe more or less

the same thing.

I saw a couple puffs of black smoke come out, and I just figured that they had

gunned the engines a little bit.

Instant later, the left wing rose up and it tipped over and went nose first into

the trees here.

Most of the witnesses agree that takeoff appeared normal until the airplane was

about 300 feet off the ground.

Almost all report smoke and fire coming from the right engine.

Several remember hearing loud bangs.

Most said that the plane then rolled abruptly into a steep right bank, went

into a nose-low spin.

and crashed the value of having so many

witness statements to look at they gave us a picture of what happened right at

the end a loss of control fire and of course the impact itself investigators

must now determine what could have caused such a catastrophic loss of

control so soon after takeoff

I'm guessing this was made by the right wing.

Yep, agreed.

Let's see what we can figure out.

After securing the crash site of Midwest Express Flight 105, investigators begin

mapping the wreckage field.

I think the whole length of the...

Impact site was about 295 feet not much more than two times the length of the

airplane So the airplane hit pretty steeply and didn't go very far Right

wing tip here and the horizontal stabilizer here

This is how the plane hit the ground a 90-degree right

roll and a right yaw Can you grab picture

Got it.

The wreckage path confirms eyewitness reports.

It very definitely indicates right up front that you've got a lost control.

It doesn't suggest why you had the lost control. It kind of opens the door to a

lot of different investigative avenues.

Yeah, it's fire damage for sure, but it looks like all the pieces are here.

Investigators study the plane's control surfaces.

starting with the ailerons.

The ailerons on each wing direct the plane's roll.

If an aileron malfunctions and gets stuck in one position, it could cause a

devastating lack of control.

Control tabs are in place. They seem to be intact.

We didn't find a problem with other components that are part of that system.

With the ailerons ruled out, investigators turn their attention to

the rudder system.

The dampers are still working.

The hinges are intact.

The fractures look like overload, not stress.

I don't think the rudder is our culprit.

We didn't find anything wrong with any of the control systems based on that

preliminary look-see.

Could the engines provide insight into the crash of Midwest Express Flight 105?

If the engines were, well, obviously banged up, they were fire damaged, and

they were along the wreckage path, which is always good, that means they didn't

fall off somewhere earlier.

There's a spacer missing here.

Most of the blades from this stage are gone.

The team finds damage inside the right engine that could not have been caused

on impact.

The missing engine pieces include a spacer and the compressor blades from

the ninth and tenth compressor stages.

They're found more than half a mile from the main wreckage site.

Finding any part of the airplane...

short of where the airplane crashed, was a suggestion of a malfunction or failure

that had occurred in flight and required further investigation.

The engine and the pieces are moved to a nearby warehouse where they can be more

closely examined.

In the end, we ended up finding about 90% by weight of the spacer, so that was

a big boon to the investigation.

Investigators study the recovered spacer.

to determine why it failed on Midwest Express Flight 105.

All on the pedestal here, it's definitely not overstress.

They discover a telltale crack on the spacer that appears to have spread over

time, what's called progressive cracking.

There may be stress marks. There may be cracks.

There may be things that suggest a direction of failure.

and a time frame where the fracture would have occurred.

It doesn't occur all at one time.

Can we determine how long that crack has been spreading for?

I'll see what I can do.

We became interested not just in detailing that, but trying to find

whether this was a fracture that could have been detected previously.

A closer look at the crack on the steel spacer reveals it's filled with nickel.

A lightweight nickel coating is applied to certain vital engine pieces like the

compressor spacers and blades to prevent corrosion.

The compressor was brought in to be refurbished in 1981.

Coating stripped, spacer examined, and here we go.

Replated.

That was four years ago.

The maintenance records tell investigators that work had been done on

the spacer that failed on Midwest Express Flight 105.

Nickel cadmium plating is the last step of the refurbishment process that would

have been done by the maintenance facility that was involved with

essentially the overhaul of that part.

Four years before the accident, the spacer was removed from the engine,

stripped of its coating, and examined for cracks.

The inspector reported no cracks and sent the part to be replated with

nickel.

Since nickel was found inside the crack, it must have been present when the

nickel was applied.

The DC-9 flew about 2,500 flights over the next four years with the damaged

spacer until it finally ruptured on flight 105.

It should have been caught during an overhaul, but it wasn't. And so the

crack kept growing and growing until it finally hit its failure point.

Investigators wonder what caused the crack to form on the spacer in the first

place.

These spacers have been used inside the JT-8D for more than 20 years.

This can't be the first one to break.

There are identical spacers in more than 14,000 JT-8D engines used on airplanes

around the world.

Is there any history of similar failures?

It turns out that Pratt & Co. were well aware of spacer failures that occurred

before. So we started to wonder, what have they done about it?

Seems to happen a lot.

While researching spacer issues in other aircraft, investigators uncover 45

similar failures prior to Midwest Express Flight 105's accident.

Well, when we heard that this part had failed 45 times before, we kind of

looked askance a bit at the manufacturer.

It's a big deal.

But of all the failures uncovered by the NTSB... Not a single loss of aircraft or

even a single injury.

Well, that says something.

In every previous incident, the plane landed safely.

So why did the rupture of a spacer on Flight 105 cause the plane to become

uncontrollable and crash?

Well, we knew that none of the previous failures had led to an accident.

So one of the questions was, what made this one different?

Let's see what got hit.

Investigators consider the possibility that pieces of the ruptured spacer on

flight 105 punctured the plane's fuselage and damaged vital control

systems, such as cables or hydraulic lines.

When a part like a spacer or a fan disc breaks under a lot of

stress, it's almost like a small explosion going off. You have fragments

of metal flying at very, very high speeds.

Okay, let's see 894.

They study the location and pattern of all the puncture marks found on the skin

of the plane to see if any were near vital control links.

No control cables, no hydraulics.

10-12.

No control cables, no hydraulics.

What about the pieces of the plane that we haven't been able to find?

There are many critical pieces of the plane that have not been recovered or

are too badly damaged to study.

Is it possible to figure out if they could have been hit by the engine

pieces?

We have what we need to run a trajectory analysis.

All right.

We're looking at could these parts that have left the engine

strike a control surface or the hydraulic system or anything else where

they could have done secondary damage that would have compounded the situation

the pilots were facing.

Knowing the plane's velocity, angle of ascent and the wind speed at the time

the engine failed, the team calculates the path of debris ejected from the

engine.

The smallest piece we found was about half an ounce.

The largest piece was just over a pound.

So everything moves away from the plane. Nothing hits it.

The trajectory analysis tells investigators that none of the ejected

engine pieces would have struck the airplane with enough force to cause

substantial damage to the control systems.

The analysis indicated that it was so unlikely that we considered it an

impossibility, that those parts leaving the engine in the direction they went

and the size.

that they had could have caused secondary damage that would cause lost

control.

If spacer fragments didn't hit vital components and cause the loss of

control, perhaps the initial explosive force opened the cowling or engine

cover, affecting the plane's aerodynamics.

If the cowling had been blown open, it would cause a lot of drag.

It might cause the airplane to roll.

It might cause the airplane to become uncontrollable.

There's a hole in the cowling. Looks to be about two square inches.

The team finds evidence that the cowling was pierced by fragments ejected from

the engine.

But all four latches, one, two, three, four, they're all latched.

Can't have opened in flight.

They found them all either latched or fully intact with no damage at all.

So that scenario kind of went out the window.

So the ruptured red engine didn't bring down this plane?

It did not.

Well, reasonably early in the investigation, the team figured out that

a spacer had failed and the engine had failed.

And now we had the rest of the accident to figure out.

Why would that cause a crash?

Investigators examine Midwest Express Flight 105's flight data recorder in

hopes of understanding how the loss of the right engine caused the death of 31

people.

Doesn't give us much, but it's all we've got to work with.

It was only recording four flight parameters.

It had airspeed, altitude, heading, and vertical acceleration.

Right here.

A sudden deceleration of 450 feet. This must be where the right engine failed.

Well, that's consistent with what the witnesses told us.

The engine failure we know occurred above the airport, so the airplane had

only been airborne for a few seconds.

But the heading doesn't change when the engine fails. They must be applying left

rudder to compensate.

Dead right engine, left rudder. That's the right move.

When the right engine fails, the remaining engine should force the plane

to the right. To counteract that, a pilot would apply left rudder to keep

the plane flying straight.

The data shows that's precisely what Captain Martin did in response to Flight

105's engine failure.

After a few seconds, they start this yacht to the right.

So, right rudder?

Four seconds after the failure of the right engine, the FDR data indicates

that the pilot moved the rudder from left to right. That forced the plane

into a sudden yaw to the right.

He's in a side slip.

He's deviating further and further to the right.

Then he starts to lose altitude.

Huh, 148 knots.

He's in a high-speed stall.

How'd they let that happen?

You went into a right skid and then followed by an abrupt right turn and

dive. That would be consistent with a stall, but the air speeds indicated were

high enough that you wouldn't normally have a stall at those speeds.

Clearly lost control of the plane. Five seconds later, they hit the ground.

The flight data tells investigators that it took only 10 seconds for the pilots

to lose control of the plane after the failure of the right engine.

I have never seen an accident sequence that brief before or since.

So... He responds correctly at first with left rudder and nose down pitch,

then he inexplicably switches to right rudder?

And that's what leads to the ah, the right roll, and the eventual stall.

Right rudder, it makes no sense.

It was clear from all the data that the pilot didn't understand what happened.

He responded inappropriately, and within seconds the airplane was lost and

everybody on board was killed in the accident.

Was there something about the DC-9's handling characteristics that led the

pilot to input the right rudder?

So it was relatively easy to control then?

Just with the control wheel?

Investigators interviewed DC-9 pilots to determine how the plane handles with

only one engine.

The DC-9 pilots that we talked to, it was overwhelmingly described as a very

easy, docile aircraft to handle in those situations.

Oh, this has been extremely helpful.

Thanks. Okay, bye.

They all say the same thing.

It's no big deal with one engine.

Because the DC-9's engines are mounted on the fuselage instead of the wings,

when one engine fails, it does not force the plane into a severe turn.

DC-9 is almost a centerline thrust aircraft, so if you lose one engine on

either side, there's not that much excess yawing or controllability

problems at all.

How easy is it to handle a DC-9 in the situation the pilots of Flight 105 found

themselves in?

It was valuable to do a...

simulation or test flight to get a better idea of exactly how the airplane

performed under the circumstances of the accident, but at a higher altitude.

Okay, when we get to 10,000 feet, I'm going to cut power to the right, leaving

the left at takeoff power.

Keep the flaps at 20, target speed is 170 knots.

They've tried to match the parameters of the airplane in terms of the flaps,

gear.

and uh airspeed and then proceeded to cut the engine power okay

cutting power

okay

thank you let's go around for another flight demonstrations showed that the

pilot didn't have to input rotor at all If he had only used the ailerons, he

could have safely fallen out of it.

Okay, this time, no reaction.

Let's see what the plane does.

With no input from the pilot after the loss of the right engine, the plane

rolls right and the nose drops.

Okay, you can recover.

Even with no immediate action, the pilot is easily able to recover the plane.

Okay.

I think we have what we need.

That may have actually been the best thing to do, was to do nothing, watch

the airplane's response, understand the airplane's response, and respond

appropriately to the nature of the emergency.

How was that?

If the DC-9 is so easy to control, why were two experienced pilots unable to

recover from the loss of their right engine?

Okay. We know they were both captains with the airline, but how much

experience did they actually have?

Investigators look into the backgrounds of the pilots of Midwest Express Flight

105 to understand why they didn't recover from a single engine failure.

Neither had a lot of turbojet experience.

One was upgraded to captain with 600 hours on the DC-9, the other with only

500.

Doesn't seem like a whole lot.

Advancement to captain happened within a year for both of these pilots at Midwest

Express Airlines at that time. And the reason why was it was a small airline.

It was growing quickly.

They were adding airplanes.

They consult large U.S. airlines to determine their qualifications for an

upgrade.

Okay, thanks.

That's the same for both airlines.

Ten years seniority and 7,500 hours as DC-9 first officer to be considered for

an upgrade.

These guys weren't anywhere near that.

While the pilots may have been experienced, neither had spent much time

piloting the DC-9 before they were promoted to captain.

The more experience someone has exercising command, the more likely they

are to exercise that command in situations that call for calm thinking,

for rational response, and so on.

So how much training did these guys get on engine failures?

To understand why the pilots of Flight 105 were not able to control their plane

following an engine failure, investigators look at their training

records. What was their experience?

What kind of pilots were they? How did they respond to unusual events? Did they

seem self-confident?

How well did they work together?

Okay, so...

The captains practiced 12 engine failures on takeoff in the simulator.

First officer, 15 sessions.

They discover that both pilots had been extensively trained to cope with engine

failure on takeoff.

Precisely the type of emergency they faced on flight 105.

But what exactly were they taught?

We looked very closely at the training these pilots received.

and interviewed the flight instructor who oversaw their training in the DC-9

from Interwest Express.

Tell me what you can remember about Captains Martin and Weiss.

They were both excellent pilots.

Trained to proficiency.

Lots of simulator training on engine failure.

You bet.

We ran them through all sorts of scenarios.

Let's focus on takeoff with simulated engine failure. What were they taught?

Pay attention to your yachts. Your best cue.

Yawing right.

Flying left rudder.

Control wheel as needed. Keep the plane level.

Nice and easy.

Good job.

Got it.

Captain Martin was trained to pay close attention to the plane's yawing motion

when identifying and correcting an engine failure.

Looks like they had excellent training.

Thanks for the insight.

We found the instructor to be very capable and very dedicated, and we found

the simulators to be okay for that time.

If Captain Martin was extensively trained to cope with engine failure on

takeoff, why didn't he do what he was trained to do?

Most of their training was for an engine failure at takeoff.

Pretty easy to identify which way you're yawing.

Investigators consider the timing of the engine failure.

If you're closer to the ground, there are much more visual cues available from

the ground to see what the airplane is encountering.

Not so easy when you're already in a steep climb. Right.

As Flight 105 was climbing, only blue sky would have been visible to the

pilots, making it more difficult to identify the direction of the yaw.

How was that?

With no outside visual cues, investigators believe Captain Martin's

initial response was a spontaneous reaction to the plane's changing motion

and not the result of an analysis of the situation.

But what triggered the incorrect rudder application a few seconds later that

resulted in a total loss of control?

All right, let's see what these guys are doing.

Clear for takeoff.

Med-X 105.

Here we go.

Investigators pin their hopes on the CVR to explain Captain Martin's actions.

Station. Mid-X-105.

Turn left heading 175.

What do we got here, Bill?

Here.

Mid-X-105, Roger.

We got an emergency here.

Mid-X-105, Roger.

Oh, crap.

Head down. Head down.

I've never heard an accident unfold so quickly.

Holy smokes.

And I remember just being amazed how quickly the accident took place.

It was like they were flying.

You hear the engine fail, and the next thing you know, the airplane's lost.

Okay. So what did we hear?

It sure happened fast.

Yep. Fifteen seconds from the sound of the engine exploding to the end of the

recording.

Astonishing.

What else?

What was that? What have we got? The captain seemed uncertain.

Agreed.

What have we got here, Bill?

Well, it was clear from the cockpit voice recorder that the pilot didn't

recognize what had happened. He articulates that. What was that?

What have we got here, Bill?

What was that?

The sound of the engine, the sensation of the right yaw, and the instruments

should have told Captain Martin that he was dealing with a failure of his right

engine.

Anything else on that tape?

Not a lot of troubleshooting or communication between them, really.

Actually, none.

What do we got here, Bill?

Yeah.

Mid-X-105, Roger.

We got an emergency here.

Investigators hear Captain Martin asking his first officer for assistance in

assessing the situation, but First Officer Weiss never replies.

That was troubling to me because I would have expected that in an emergency

situation, and especially when asked for help, that the other crew member would

have helped if he could.

So that prompted some further investigation.

Okay, let's start with that.

Why was the captain confused about what had happened?

Should have been obvious.

4,600 flight hours, but he's never experienced an actual engine failure.

NTSB investigators looking into the crash of Midwest Express Flight 105 have

learned that Captain Martin's only experience with a single engine failure

was in a simulator.

Yaw and deceleration don't feel the same, plus all the sounds are different.

It's not the same as a real thing.

It was a catastrophic engine failure where they had thrust and then they

didn't. It was immediate.

I don't think they were trained in that kind of scenario.

What have we got here, Bill?

The limitations of the simulator may explain Captain Martin's initial

confusion, but it does not explain the First Officer's unusual silence.

Let's have another listen.

They returned to the cockpit voice recording to better understand how First

Officer Weiss reacted to the emergency.

That's two.

That's three.

He asked for help three times, but the FO says absolutely nothing to him the

entire time. Not even, I'm not sure, or I don't know.

Not a word.

In this situation, seconds were critical.

Had he said right away, I don't know, then the captain would immediately have

tried to rely on his own understanding of the situation, rather than expect the

first officer to come up with an explanation of what he had.

What could have caused the first officer's silence?

Hang on.

What altitude were they at when the engine failed?

450 feet.

This could be it, then.

Pilots say they were taught not to make call-outs or even verbalise the nature

of an emergency after 100 knots before reaching 800 feet.

Let me see.

Not even to identify a failed engine.

No communication at all until 800 feet.

I couldn't imagine a management team at Midwest Express thinking this silent

cockpit business made any sense.

It just flies in the face of communication, which is kind of the

fabric that holds a safe flight together.

So it just blew my mind, frankly.

How was that?

Investigators think First Officer Weiss may have been following a Midwest

Express unwritten rule of not discussing emergencies until the plane reaches 800

feet.

It's important during an emergency for people to talk to each other because

they both may not be looking at the same instrument. They both may not be looking

at the same thing outside. You have to talk. You have to talk.

One last thing troubles investigators.

In spite of his first officer's silence, Captain Martin could have easily

identified which engine had failed simply by looking at his instruments.

Both pilots had visual information from the engine flight instruments, like turn

indicator and so on, that would have told them what was going on in terms of

the loss of thrust.

I don't think either of them were scanning the instruments.

Agreed. They never saw which engine had failed, the direction of the awe, or the

severity of the roll.

The NTSB concludes that the pilot's failure to monitor their instruments

contributed to Captain Martin's incorrect rudder application.

With the lack of visual cues, the only sure way they would have had to

recognize which engine failed would be to look at the instruments.

The pilot's lack of awareness ultimately led to the stall that brought down

Flight 105.

Up until here, he should be fine.

But then there's no communication and no instrument scanning.

And this, this is what dooms them.

He puts in a right rudder, which sends them into a side slip, an extreme right

roll, and a stall.

If they had talked to each other, scanned their instruments, you don't get

this accident.

The reality of the situation is the mismanagement inside the cockpit.

caused the accident, not the engine failure.

The NTSB's final report makes several key recommendations to prevent this type

of accident from happening again.

They recommend a directive requiring airlines to replace the existing spacers

with a new type of spacer which is less likely to fail.

They also recommend that airlines are advised to teach their pilots to

communicate during onboard emergencies.

It should never be overstated the importance of crew coordination.

Every crew member has something valuable to add. Anything that is an abnormality

needs to be discussed at the proper time.

They also suggest that airlines review their simulator training to ensure

pilots are taught to use their instruments to assess the nature of

engine failures.

I think this accident still has an impact today because

It's such a clear illustration of what happens when you don't do what

you should do in an emergency.

Anytime there's an emergency, you should have the exact same procedures and you

should run right to them.

Midwest Express Airlines continued to operate until 2009.

It merged with Frontier Airlines in 2010.