Air Emergency (2003–…): Season 15, Episode 9 - Steep Impact - full transcript
When the crash of a commuter plane in Georgia kills a NASA astronaut, a four-term U.S. senator, and 21 others, the disaster grabs headlines around the world. Investigators soon come up with a theory to explain the crash. There's just one problem-the manufacturer says it's impossible. Gambling on a hunch, one investigator puts his reputation on the line to uncover the startling truth about what brought down Atlantic Southeast Airlines Flight 2311.
Grab your ankles, head down
A plane crash in Georgia leaves 23 dead
including a NASA astronaut and a US senator.
Any time there is a high profile case
there is work-stress and pressure on you.
In search of the cause...
We've got to see what happens in the air
...investigator Tom Haueter
gambles on a risky hunch
When I first proposed to do a flight test
it was not well received
He puts his reputation
and the life of a skilled test pilot on the line.
This was in my opinion
a very dangerous manoeuvre.
What if it crashes?
What if we loose the airplane?
The risks of course is part of the game.
Air Crash Investigation
Season 15 Edition 09
This is a true story
It is based on official reports
and eyewitness accounts
STEEP IMPACT
Atlantic South-East Airlines Flight 2311
cruises at 15,000 ft.
April 1991
Sorry, I don't want to get my hopes up
Last year still hurts
At the controls is Captain Mark Friedline
The 34 is an experienced pilot
with almost 12,000 flight hours
We got storm-clouds ahead of us
Yep, what are want to do?
First Officer Hank Johnston is 36
He's been flying with Atlantic SE Airlines
for nearly 3 years.
We are going to go around it
Tell everybody we will go for a ride
Center, Atlantic SE Airlines 2311
we'd like to do our request to
do a deviation for some weather
This was a normal day in the life of the crew
Nothing on toward that happened
I doubt they were expecting any difficulties
with the flight.
ASE 2311, Roger that
the weather deviation approved
to your discretions.
Maintain correct altitude
OKAY, thanks, Center, 2311
Let's go 20° to the right
Today's flight is a short commuter route
from Atlanta, Georgia to the city of Brunswick
on the Atlantic coast.
This is captain Friedline on the flight deck
We'll get a bit of weather ahead of us
but will go around it
If you bounce there is nothing to worry about
We do ask you to return to your seats
and make sure your safety belts are fastened.
The Embraer 120 banks gently
as the pilots deviate around the storm.
The Embraer B120 is designed for
commuter type operations
It takes approx. 30 people on short haul flights
between city centres.
The commuter airplane
is powered by 2 turboprop engines.
Turbo prop differs from jet engines because
instead of using the exhaust to power the aircraft
they use a propeller which is more fuel efficient
and is more suited for this smaller aircraft.
Today, there are 20 passengers aboard
including NASA astronaut Sonny Carter
As part of the space shuttle Discovery's crew
Sonny Carter orbited the earth for a 120 hours
circling the planet 79 times.
Fredrich Gregory surfed with Carter
on that mission.
Sonny was not only an astronaut,
he was a renaissance man
He was an engineer...
...he was a medical doctor...
...he was a NAVY pilot.
At a very young age he had accomplished things
that even one of which
would have been an outstanding achievement
for just any other person.
Also on the flight is another high profile figure:
former Senator John Tower.
Tower served 4 terms in the US senate
a leading republican,
he was an advisor to presidents Ronald Reagan
and George H.W. Bush
And we are still running 2 minutes late
Couldn't be helped
Before take off, Flight 2311 was delayed,
due to a mechanical problem.
The crew was forced to switch planes
It was a last minute change in the aircraft
which is not extraordinary
and it would not have thrown the crew.
The new plane is running smoothly
as it leaves the storm clouds behind
and nears its destination.
Glynco Airport is a former airbase
with just one runway.
It is used by private planes and small commuter
airlines flights to the Georgia coast
The flight attendant prepares the cabin
for landing.
The runway is in sight.
The crew is just 5 minutes from touching down.
ASA 2311, cleared direct to F1
Glynco, report the airport in sight
expect the visual
We do have it in sight
00:06:08,721 --> 00:06:11,554
Slowing for approach speed.
The aircraft was normal,
there was nothing unexpected
Gear down
Gear down
3 green
Then, the captain notices an unusual sound
It is weird
The other one seems to be spinning faster.
The left is...
let's just pull it a bit more
Bringing power down to the left
Captain Friedline tries to compensate
for the plane's unexplained pull to the left.
Flight 2311 is less than a 1000 ft from the ground
and the plane
is getting more and more difficult to control.
What is going on?
The crew were apparently caught completely
by surprise by something
The plane is rolling to the left
and the crew doesn't know why.
The crew instinctively applied opposite aileron
moving the stick to the right
to try and prevent the aircraft from rolling
Come on, guys,
get us out of this
I think they would have been preoccupied
with trying to stop the roll
Maybe to the extend of
not knowing just how much trouble they were in
Do what I do:
Grab your ankles, head down
Captain Friedline fights desperately
to save his plane.
It's no use
The aircraft crashed in the middle of woods
which would have made the job of the first
responders and fighting teams quite difficult
Rescues make their way through the dense brush
but when they reach the crash site
they find no one alive to rescue.
All 20 passengers and 3 crew-members are dead,
killed at the moment of impact.
This was a catastrophic crash
There was no chance for survival.
With two high profile passengers on board
the crash of Flight 2311 makes headlines
around the world.
In this particular accident Senator Tower on board
was 4 terms senator was killed
which raised the profile of
the accident a great deal
both from the investigation point of view
and the public and the media.
I was in Spain when the accident occurred
When I got on the plane to head back
the pilot told me that one of
the astronauts had been killed
and that his name was Carter
I was in shock,
because Sonny and I had had
an amazing relationship for years
We were like brothers.
Wreckages still smouldering
when Jim Ritter arrives at the crash site
Though an experienced investigator
with the NTSB
Ritter is struck by what he sees.
My God!
There was a lot of fire damage in the wreckage
and the airplane was totally destroyed
Ritter realizes he is facing
one of the biggest challenges of his career.
Any time there is a high profile case,
we are a little more nervous
There is more stress and pressure on you
to come up with the cause for the accident.
The pressure is on
to figure out what happened aboard Flight 2311
In Glynn County, Georgia
investigators search for answers at the crash site
Several people at a nearby trailer-park
saw the plane go down.
They come right at the top of the house
and they got real loud
It was coming right over these trees here
and then extra loud
The eyewitnesses saw an extreme left roll angle
from the airplane
so we knew that it was a catastrophic failure
a very abrupt failure that would have been
difficult for the flight crew to overcome.
OKAY, let's start here
and work backwards to first impact.
I want a record of everything.
You need to look at the crash site
to collect the physical evidence
that is the most important aspect
of any air-crash investigation.
They look at the wreckage to try and evaluate
the sequence of events, leading to the crash
They look for the 2 wing-tips,
the tail, the tip of the nose
A survey of the crash site soon confirms
those key pieces are among the wreckage.
It tells us that there were no in-flight break up
and that the airplane stayed intact
until it collided with the ground
Get a shot of this wing for me, please
For expert investigators,
the size and shape of the impact zone
is also a clue.
If an aircraft flies through trees,
a long cut through the trees
the investigators get an idea that the aircraft
hit the ground approx. flat in a shallow descend
In this case,
the wreckage was very concentrated
which indicated that
the aircraft hit the ground at a very steep angle.
We need to get some measurements on these trees
Investigators hope the broken trees
will shed light on witness statements
about the way the plane was flying.
It was like it made a wild right turn,
but it was unusual, because it turned out
the nose went into the ground
I was right there when the engines touched
I heard the explosion
I saw the fire and the smoke
When the plane crashed,
it sheered the tops off of many of the trees
in the woods there
one of the things that we did was,
we measured the heights of the trees
where the tops were sheered off
to give us a fairly good idea of the roll angle
which was a large left roll
In the plane you felt light in your seat
you would have turned left at almost 90°
That would have been a traumatic experience
What could make it roll so far over?
When the airplane rolled to the left,
it could be only due to 2 things
Perhaps the pilot wanted to roll to the left
or there was a malfunction
that the pilots couldn't counteract
What do you got for me?
Ritter examines airport flight records.
He is looking for anything
that might have triggered the deadly roll
Clear skies on approach
almost zero traffic.
There were no other airplanes in the area to avoid
It was a nice clear day
so we didn't have an explanation for
why the airplane rolled so violently to the left.
Ritter won't be getting an explanation
for many on board flight recorders
At the time of the crash,
commuter-planes aren't required to carry them
Investigators will have to solve the mystery
without one of their most useful tools
Not having those
made the physical evidence all themore important
What have you got for me?
This is the third one
The engines have been badly damaged in the crash
the propeller blades have been torn off
Did Flight 2311 suffer
some kind of engine failure?
It is too soon to say
One thing Ritter does know is
that the pressure he is feeling from the media
is not about to let up.
The deaths of senator Tower
and astronaut Sonny Carter
are sure to keep the investigation
in the spotlight
Ritter is determined
not to let the pressure get to him.
Sometimes we feel pressure
to do investigation quickly,
but for the most part,
it is more important to get it right
I'd rather take the time
and have a good analysis of the evidence
before they come out with the probable cause.
You won't find their answers here
Let's get what we can
back to the hangar
Wings, tail, engines, instruments
The question now:
Can he find enough evidence to ever solve
the mystery of Flight 2311?
Investigators begin the painstaking task
of sorting through the wreckage.
They are searching for any evidence
that might hint at why the Embraer 120
rolled sharply to the left and crashed,
killing every one on board.
Al-right guys, let's start with the ailerons
They wonder if one of the flight control surfaces
was malfunctioning.
The control surfaces
are the movable flaps
typically at the back of
the wings, the tail and the fin
which allow the pilot to roll the aircraft,
to pitch the aircraft and to yaw the aircraft
the three motions an aircraft can do
They need to examine
every component of the flight control system
What you are looking for is
☮Continuity of the controls
☮Are the hinges all intact
☮All the actuators are in their proper position
☮Were the control surfaces themselves
in a reasonable position at the time of impact?
The ailerons look good
When we examined the control surfaces,
we didn't find anything unusual at all
Everything checked out normally.
So, the control-systems were all working
Why would a plane do this?
Left bank, almost 90°
and almost straight down.
Maybe the engine
Without the black boxes,
it is basically a process of elimination
We analyse all of the physical evidence
and come up with the most compelling scenario
that matches that evidence
OKAY, let's see what we can find in here
If one of the 2 turbo prop engines failed
it might explain
why the plane went into such a steep roll
We wanted to look for
any indications of an engine problem
so we retrieved both the left and the
right engine and their propeller systems
and took those back to the laboratory
for further examination.
Investigators soon find some tell tale evidence
leaves and branches inside the engines.
When we find vegetation in the engines
that tells us that the engine is operating
because it is sucking in air
and it is pulling in the leaves and the vegetation
That is an indication that it is making power
They must have spinning to the very end,
no doubt about it.
We confirmed that both engines were operating,
right up until the moment of impact.
With engine failure ruled out,
Ritter turns his attention to
the other main part of the propulsion system
Let's take a look at these propellers.
Tom Haueter is an NTSB investigator
When we started doing tests
in the propeller system,
we didn't know
where it is going to lead this
But it was something we had to eliminate,
and nothing else.
Deep inside the propeller unit
investigators uncover an important clue
Aha, there we are!
Take a look
There is a small mark where two parts of
the propeller mechanism slammed together on impact
The witness mark might be enough to tell
investigators how the propellers were operating
You can literally match up the scratch marks
between both pieces
and you'll know what the angle
of the propeller-blade was from that measurement
Mark this one 22°
The Embraer 120 has a constant speed propeller
The blades spin at a steady rate in flight
When the pilots need more power,
the blades twist
changing their angle to take
a bigger byte out of the air
and provide more thrust.
And in flight,
it acts like the automatic transmission in a car
It is as if it is changing gears
to match the engine load, the speed
whether the plane is climbing or descending
and it does so constantly and automatically
without intervention by the crew.
There are witness marks inside both propellers.
Now, this one looks like 3°
The marks tell Ritter the angle of the blades
when the plane slammed into the ground
We immediately noticed the difference between
some of the blade angle measurements
for the left engine
versus the right engine
These guys had a big problem
with their left propeller.
The blades were almost flat
The left side propeller blades
are at a dangerously low angle
one that is never used during flight.
At 3° the blades are so flat,
they would act like a wall
blocking the flow of air the plane needs
to maintain lift
There is certain speeds,
certain regimes of flight,
if the propeller goes flat enough
you have situations you cannot control the plane
Investigators study the mechanism
used to control the left propeller.
Hey, look at this
They make a disturbing discovery
It is completely worn down
The teeth on a key piece of the gear-mechanism,
known as the quill are almost entirely worn away
Investigators may finally have the lead
they've been looking for
This is what is supposed to look like
With its teeth worn away,
the quill can't lock onto the gear system
that controls the angle of the propeller blades
The discovery might explain
why the propeller blade slipped
to such a dangerously low angle.
Once we noticed
that the quill teeth were severely worn
we started theorizing
what would happen in that situation
It was pretty clear
that the control of the blade-angles could be lost
That was a Eureka moment for us
because now we had a serious malfunction
that we could examine.
This could definitively be it
Ritter digs into manufacturing's reports
trying to find out how such a vital part failed.
But what he finds only adds to the mystery
The quill teeth are made
of an extremely durable case hardened metal
They are definitively made to last
It is hard to imagine what could have caused
such rugged teeth to wear down so badly.
Something didn't work as planned.
Let's find out everything we can
about every one of these pieces
Ritter is certain he has found
the critical clue:
worn down teeth on the quill
that could have allowed the propeller blades
to slip to a dangerous angle.
but he soon learns
there is a big problem with his theory
The manufacturer says it is impossible.
Engineers at Hamilton Standard
included a fail safe feature
when they designed the propeller
It should be impossible for the blades
to go flat during flight.
Manufacturers have to demonstrate
that their systems are fail safe
All the test and research have been done
before this said:
even if you disconnect,
that will not result into an accident
If there is ever a problem
with the mechanism controlling the angle,
the blades are designed to move on their own
to the feathered position.
The feathered propeller blade,
the leading edge of the blade
is directly into the wind
that is minimal drag,
no thrust, but very little drag.
A feathered propeller
can't endanger the safety of the flight
If this rod disengaged because of any cause
the propeller should go to feather
which would result in the loss of the engine
and the crew would be able to cope with that
and land on one engine
as they are trained to do
The runway is in sight
This got to be it
It has to be connected somehow
But Ritter isn't convinced
by the manufacturer's assurances.
His gut tells him the worn quill did allow
the propeller blades to move to a dangerous angle.
But without flight data
his investigation has hit a wall
He has no way to prove
the quill brought down Flight 2311
Tom, welcome to the team
Tom Haueter:
No problem, Jim I'm glad to help
With the investigation stalled
Tom Haueter joins Jim Ritter to hunt for answers
in the crash of Atlantic SE Flight 2311
I was a little nervous about this when in that
when I first jumped in to it,
I didn't really know what was going on
other than it had a standstill
and they were looking for me to get it moving
What do we have?
I think everything we need to know
is right here on this table
We have a part of the propeller control assembly
and the teeth on the gear are essentially gone
That is very unusual
Could that be a part of the accident,
we didn't know
Haueter and Ritter
study the design of the mechanism
Tom, look at this
We don't see anything obviously wrong
with the crew
We don't see anything wrong
with the structure
We don't see anything wrong
with the engines and the flight control system
but we do have
a severely worn transfer to the quill
Is it possible that this could have resulted in
the loss of control?
They discover that shortly before the accident
Hamilton Standard started using a harder,
more abrasive coating on a key part,
known as the transfer tube
Its grooves mesh like clockwork
into the teeth of the quill
The change had an unexpected consequence
It turned into a giant file
The splines on the transfer tube were
much harder and rougher than the quill teeth
and it was almost like sandpaper
The tube was wearing down the teeth on the quill
The discovery explains
the worn teeth on the quill.
but it still doesn't explain the crash.
The design of the propeller should ensure
it snaps to a safe position,
even with the worn part.
The propeller manufacturer believed
that they had a fail safe condition
so that even if they have this problem
the blades would be slowly driven
to the feather position.
In spite of what all their data says
the investigators want to see for themselves.
They set up a test
at the manufacturer's facility
We had an engine and a propeller combination
mounted in a test cell.
In order to determine that something is fail safe
the engineers use a combination of
mathematical analysis, a very structured analysis
testing in a laboratory
and by these methods,
HS convinced themselves that this propeller would
fall in a safe direction i.e. towards feather
Jim Ritter's doubts about the fail safe design
may be confirmed in just a few seconds
The technician flips a switch to free
the transfer tube from the teeth on the quill
but as the test unfolds
instead of moving to flat,
the propeller blades move to the safe position
The fail-safe system performs exactly
as it was designed to
It suddenly seems that investigators
are on the wrong path altogether
When we tested the quill, with the worn teeth
at the manufacturer's facility
We found that the propeller-blades
went to feather
So, at that point, we are basically stunned
Can you play that again?
The test results leave
Ritter and Haueter wondering
Could there be some other factor
they've overlooked?
Tom Haueter:
Hold on, that's it
It is bolted to the ground
The question I raise
well, in flight the aircraft is in turbulence
is bouncing around, there is different vibrations
With the aircraft on the ground
or the engine is mounted solidly to the ground
the airplane doesn't behave the same,
there is different vibration modes
Is it possible that could change
the outcome of the analysis?
I think we have to see it in the air
One of the problems of doing testing
in a laboratory environment
is that you can't always
anticipate what will happen
when you go out
into the real world
Imagine for example an orchestra
that rehearses in an acoustically
perfectly concert-hall
and then performs outside
with random noises, without the sound control
You could see how one might miss
important factors about the sound
We've got to see what happens in the air
I said: Well, the only way to really know
is to do a flight test and find out
Let's do something that actually determines
that this is the case
We are at a point in the investigation
we need to start eliminating things
The investigation into the crash of Flight 2311
moves to Embraer headquarters in Brazil
We really wanted to look at
what would happen in an actual flight,
with the same malfunction
Tom Haueter meets with representatives
from Embraer and propeller manufacturer HS
The team immediately starts preparing
for the test flight
Thank you for doing this
When I first proposed doing the flight test,
it was not well received by almost everybody
People saw no need
why we were spending the time, the money
why we were doing this
But I pushed hard
we need to really prove this
Embraer's chief test pilot, Gilberto Schittini
has agreed to put the investigator's theory
to the test
The risk is part of the game
You have to reduce the risk as much as possible,
but you cannot avoid this all
You just have to accept it
The mindset of a test-pilot
is not of heroic bravery
but certainly one wants to be unflappable in
the face of danger and unusual situations
because you are not very productive if you are
terrified by the goings on in the cockpit
Schittini will fly an Embraer 120
that has been modified to recreate the failure
So, we modified the quill
The teeth have been worn down,
just like Flight 2311
A worn quill will be placed
inside the propeller unit.
This will potentially very high risk
because once we disconnected the transfer tube
in flight
the pilots will have no way
to control the propeller
We put a pitch-lock here,
it won't go past 23°
A mechanical lock has also been added
to stop the propeller from going flatter than 22°
It would be too dangerous in the test-flight
to have blades that go all the way to flat pitch
You'd loose control of the airplane
If the blades move to 22°
then we know they would have gone flat
The blades don't need to go completely flat
to prove that Haueter is on the right track
It was a high risk test,
we had to take extra precautions
in order not to repeat the accident.
Real time data will tell the team on the ground
whether the blades will go safely towards feather
or moving dangerously flat
Everything says that if you disconnect,
the normal frictional drag within the system..
...will cause the components to rotate
towards the feather position
But will that really happen in flight
the same way?
They are about to find out
We are comfortable
that this test can be safely accomplished
That is the difference between
believing and knowing
As the test plane climbs
Schittini takes it over an unpopulated area
just in case
There was a potential
for a severe controllability problem
such that the pilots might have to abandon
the aircraft and parachute to the ground
The propellers have been set
for a normal angle for flight...
...around 30°
EMB120, do you copy
There was a lot in terms of facts that..
What if this airplane crashes?
What if we loose the airplane?
I am the investigator in charge,
I am the one who is running this test
This could be all my responsibility
Copy we are ready to disengage the prop
They would pull the lever that would disconnect
the Propeller Control System from the propeller
and see what happens
The most dangerous part of the test flight
is now under way
Let's do what we have to do
and get this airplane on the ground
This was in my opinion a very dangerous manoeuvre,
because
We didn't know what the outcome was going to be
and they already knew
the aircraft was uncontrallable
under certain circumstances
propeller blades angle
no problems
No control issues
We were thinking about
the test to do and nothing else
You have to focus in the job you have to do
Once you focus,
forget about everything else.
As soon as the propeller
quill is disengaged
the blades begin doing what
they were designed to do
Moving towards the feathered position
As they start drifting towards feather
we thought: OKAY
this test is just going to prove
what everybody said it would do
You could see on their faces
and they didn't say anything
but you could hear the relief
The propeller blades keep moving
towards the fail safe position
For Tom Haueter,
Brazil is a long way to come for failure
It seems like the test that he hoped
would solve the mystery of flight 2311
is going to leave the investigation
back at square 1
With the cause of the Atlantic SE crash
still unknown
and the entire investigation on the line
the Brazilian test flight takes a dramatic turn
We saw that the RPM was increasing
Slowly, but it was increasing
so I knew that the propeller
was going in the wrong direction.
As the flight continues, the blades begin
drifting in the opposite direction
towards the dangerous flat position.
Here we go
The aerodynamics, the vibration,
the actual in flight loads on the propeller
are just something you can't predict in a test
It is something that actually had be flown
Schittini remains smooth and precise
on the controls
We started feeling a rolling moment
to the left
The blades go as flat
as this test will allow: 22°
For Haueter, it is a victory
The risky test-flight has paid off
Their theory about the crash
is back on solid ground
The Hamilton Standard and FAA people
just went white,
I mean, they were in a shock
You could tell that everything they believed
it just have been thrown out the window
That is the slightest they are going to get
Do you have what you need?
Affirmative, we have everything we need
OKAY, we are going back
Once the test-flight is finished
it is finished
Don't stay around looking for trouble
Go back home
Seen the data right then,
it took a load of
I said Wow,
We now now what happened
It was obvious
The team has just one more question to answer
before they can explain the crash
What happens when blades go completely flat?
In de flight test,
they were only able to go to
a blade angle of 22° for safety-reasons
but during my flight simulations,
I went down to flight angles as low as 3°
In a flight simulator,
Jim Ritter recreates the fatal propeller
malfunction aboard ASA 2311
As the blade angle get to very low values
the airplane was uncontrollable.
The simulation allows him to experience
what the test pilots couldn't
The flight test proved:
we could have a blade go to flat pitch
The simulator proved:
You loose control evidently
That was helpful
You could say that the simulator testing
was the final piece of the puzzle
They finally understand the full story
behind the deadly crash
When the flight-crew began
preparations for landing
the teeth on the quill were worn down
still operational
Preparing for landing put renewed pressure
on the already worn teeth in the quill
They could no longer hold
Once the teeth gave way,
the blades were free to
drift to a different angle
As the turbulence of flight and mechanical
vibrations from the engine shook the propeller
the blades moved to a dangerously flat position
The fail safe....failed
The propeller design couldn't hold the blades
at a safe angle
and the plane
became less and less controllable
For a while you could fight this,
you can control movements
with the wheels, the rudder
But unfortunately they were in a situation
where it didn't matter
even if you were the best pilot on the planet
you were not able to control that airplane
It's going to roll over,
it is going to dive towards the ground
Ironically, human error is almost always
underneath the cause of an accident
even it wasn't the pilot or a mechanic
In this particular case,
an engineering change was made
which, who had the intention to
didn't work out
The worn quill teeth was a time bomb
waiting to go off.
There was nothing he could have done
to safe that plane
What I really felt was:
OKAY, now we now it,
we'd better tell the rest of the world.
and get operators
to start changing their equipment right now
The fail safe system doesn't work
After the accident, additional safe guards
were added to prevent this type of failure
The change that affected not only the Embraer,
but several other turbo prop aircraft as well
The FAA also demanded
more frequent inspections of propeller-quills
The most important thing
we've learned from the accident of Brunswick...
...is assumptions:
You cannot rely on assumptions
It was assumed
the ground test was as good as flight test
That one assumption unfortunately
turned out to be fatal for some people
Narrator
Jonathan Aris
Subtitles
Rein Croonen
A plane crash in Georgia leaves 23 dead
including a NASA astronaut and a US senator.
Any time there is a high profile case
there is work-stress and pressure on you.
In search of the cause...
We've got to see what happens in the air
...investigator Tom Haueter
gambles on a risky hunch
When I first proposed to do a flight test
it was not well received
He puts his reputation
and the life of a skilled test pilot on the line.
This was in my opinion
a very dangerous manoeuvre.
What if it crashes?
What if we loose the airplane?
The risks of course is part of the game.
Air Crash Investigation
Season 15 Edition 09
This is a true story
It is based on official reports
and eyewitness accounts
STEEP IMPACT
Atlantic South-East Airlines Flight 2311
cruises at 15,000 ft.
April 1991
Sorry, I don't want to get my hopes up
Last year still hurts
At the controls is Captain Mark Friedline
The 34 is an experienced pilot
with almost 12,000 flight hours
We got storm-clouds ahead of us
Yep, what are want to do?
First Officer Hank Johnston is 36
He's been flying with Atlantic SE Airlines
for nearly 3 years.
We are going to go around it
Tell everybody we will go for a ride
Center, Atlantic SE Airlines 2311
we'd like to do our request to
do a deviation for some weather
This was a normal day in the life of the crew
Nothing on toward that happened
I doubt they were expecting any difficulties
with the flight.
ASE 2311, Roger that
the weather deviation approved
to your discretions.
Maintain correct altitude
OKAY, thanks, Center, 2311
Let's go 20° to the right
Today's flight is a short commuter route
from Atlanta, Georgia to the city of Brunswick
on the Atlantic coast.
This is captain Friedline on the flight deck
We'll get a bit of weather ahead of us
but will go around it
If you bounce there is nothing to worry about
We do ask you to return to your seats
and make sure your safety belts are fastened.
The Embraer 120 banks gently
as the pilots deviate around the storm.
The Embraer B120 is designed for
commuter type operations
It takes approx. 30 people on short haul flights
between city centres.
The commuter airplane
is powered by 2 turboprop engines.
Turbo prop differs from jet engines because
instead of using the exhaust to power the aircraft
they use a propeller which is more fuel efficient
and is more suited for this smaller aircraft.
Today, there are 20 passengers aboard
including NASA astronaut Sonny Carter
As part of the space shuttle Discovery's crew
Sonny Carter orbited the earth for a 120 hours
circling the planet 79 times.
Fredrich Gregory surfed with Carter
on that mission.
Sonny was not only an astronaut,
he was a renaissance man
He was an engineer...
...he was a medical doctor...
...he was a NAVY pilot.
At a very young age he had accomplished things
that even one of which
would have been an outstanding achievement
for just any other person.
Also on the flight is another high profile figure:
former Senator John Tower.
Tower served 4 terms in the US senate
a leading republican,
he was an advisor to presidents Ronald Reagan
and George H.W. Bush
And we are still running 2 minutes late
Couldn't be helped
Before take off, Flight 2311 was delayed,
due to a mechanical problem.
The crew was forced to switch planes
It was a last minute change in the aircraft
which is not extraordinary
and it would not have thrown the crew.
The new plane is running smoothly
as it leaves the storm clouds behind
and nears its destination.
Glynco Airport is a former airbase
with just one runway.
It is used by private planes and small commuter
airlines flights to the Georgia coast
The flight attendant prepares the cabin
for landing.
The runway is in sight.
The crew is just 5 minutes from touching down.
ASA 2311, cleared direct to F1
Glynco, report the airport in sight
expect the visual
We do have it in sight
00:06:08,721 --> 00:06:11,554
Slowing for approach speed.
The aircraft was normal,
there was nothing unexpected
Gear down
Gear down
3 green
Then, the captain notices an unusual sound
It is weird
The other one seems to be spinning faster.
The left is...
let's just pull it a bit more
Bringing power down to the left
Captain Friedline tries to compensate
for the plane's unexplained pull to the left.
Flight 2311 is less than a 1000 ft from the ground
and the plane
is getting more and more difficult to control.
What is going on?
The crew were apparently caught completely
by surprise by something
The plane is rolling to the left
and the crew doesn't know why.
The crew instinctively applied opposite aileron
moving the stick to the right
to try and prevent the aircraft from rolling
Come on, guys,
get us out of this
I think they would have been preoccupied
with trying to stop the roll
Maybe to the extend of
not knowing just how much trouble they were in
Do what I do:
Grab your ankles, head down
Captain Friedline fights desperately
to save his plane.
It's no use
The aircraft crashed in the middle of woods
which would have made the job of the first
responders and fighting teams quite difficult
Rescues make their way through the dense brush
but when they reach the crash site
they find no one alive to rescue.
All 20 passengers and 3 crew-members are dead,
killed at the moment of impact.
This was a catastrophic crash
There was no chance for survival.
With two high profile passengers on board
the crash of Flight 2311 makes headlines
around the world.
In this particular accident Senator Tower on board
was 4 terms senator was killed
which raised the profile of
the accident a great deal
both from the investigation point of view
and the public and the media.
I was in Spain when the accident occurred
When I got on the plane to head back
the pilot told me that one of
the astronauts had been killed
and that his name was Carter
I was in shock,
because Sonny and I had had
an amazing relationship for years
We were like brothers.
Wreckages still smouldering
when Jim Ritter arrives at the crash site
Though an experienced investigator
with the NTSB
Ritter is struck by what he sees.
My God!
There was a lot of fire damage in the wreckage
and the airplane was totally destroyed
Ritter realizes he is facing
one of the biggest challenges of his career.
Any time there is a high profile case,
we are a little more nervous
There is more stress and pressure on you
to come up with the cause for the accident.
The pressure is on
to figure out what happened aboard Flight 2311
In Glynn County, Georgia
investigators search for answers at the crash site
Several people at a nearby trailer-park
saw the plane go down.
They come right at the top of the house
and they got real loud
It was coming right over these trees here
and then extra loud
The eyewitnesses saw an extreme left roll angle
from the airplane
so we knew that it was a catastrophic failure
a very abrupt failure that would have been
difficult for the flight crew to overcome.
OKAY, let's start here
and work backwards to first impact.
I want a record of everything.
You need to look at the crash site
to collect the physical evidence
that is the most important aspect
of any air-crash investigation.
They look at the wreckage to try and evaluate
the sequence of events, leading to the crash
They look for the 2 wing-tips,
the tail, the tip of the nose
A survey of the crash site soon confirms
those key pieces are among the wreckage.
It tells us that there were no in-flight break up
and that the airplane stayed intact
until it collided with the ground
Get a shot of this wing for me, please
For expert investigators,
the size and shape of the impact zone
is also a clue.
If an aircraft flies through trees,
a long cut through the trees
the investigators get an idea that the aircraft
hit the ground approx. flat in a shallow descend
In this case,
the wreckage was very concentrated
which indicated that
the aircraft hit the ground at a very steep angle.
We need to get some measurements on these trees
Investigators hope the broken trees
will shed light on witness statements
about the way the plane was flying.
It was like it made a wild right turn,
but it was unusual, because it turned out
the nose went into the ground
I was right there when the engines touched
I heard the explosion
I saw the fire and the smoke
When the plane crashed,
it sheered the tops off of many of the trees
in the woods there
one of the things that we did was,
we measured the heights of the trees
where the tops were sheered off
to give us a fairly good idea of the roll angle
which was a large left roll
In the plane you felt light in your seat
you would have turned left at almost 90°
That would have been a traumatic experience
What could make it roll so far over?
When the airplane rolled to the left,
it could be only due to 2 things
Perhaps the pilot wanted to roll to the left
or there was a malfunction
that the pilots couldn't counteract
What do you got for me?
Ritter examines airport flight records.
He is looking for anything
that might have triggered the deadly roll
Clear skies on approach
almost zero traffic.
There were no other airplanes in the area to avoid
It was a nice clear day
so we didn't have an explanation for
why the airplane rolled so violently to the left.
Ritter won't be getting an explanation
for many on board flight recorders
At the time of the crash,
commuter-planes aren't required to carry them
Investigators will have to solve the mystery
without one of their most useful tools
Not having those
made the physical evidence all themore important
What have you got for me?
This is the third one
The engines have been badly damaged in the crash
the propeller blades have been torn off
Did Flight 2311 suffer
some kind of engine failure?
It is too soon to say
One thing Ritter does know is
that the pressure he is feeling from the media
is not about to let up.
The deaths of senator Tower
and astronaut Sonny Carter
are sure to keep the investigation
in the spotlight
Ritter is determined
not to let the pressure get to him.
Sometimes we feel pressure
to do investigation quickly,
but for the most part,
it is more important to get it right
I'd rather take the time
and have a good analysis of the evidence
before they come out with the probable cause.
You won't find their answers here
Let's get what we can
back to the hangar
Wings, tail, engines, instruments
The question now:
Can he find enough evidence to ever solve
the mystery of Flight 2311?
Investigators begin the painstaking task
of sorting through the wreckage.
They are searching for any evidence
that might hint at why the Embraer 120
rolled sharply to the left and crashed,
killing every one on board.
Al-right guys, let's start with the ailerons
They wonder if one of the flight control surfaces
was malfunctioning.
The control surfaces
are the movable flaps
typically at the back of
the wings, the tail and the fin
which allow the pilot to roll the aircraft,
to pitch the aircraft and to yaw the aircraft
the three motions an aircraft can do
They need to examine
every component of the flight control system
What you are looking for is
☮Continuity of the controls
☮Are the hinges all intact
☮All the actuators are in their proper position
☮Were the control surfaces themselves
in a reasonable position at the time of impact?
The ailerons look good
When we examined the control surfaces,
we didn't find anything unusual at all
Everything checked out normally.
So, the control-systems were all working
Why would a plane do this?
Left bank, almost 90°
and almost straight down.
Maybe the engine
Without the black boxes,
it is basically a process of elimination
We analyse all of the physical evidence
and come up with the most compelling scenario
that matches that evidence
OKAY, let's see what we can find in here
If one of the 2 turbo prop engines failed
it might explain
why the plane went into such a steep roll
We wanted to look for
any indications of an engine problem
so we retrieved both the left and the
right engine and their propeller systems
and took those back to the laboratory
for further examination.
Investigators soon find some tell tale evidence
leaves and branches inside the engines.
When we find vegetation in the engines
that tells us that the engine is operating
because it is sucking in air
and it is pulling in the leaves and the vegetation
That is an indication that it is making power
They must have spinning to the very end,
no doubt about it.
We confirmed that both engines were operating,
right up until the moment of impact.
With engine failure ruled out,
Ritter turns his attention to
the other main part of the propulsion system
Let's take a look at these propellers.
Tom Haueter is an NTSB investigator
When we started doing tests
in the propeller system,
we didn't know
where it is going to lead this
But it was something we had to eliminate,
and nothing else.
Deep inside the propeller unit
investigators uncover an important clue
Aha, there we are!
Take a look
There is a small mark where two parts of
the propeller mechanism slammed together on impact
The witness mark might be enough to tell
investigators how the propellers were operating
You can literally match up the scratch marks
between both pieces
and you'll know what the angle
of the propeller-blade was from that measurement
Mark this one 22°
The Embraer 120 has a constant speed propeller
The blades spin at a steady rate in flight
When the pilots need more power,
the blades twist
changing their angle to take
a bigger byte out of the air
and provide more thrust.
And in flight,
it acts like the automatic transmission in a car
It is as if it is changing gears
to match the engine load, the speed
whether the plane is climbing or descending
and it does so constantly and automatically
without intervention by the crew.
There are witness marks inside both propellers.
Now, this one looks like 3°
The marks tell Ritter the angle of the blades
when the plane slammed into the ground
We immediately noticed the difference between
some of the blade angle measurements
for the left engine
versus the right engine
These guys had a big problem
with their left propeller.
The blades were almost flat
The left side propeller blades
are at a dangerously low angle
one that is never used during flight.
At 3° the blades are so flat,
they would act like a wall
blocking the flow of air the plane needs
to maintain lift
There is certain speeds,
certain regimes of flight,
if the propeller goes flat enough
you have situations you cannot control the plane
Investigators study the mechanism
used to control the left propeller.
Hey, look at this
They make a disturbing discovery
It is completely worn down
The teeth on a key piece of the gear-mechanism,
known as the quill are almost entirely worn away
Investigators may finally have the lead
they've been looking for
This is what is supposed to look like
With its teeth worn away,
the quill can't lock onto the gear system
that controls the angle of the propeller blades
The discovery might explain
why the propeller blade slipped
to such a dangerously low angle.
Once we noticed
that the quill teeth were severely worn
we started theorizing
what would happen in that situation
It was pretty clear
that the control of the blade-angles could be lost
That was a Eureka moment for us
because now we had a serious malfunction
that we could examine.
This could definitively be it
Ritter digs into manufacturing's reports
trying to find out how such a vital part failed.
But what he finds only adds to the mystery
The quill teeth are made
of an extremely durable case hardened metal
They are definitively made to last
It is hard to imagine what could have caused
such rugged teeth to wear down so badly.
Something didn't work as planned.
Let's find out everything we can
about every one of these pieces
Ritter is certain he has found
the critical clue:
worn down teeth on the quill
that could have allowed the propeller blades
to slip to a dangerous angle.
but he soon learns
there is a big problem with his theory
The manufacturer says it is impossible.
Engineers at Hamilton Standard
included a fail safe feature
when they designed the propeller
It should be impossible for the blades
to go flat during flight.
Manufacturers have to demonstrate
that their systems are fail safe
All the test and research have been done
before this said:
even if you disconnect,
that will not result into an accident
If there is ever a problem
with the mechanism controlling the angle,
the blades are designed to move on their own
to the feathered position.
The feathered propeller blade,
the leading edge of the blade
is directly into the wind
that is minimal drag,
no thrust, but very little drag.
A feathered propeller
can't endanger the safety of the flight
If this rod disengaged because of any cause
the propeller should go to feather
which would result in the loss of the engine
and the crew would be able to cope with that
and land on one engine
as they are trained to do
The runway is in sight
This got to be it
It has to be connected somehow
But Ritter isn't convinced
by the manufacturer's assurances.
His gut tells him the worn quill did allow
the propeller blades to move to a dangerous angle.
But without flight data
his investigation has hit a wall
He has no way to prove
the quill brought down Flight 2311
Tom, welcome to the team
Tom Haueter:
No problem, Jim I'm glad to help
With the investigation stalled
Tom Haueter joins Jim Ritter to hunt for answers
in the crash of Atlantic SE Flight 2311
I was a little nervous about this when in that
when I first jumped in to it,
I didn't really know what was going on
other than it had a standstill
and they were looking for me to get it moving
What do we have?
I think everything we need to know
is right here on this table
We have a part of the propeller control assembly
and the teeth on the gear are essentially gone
That is very unusual
Could that be a part of the accident,
we didn't know
Haueter and Ritter
study the design of the mechanism
Tom, look at this
We don't see anything obviously wrong
with the crew
We don't see anything wrong
with the structure
We don't see anything wrong
with the engines and the flight control system
but we do have
a severely worn transfer to the quill
Is it possible that this could have resulted in
the loss of control?
They discover that shortly before the accident
Hamilton Standard started using a harder,
more abrasive coating on a key part,
known as the transfer tube
Its grooves mesh like clockwork
into the teeth of the quill
The change had an unexpected consequence
It turned into a giant file
The splines on the transfer tube were
much harder and rougher than the quill teeth
and it was almost like sandpaper
The tube was wearing down the teeth on the quill
The discovery explains
the worn teeth on the quill.
but it still doesn't explain the crash.
The design of the propeller should ensure
it snaps to a safe position,
even with the worn part.
The propeller manufacturer believed
that they had a fail safe condition
so that even if they have this problem
the blades would be slowly driven
to the feather position.
In spite of what all their data says
the investigators want to see for themselves.
They set up a test
at the manufacturer's facility
We had an engine and a propeller combination
mounted in a test cell.
In order to determine that something is fail safe
the engineers use a combination of
mathematical analysis, a very structured analysis
testing in a laboratory
and by these methods,
HS convinced themselves that this propeller would
fall in a safe direction i.e. towards feather
Jim Ritter's doubts about the fail safe design
may be confirmed in just a few seconds
The technician flips a switch to free
the transfer tube from the teeth on the quill
but as the test unfolds
instead of moving to flat,
the propeller blades move to the safe position
The fail-safe system performs exactly
as it was designed to
It suddenly seems that investigators
are on the wrong path altogether
When we tested the quill, with the worn teeth
at the manufacturer's facility
We found that the propeller-blades
went to feather
So, at that point, we are basically stunned
Can you play that again?
The test results leave
Ritter and Haueter wondering
Could there be some other factor
they've overlooked?
Tom Haueter:
Hold on, that's it
It is bolted to the ground
The question I raise
well, in flight the aircraft is in turbulence
is bouncing around, there is different vibrations
With the aircraft on the ground
or the engine is mounted solidly to the ground
the airplane doesn't behave the same,
there is different vibration modes
Is it possible that could change
the outcome of the analysis?
I think we have to see it in the air
One of the problems of doing testing
in a laboratory environment
is that you can't always
anticipate what will happen
when you go out
into the real world
Imagine for example an orchestra
that rehearses in an acoustically
perfectly concert-hall
and then performs outside
with random noises, without the sound control
You could see how one might miss
important factors about the sound
We've got to see what happens in the air
I said: Well, the only way to really know
is to do a flight test and find out
Let's do something that actually determines
that this is the case
We are at a point in the investigation
we need to start eliminating things
The investigation into the crash of Flight 2311
moves to Embraer headquarters in Brazil
We really wanted to look at
what would happen in an actual flight,
with the same malfunction
Tom Haueter meets with representatives
from Embraer and propeller manufacturer HS
The team immediately starts preparing
for the test flight
Thank you for doing this
When I first proposed doing the flight test,
it was not well received by almost everybody
People saw no need
why we were spending the time, the money
why we were doing this
But I pushed hard
we need to really prove this
Embraer's chief test pilot, Gilberto Schittini
has agreed to put the investigator's theory
to the test
The risk is part of the game
You have to reduce the risk as much as possible,
but you cannot avoid this all
You just have to accept it
The mindset of a test-pilot
is not of heroic bravery
but certainly one wants to be unflappable in
the face of danger and unusual situations
because you are not very productive if you are
terrified by the goings on in the cockpit
Schittini will fly an Embraer 120
that has been modified to recreate the failure
So, we modified the quill
The teeth have been worn down,
just like Flight 2311
A worn quill will be placed
inside the propeller unit.
This will potentially very high risk
because once we disconnected the transfer tube
in flight
the pilots will have no way
to control the propeller
We put a pitch-lock here,
it won't go past 23°
A mechanical lock has also been added
to stop the propeller from going flatter than 22°
It would be too dangerous in the test-flight
to have blades that go all the way to flat pitch
You'd loose control of the airplane
If the blades move to 22°
then we know they would have gone flat
The blades don't need to go completely flat
to prove that Haueter is on the right track
It was a high risk test,
we had to take extra precautions
in order not to repeat the accident.
Real time data will tell the team on the ground
whether the blades will go safely towards feather
or moving dangerously flat
Everything says that if you disconnect,
the normal frictional drag within the system..
...will cause the components to rotate
towards the feather position
But will that really happen in flight
the same way?
They are about to find out
We are comfortable
that this test can be safely accomplished
That is the difference between
believing and knowing
As the test plane climbs
Schittini takes it over an unpopulated area
just in case
There was a potential
for a severe controllability problem
such that the pilots might have to abandon
the aircraft and parachute to the ground
The propellers have been set
for a normal angle for flight...
...around 30°
EMB120, do you copy
There was a lot in terms of facts that..
What if this airplane crashes?
What if we loose the airplane?
I am the investigator in charge,
I am the one who is running this test
This could be all my responsibility
Copy we are ready to disengage the prop
They would pull the lever that would disconnect
the Propeller Control System from the propeller
and see what happens
The most dangerous part of the test flight
is now under way
Let's do what we have to do
and get this airplane on the ground
This was in my opinion a very dangerous manoeuvre,
because
We didn't know what the outcome was going to be
and they already knew
the aircraft was uncontrallable
under certain circumstances
propeller blades angle
no problems
No control issues
We were thinking about
the test to do and nothing else
You have to focus in the job you have to do
Once you focus,
forget about everything else.
As soon as the propeller
quill is disengaged
the blades begin doing what
they were designed to do
Moving towards the feathered position
As they start drifting towards feather
we thought: OKAY
this test is just going to prove
what everybody said it would do
You could see on their faces
and they didn't say anything
but you could hear the relief
The propeller blades keep moving
towards the fail safe position
For Tom Haueter,
Brazil is a long way to come for failure
It seems like the test that he hoped
would solve the mystery of flight 2311
is going to leave the investigation
back at square 1
With the cause of the Atlantic SE crash
still unknown
and the entire investigation on the line
the Brazilian test flight takes a dramatic turn
We saw that the RPM was increasing
Slowly, but it was increasing
so I knew that the propeller
was going in the wrong direction.
As the flight continues, the blades begin
drifting in the opposite direction
towards the dangerous flat position.
Here we go
The aerodynamics, the vibration,
the actual in flight loads on the propeller
are just something you can't predict in a test
It is something that actually had be flown
Schittini remains smooth and precise
on the controls
We started feeling a rolling moment
to the left
The blades go as flat
as this test will allow: 22°
For Haueter, it is a victory
The risky test-flight has paid off
Their theory about the crash
is back on solid ground
The Hamilton Standard and FAA people
just went white,
I mean, they were in a shock
You could tell that everything they believed
it just have been thrown out the window
That is the slightest they are going to get
Do you have what you need?
Affirmative, we have everything we need
OKAY, we are going back
Once the test-flight is finished
it is finished
Don't stay around looking for trouble
Go back home
Seen the data right then,
it took a load of
I said Wow,
We now now what happened
It was obvious
The team has just one more question to answer
before they can explain the crash
What happens when blades go completely flat?
In de flight test,
they were only able to go to
a blade angle of 22° for safety-reasons
but during my flight simulations,
I went down to flight angles as low as 3°
In a flight simulator,
Jim Ritter recreates the fatal propeller
malfunction aboard ASA 2311
As the blade angle get to very low values
the airplane was uncontrollable.
The simulation allows him to experience
what the test pilots couldn't
The flight test proved:
we could have a blade go to flat pitch
The simulator proved:
You loose control evidently
That was helpful
You could say that the simulator testing
was the final piece of the puzzle
They finally understand the full story
behind the deadly crash
When the flight-crew began
preparations for landing
the teeth on the quill were worn down
still operational
Preparing for landing put renewed pressure
on the already worn teeth in the quill
They could no longer hold
Once the teeth gave way,
the blades were free to
drift to a different angle
As the turbulence of flight and mechanical
vibrations from the engine shook the propeller
the blades moved to a dangerously flat position
The fail safe....failed
The propeller design couldn't hold the blades
at a safe angle
and the plane
became less and less controllable
For a while you could fight this,
you can control movements
with the wheels, the rudder
But unfortunately they were in a situation
where it didn't matter
even if you were the best pilot on the planet
you were not able to control that airplane
It's going to roll over,
it is going to dive towards the ground
Ironically, human error is almost always
underneath the cause of an accident
even it wasn't the pilot or a mechanic
In this particular case,
an engineering change was made
which, who had the intention to
didn't work out
The worn quill teeth was a time bomb
waiting to go off.
There was nothing he could have done
to safe that plane
What I really felt was:
OKAY, now we now it,
we'd better tell the rest of the world.
and get operators
to start changing their equipment right now
The fail safe system doesn't work
After the accident, additional safe guards
were added to prevent this type of failure
The change that affected not only the Embraer,
but several other turbo prop aircraft as well
The FAA also demanded
more frequent inspections of propeller-quills
The most important thing
we've learned from the accident of Brunswick...
...is assumptions:
You cannot rely on assumptions
It was assumed
the ground test was as good as flight test
That one assumption unfortunately
turned out to be fatal for some people
Narrator
Jonathan Aris
Subtitles
Rein Croonen