Episode Transcript
Transcripts are displayed as originally observed. Some content, including advertisements may have changed.
Use Ctrl + F to search
0:00
I'm Dr Carl, coming to you from
0:02
the lands of the Gadigal people of
0:04
the Eora nation. I acknowledge Aboriginal and
0:06
Torres Strait Islander peoples as the first
0:09
Australians and traditional custodians of the lands
0:11
where we live, learn and work. G'day
0:15
Dr Carl, Shirtloads of Science University of Sydney,
0:18
talking with Dr Sonia Brown. Thanks
0:20
Dr Carl. And you're a lecturer
0:22
at the University of New South
0:24
Wales. Correct. And your field is?
0:27
Airspace Design. And a few airplanes have
0:29
been doing bad things recently. We're talking
0:31
now in early 2024. And
0:34
the Alaska Airlines airplane was
0:36
a 737. Now the
0:39
737, it's been around for about
0:41
half a century. Is that right? Yeah, it's been
0:43
around for over 50 years. The original 737s came out in
0:45
the 60s. Oh
0:48
my God. So it went through the 737 original,
0:52
then the classic, then the next generation. Now we've
0:54
got the MAX. And Alaska
0:56
Airlines ones, what model was that? A
0:59
Boeing 737 MAX 9. And
1:02
it was a plug for an
1:04
additional optional emergency exit door. Why
1:07
would you have an emergency exit door optional? Like
1:09
if when I go into a chemist I don't
1:11
say, oh can I just have the average grade?
1:14
I want the full maximum. So why don't they
1:16
have the full maximum number of emergency doors? On
1:19
aircraft the amount of emergency exits you need is based
1:21
on the number of people you're carrying because you need
1:23
to be able to get them all off an emergency,
1:25
which needs to be certified to 90 seconds.
1:28
With particular aircraft you can certify them to
1:31
a maximum number of people, but certain airlines
1:33
will choose to carry less than that. For
1:35
example, 737 MAX 9 is
1:37
certified to carry over 200 people maximum. However,
1:42
if an airline wants to run a
1:44
dual class configuration, so say a spacious
1:46
business area up the front, and then
1:48
a small economy section it might be carrying less than
1:50
200 people, so it
1:53
may not need the extra emergency exits.
1:55
And therefore because of weight reasons and
1:57
complexity reasons, they'll replace it with this.
2:00
plug. It's just an aluminium plug
2:02
in the door. The emergency exit
2:04
door itself weighs a lot more because it
2:06
has to have all the features to actually
2:08
be able to open out in emergency plus
2:10
all the emergency systems. So for example, extra
2:13
emergency exit slides that have to be able
2:15
to deploy from there. You'd also have to
2:17
have those seats spaced further apart of that
2:19
location to have that emergency exit aisle as
2:22
well. It is actually significant for the airlines
2:24
if they don't want to carry that amount
2:26
of people to be having that whole extra
2:28
systems two times out on both sides of the
2:30
aircraft. I had no idea that they would put a
2:32
plug in the side of the 737. How long have
2:35
they been doing this offering various versions of the 737
2:37
which is one of the most popular
2:39
light jets ever made, isn't it? There's
2:41
thousands of 737s out there. The plug that is on
2:44
the 737
2:46
MAX 9, a very similar plug was also
2:49
used in the next gen series of aircraft.
2:51
So there was the Boeing 737-900ER which was
2:53
introduced in around 2007 which also has variants
2:55
which include
3:00
this plug. So these have been flying for
3:02
over a decade, 15 years or so. In
3:05
this case, the door popped when
3:08
it got to 60,000 feet. Does
3:10
the fact that the door popped halfway
3:13
to maximum altitude so it wasn't full pressure
3:15
difference between the inside and the outside, does
3:17
that give you any hints as to what
3:19
might be going on? It does
3:21
at 16,000 feet. We note that
3:24
inside the aircraft typically we pressurised
3:26
to around 8,000 feet.
3:28
So somewhere in the 6 to 8,000 feet but we assume
3:30
8,000 feet. That means our
3:33
pressure difference at that point was around 20,000 kilopascals
3:36
whereas to the maximum certified altitude,
3:39
it's almost three times that pressure difference.
3:41
So the fact that it has come
3:43
off so early
3:45
indicates that there's probably
3:47
something physically wrong. So
3:50
either something went
3:52
wrong in manufacturer installation or potentially
3:54
there's been damage to it since
3:56
it's been in service that hasn't been detected rather
3:59
than I guess what would be most scary
4:01
or critical which would be an actual design issue.
4:04
It's wonderful listening to your engineer brain
4:06
work through the different possibilities and
4:09
kind of put down at
4:11
the bottom of the list a design
4:13
problem because it's been flying for so
4:15
long. You casually mentioned 90 seconds.
4:17
I keep on reading about this and I've read
4:19
in Aviation Week of Space Technology that what they
4:21
do when they want to certify a plane for
4:24
96 and they'll actually prove it can be done.
4:26
They get all the cabin crew, they train them
4:28
up, they pick all the skinny ones and then
4:30
they say go and they can get all of
4:32
them out in 90 seconds. But let's go
4:34
back to the I think it was
4:36
the first of January when there was
4:39
a collision between two planes at a
4:41
airport in Japan or was it two places? It
4:44
was a Japan Airlines Airbus A350.
4:46
So this is a big
4:48
plane twin aisle as compared
4:50
to single aisle and a
4:52
Coast Guard Bombardier Dush 8.
4:55
It took more than 90 seconds and there
4:57
was a report of some person slowing people
5:00
down by reaching up and then grabbing his
5:02
bag over had been and then putting it
5:04
on the slide and then tearing the
5:06
slide on the way down. So the 90
5:09
seconds is the optimal situation is what we
5:11
certify for but it does assume
5:13
that everything is operating as expected
5:15
which also includes things like all
5:17
of the exits and all of
5:19
the slides operating. So in the
5:21
case of the Airbus A350 it's
5:24
fitted with eight emergency exits
5:26
and eight emergency slides. My
5:29
understanding is they only had
5:31
three emergency exits and emergency slides
5:34
operating for that evacuation. So that
5:36
does obviously make it take
5:38
significantly longer which is why
5:40
in design also we want to make
5:42
sure that when there is an emergency exit as
5:44
long as possible is available which as I say
5:47
thankfully for the Japan Airlines case everyone got off
5:49
safely. Both the forward
5:51
door and the rear door near the
5:53
tail were operating on the side where
5:56
the wing was damaged and on
5:58
the other side which wasn't
6:00
where the damage was, only the
6:02
aftmost, only the one near the tail was
6:05
operating. So they didn't operate on either side
6:07
near the wings, obviously on the case where
6:10
the damage is. My understanding is
6:12
they also didn't operate the front doors
6:14
on the side where the damage wasn't
6:17
because they some reports that
6:19
I haven't been able to confirm that
6:21
engine was not shut down properly. So
6:24
you can't have people evacuating out in
6:26
front of an engine that is not
6:28
fully shut down. And if people get sucked into
6:30
an engine that's very bad. That would be very
6:33
bad. On an aeroplane, how
6:35
much attention do you play to the flight crew because what
6:37
I do is I stare at them
6:39
and I give them a little wave at the beginning and then
6:42
that way I know they know I'm paying attention
6:44
because I feel so sorry for them when people
6:46
ignore them. Because I know that in
6:48
an emergency, I discovered this in medicine,
6:51
in an emergency you drop to
6:53
your lowest level of training. And
6:56
so on one hand I've seen the photo
6:58
of people on the
7:00
wing of the A320 that landed
7:03
in the Hudson River and they're standing on the
7:05
wing and they're all wearing their life jackets inflated
7:08
and about three-quarters have got them inflated wrongly. They
7:10
haven't paid attention. Some of them are up in
7:12
the air above their head. I haven't
7:15
done the place the strap around your waist and
7:17
clip it tightly and pull on the strap to
7:19
tighten. They've missed that bit. Yeah so I
7:21
always firstly wave to them then pay attention
7:24
firstly because they will look after me I
7:26
hope if things go bad and then secondly
7:28
as an act of solidarity with a fellow
7:30
professional performer because I feel sorry for the
7:33
people not paying attention. You've probably memorized all
7:35
this stuff anyway though. I've pretty much got
7:37
it memorized in my head but I'm certainly always
7:39
looking at them and making sure I am
7:41
aware of where I'm seated relative to
7:43
those exits. And I think
7:46
one of the most important things that
7:48
the flight attendants say is
7:50
in the case of a loss of pressure
7:53
which we had already on the Boeing 737
7:55
MAX 9. When those
7:58
masks come down you should fit your own mask
8:00
first before helping other people including
8:02
children. Now as a parent myself
8:05
it's a conflict right the thing you most care
8:07
about is your child but the reason they say
8:10
that is because in the event of a loss
8:12
of pressure depending obviously on the altitude without
8:14
oxygen yourself you can lose consciousness and
8:16
then you can't help yourself or your
8:18
child. So it's actually one of
8:20
the most important things they say is fit your
8:23
own mask first. Order
8:25
is important and listening to the flight
8:27
crew in those situations and as
8:29
quickly as you can safely. We've
8:32
talked about two sets
8:35
of airplane bad things the
8:37
Airbus 380 and the 737 then
8:40
we had a bit of what
8:42
looked like a design problem stuck
8:44
through in the first of the
8:46
max theories where according
8:49
to various articles I've written in New
8:51
York Times and Aviation Week in Space
8:53
Technology there was a conflict between the
8:55
engineers and the bean counters or the
8:58
accountants in the company and in this
9:00
case the summary is I guess that
9:02
the bean counters one and two planes
9:05
crashed. I'm not sure
9:07
I want to you know go labeling
9:09
blame within Boeing on this
9:11
one but the big design fault
9:13
in this case was that they
9:16
had an angle of
9:18
attack sensor which basically monitors
9:20
the angle of the aircraft
9:23
relative to the oncoming air flow. It's like a
9:25
little flippy thing that sticks out the side and if
9:27
the wind is running parallel to the
9:30
length of the airplane it'll be dead flat
9:32
and it can swivel around up and down
9:34
depending on if the airflow is going from
9:36
top to bottom or bottom to top or
9:38
just along the fuselage it swivels around. Yeah
9:40
so they had multiple of the sensors
9:42
but they had one of these sensors
9:44
that was providing input to a particularly
9:46
new system that they installed which
9:49
has a really long
9:51
complicated name called the Maneuvering
9:54
Characteristics Augmentation System MCAS
9:56
but basically it means an anti-stall system
9:59
so what it was to do was
10:01
detect if that angle of attack became
10:03
too high due to drift
10:05
over time that maybe the pilots didn't notice
10:07
and then push the nose down in that
10:09
case. That was what it was designed to
10:11
but one of the issues with the
10:14
system was that it was just reading off of one
10:16
sensor. There's two sensors and they deliberately
10:19
ran only off one and I read a
10:21
report by an engineer who said he tried
10:23
to get them to read off both
10:25
and have it what he called a
10:27
synthetic airspeed system but his bosses wouldn't
10:29
let him do it. I don't know
10:31
about the internal decisions there
10:33
but unfortunately what went
10:35
live was reading off of one
10:37
sensor which is basically a no-no. Because you got
10:39
no redundancy. Actually we should just back up a
10:42
bit here. So what had happened was that they
10:44
had the series the first ones in the classics
10:46
and the next gen they went for the max
10:49
and they were limited by how big an engine they
10:51
could put on it and Airbus was coming out with
10:53
new designs and it was easy for them to put
10:55
on a new bigger engine because they
10:57
were working from scratch but here they're putting on a
11:00
new engine on a 50-year-old plane and it
11:02
was so big they decided to mount it
11:04
more forward and more up. So
11:07
it changed the stability characteristics so
11:09
that's why they thought it might
11:11
gradually drift up because the stability
11:13
profile of the max was different
11:16
to those particularly those next-gen aircraft
11:18
that it was building off of. By the age
11:20
of a bit of background which I never realised
11:22
before reading all this and airplane normally flies a
11:24
couple of degrees nose up. Ideally
11:27
mostly in cruise we
11:29
would want the fuselage to be mostly
11:31
horizontal however the wing should
11:33
be set at what we call an
11:35
incidence angle so the wing will be
11:38
slightly up relative to the fuselage to
11:40
produce the lift required for cruise. So
11:42
I was wrong when I gave the information
11:44
that they're tilted up by a degree or
11:46
so with the whole airplane. The air is
11:49
running parallel to the fuselage. Ideally
11:51
for minimum drag we'd want it
11:53
parallel but to do that and
11:55
to still get the lift required our
11:57
wing will be tilted effectively a little.
12:00
bit. The entire wing on installation,
12:03
on installation, it will have an angle relative
12:05
to the fuse large which is called an
12:07
incidence angle. I kind of love talk to engineers
12:09
who tell me where I'm wrong and get me
12:11
started on getting things right. So with the Mac
12:14
the engine was mounted higher, it
12:17
was a bigger more powerful engine, it was mounted more
12:19
forward and higher and it led to a
12:22
possible tendency for the nose to come up
12:24
and so they put in some software to
12:27
bring the nose down and didn't
12:29
tell the pilots about it, didn't make
12:31
it part of the training. Yeah so
12:33
it wasn't part of any of the
12:35
original training, apparently in the
12:38
original manuals there was something that
12:40
defined the term MCAS to say
12:42
the MCAS equals maneuvering characteristics augmentation
12:44
system but that was the only mention.
12:46
Articles that I read and I could
12:48
be wrong on this implied that Boeing
12:51
wanted pilots who trained on previous 737s
12:53
to be able to walk into this
12:55
one and not do any extra training.
12:58
Well it was a very limited training
13:00
like on an iPad training rather than
13:02
full simulator training. Right, some of the
13:04
engineers were saying maybe
13:06
they should have gone through recertification training
13:09
to handle this MCAS system. It certainly
13:11
would have helped but the problem
13:13
that we had with the Macs back
13:15
then and noting that obviously the system's
13:18
been updated since the incident
13:20
in 2018 and 2019 even
13:23
after the first incident which was lying
13:25
there. Boeing then put out
13:27
a bulletin which outlined
13:29
the system said what you should do if
13:32
it played up and
13:34
the Ethiopian pilots when
13:36
this happened to them, the second
13:38
plane in March 2019, they
13:41
tried to follow that process. It was
13:43
very clear they tried to follow the
13:45
process of Boeing outline so I don't
13:47
think even originally just training was sufficient.
13:50
There were design faults which I say
13:52
now have been addressed thankfully but should
13:54
have been addressed from the get-go. So
13:57
now if those sensors differ by more than
13:59
five... five and a half degrees, the
14:02
MCAD system won't activate. It
14:04
fully cuts it out. It now reads
14:06
from both angle of attack sensors and
14:09
importantly checks the difference between them. So
14:11
in the Ethiopian case at least, it
14:13
was reading off one and that sensor
14:16
read something like 75 degrees, it
14:18
was clearly an erroneous reading. So
14:20
now because it reads off two, it checks them. And
14:23
if the difference is more than five and a
14:25
half degrees, it basically describes them
14:27
both and the system will not activate. The
14:30
other important thing is even if
14:32
they both said it was wrong,
14:35
it now won't activate repeatedly off
14:37
a single elevated reading. So again,
14:39
when it happened, it
14:41
kept reading that it was
14:43
like 75 degrees or whatever it was. So even
14:45
if they both read a number, which meant the
14:48
system would normally activate, if the
14:50
pilots have corrected for it,
14:52
it won't keep activating repeatedly, which it did
14:55
in both the Lion Air and the Ethiopian
14:57
case. And then the other
14:59
thing they did is they reduced control
15:01
authority of that MCAS system.
15:03
Because again, the Ethiopian case,
15:05
it was really clear, both the
15:08
pilot and the co-pilot were pulling back, which just
15:10
basically meant they were trying to use the elevator.
15:13
So the elevator also changes the pitch of the
15:15
aircraft. Is the elevator the flappy thing on the
15:17
back of the wings? It's on the back of
15:19
the tail. Oh, so there was one
15:22
way at the back of the airplane. Is it also
15:24
called a stabilizer? At
15:26
the back, we have the horizontal
15:28
stabilizer, which in the 737 is
15:30
an all-moving stabilizer. So that's what
15:32
the MCAS was controlling this stabilizer.
15:34
So it's technically set
15:36
at an angle, and it kind
15:38
of stays there, but it can move
15:40
small amounts. The front of the wing
15:42
of that stabilizer elevator thing, either up
15:44
or down, but you're implying that it
15:46
can also have one side up and
15:48
one side down. So if you were
15:50
thinking of the horizontal tail of an
15:52
aircraft. Yep, the horizontal bit, not the
15:54
vertical, the
15:56
horizontal bit. The first two thirds
15:58
of that, right? think of that as a
16:01
six part but that's standard
16:19
pitch control so the standard pitch
16:21
control the pilots would use would
16:23
be the elevator the all moving
16:25
horizontal tail or the all moving
16:27
stabilizer that's the front bit and
16:29
that would normally be set to position
16:32
what we call is trim the aircraft basically
16:35
make it stable in straight and level flight
16:37
so it would move that so that we
16:39
have to have minimum control inputs when
16:41
we're in straight and level flight basically
16:44
balance what we call the moments on
16:46
the aircraft. Now what
16:48
do you call the flappy things on the backs of
16:50
the big wing? Inboard will
16:52
have flaps and outboard
16:54
will have ailerons. Oh so neither of
16:56
them are called elevators? No elevators are only on
16:58
the tail. I've also heard the word flaperon. Flaperon
17:01
is basically a cross between a flap
17:03
and an aileron so there's elevons for
17:05
example which is a cross between an
17:08
elevator and an aileron so it depends
17:10
on the control but the most basic
17:12
controls we have is on
17:14
the tail the flappy bit at the back of
17:16
the elevator and on the wing the
17:18
flappy bits of the back inboard is
17:21
our flaps and outboard
17:23
is ailerons but as I say there
17:25
are combinations which have funny names like
17:27
elevons or flaperons etc. Are
17:30
you happy to fly in airplanes
17:32
in general do you think there's still a safe
17:34
way of transport? Flying is the safest
17:36
mode of transport we have not only
17:39
am I happy to fly and fly when
17:41
I can I happen to live under a
17:43
flight path. You're afraid of your flight path you
17:45
can see them dangle at Dunlops as they come over you?
17:47
I think that's a little after me I don't think
17:49
I'm quite that close to the airport but I
17:51
do get to watch the planes come over sometime
17:53
which is nice. How can people follow you and
17:55
you'll find work or if they want to become
17:57
a student of yours do you take students? Yes,
18:00
I take students. I take research students. They
18:02
can just look me up, Sonia Brown at
18:04
UNSW or if anyone wants to follow me,
18:06
they can look me up on LinkedIn. I
18:09
don't know why, but for a long time I've believed,
18:11
but wrongly it turns out that the fuselage
18:13
was slightly tipped up. I
18:16
didn't think it through a course or not
18:18
because you want to get dragged. You just
18:20
tip the wings up slightly, the fixed wings.
18:22
Thank you very much Dr. Sonia. Thanks Dr.
18:24
Carl. Shirtloads of science is washed, spun and
18:26
aired by the University of Sydney.
Podchaser is the ultimate destination for podcast data, search, and discovery. Learn More