0:00
How do you actually stop something
0:02
that's moving at 25,000 mph?
0:06
That's the speed you're traveling when
0:08
you come back from the moon. And you
0:10
need to go from that speed to zero,
0:13
preferably in one piece. There's no
0:16
single answer to that question.
0:19
The concept goes back over a 100red
0:21
years. Robert Goddard was thinking about
0:26
But for the past seven decades,
0:28
engineers have been building real
0:30
solutions. At least half a dozen
0:32
completely different approaches. Each
0:34
one involving trade-offs, failures, and
0:37
some genuinely surprising engineering.
0:41
We already talked about the Aremis heat
0:43
shield in previous videos,
0:45
but that heat shield is just one answer.
0:52
When a spacecraft re-enters the
0:53
atmosphere at orbital speed, it's
0:55
carrying an enormous amount of kinetic
0:58
energy. Now, you might think, why not
1:00
just fire your rockets in the opposite
1:02
direction and slow down? In theory, that
1:06
works. In practice, you'd need almost as
1:09
much fuel as it took to get you up there
1:11
in the first place. And fuel is heavy,
1:14
which means you need more fuel to carry
1:17
that fuel. It's a vicious cycle that
1:19
rocket engineers call the tyranny of the
1:22
rocket equation. So instead, we use the
1:26
The atmosphere is your brake pedal. When
1:30
you hit it at hypersonic speed, it does
1:32
an incredible job of slowing you down.
1:36
The problem is that all that kinetic
1:38
energy has to go somewhere and it goes
1:41
into heat. And here's the misconception
1:44
that heat is not mainly caused by
1:47
friction. It's caused by compression.
1:51
The spacecraft is moving so fast that
1:54
the air in front of it can't get out of
1:56
the way. It gets compressed so violently
1:59
that it heats up to thousands of
2:01
degrees. It actually becomes plasma.
2:05
A superheated shock wave forms in front
2:09
Now, this is where one of the greatest
2:12
counterintuitive discoveries in
2:14
aerospace history comes in. In the early
2:17
1950s, two engineers at NACA, that's the
2:21
predecessor to NASA, named Harvey Allen
2:24
and Alfred Edgars were working on the
2:26
re-entry problem. Everyone assumed that
2:29
a sleek pointed shape would be best for
2:31
surviving re-entry. That's what you'd
2:33
think, right? Cut through the air like a
2:36
knife. Allen and Edgars prove the
2:39
opposite. They showed that a blunt
2:41
shape, a rounded stubby shape, actually
2:44
survives re-entry far better than a
2:46
pointed one. And the reason is elegant.
2:49
A blunt body pushes that superheated
2:52
shock wave forward away from the
2:54
vehicle. The compressed air forms a
2:56
cushion. Most of the extreme heat stays
2:59
in the shocked gas and flows around the
3:01
vehicle rather than into it. That
3:04
discovery was classified as a military
3:07
secret. It wasn't published until 1958,
3:11
but it's the reason every crew capsule
3:13
ever built from Mercury to Gemini to
3:15
Apollo to Soyos to Orion has that same
3:19
blunt rounded shape. It's all because of
3:24
So, with that in mind, let's look at how
3:27
we actually deal with the heat that does
3:33
The oldest and most proven approach is
3:35
the ablative heat shield. And the
3:37
concept is beautifully simple. You build
3:40
a shield out of material that's designed
3:42
to sacrifice itself. As the spacecraft
3:46
re-enters, the outer layer chars, melts,
3:49
and vaporizes, carrying heat away as it
3:51
goes. But underneath that charring
3:54
surface, something else is happening.
3:57
The inner layers decompose chemically,
3:59
producing gases that filter outward
4:01
through the porous char. Those gases
4:04
create a thin boundary layer that
4:06
actually pushes the superheated plasma
4:08
away from the surface. The Apollo heat
4:11
shield used a material called AV coat
4:15
packed into a fiberglass honeycomb. More
4:18
than 300,000 individual cells filled by
4:21
hand. It worked perfectly on every
4:24
mission. On the robotic side, there's
4:27
Pika, phenolic impregnated carbon aber,
4:30
which flew on the Stardust probe at the
4:32
fastest re-entry speed ever recorded.
4:35
And later on, Curiosity's Mars landing
4:38
and SpaceX's Dragon capsule. These
4:41
materials have long track records, but
4:44
here's the thing about proven materials.
4:46
They can still surprise you. When NASA
4:49
built the Orion heat shield for Artemis,
4:52
they went back to Avoat.
4:54
Same name, same concept, but the
4:56
original hadn't been manufactured in
4:59
decades. Some ingredients were no longer
5:02
available. Some of the knowhow had been
5:04
lost. NASA spent over $25 million and 5
5:09
years recreating it, and the version
5:12
they ended up with behaved differently
5:14
than the Apollo original in ways nobody
5:19
I covered the full story in my previous
5:21
video. What went wrong on Artemis 1?
5:23
what NASA found during the investigation
5:26
and why they decided to fly Artemis 2
5:28
anyway with a modified re-entry
5:30
trajectory. I'll link that in the
5:32
description if you want the complete
5:34
breakdown. What I can tell you today is
5:37
that it eventually worked. NASA says
5:40
Artemis 2 landing data, including heat
5:42
shield performance, will shape the
5:44
Aremis 3 timeline. Ablative shields
5:48
work, but they're single use. Every time
5:50
you fly, you need a new one. So, what if
5:53
you could build a thermal protection
5:55
system that survives re-entry and can be
6:01
That was the philosophy behind the space
6:03
shuttle's thermal tiles. Instead of
6:06
burning away, these tiles insulate. They
6:09
absorb the heat, reriate it, and come
6:11
back ready to fly again. The material
6:14
itself doesn't change. It just manages
6:17
the energy. The shuttle's thermal
6:19
protection system was an engineering
6:22
marvel. Roughly 24,000 individual tiles
6:26
made from ultra pure silica fibers,
6:29
essentially sand, processed into a
6:31
material that was 94% air. They were
6:34
astonishingly good insulators, but they
6:37
were also extremely fragile. You could
6:40
crumble one in your fingers. And here's
6:43
what made the shuttle system truly
6:45
daunting. Nearly every one of those
6:48
24,000 tiles was unique. Each one was
6:52
individually shaped to fit its specific
6:54
position on the orbiter. They couldn't
6:57
be mass- prodduced. And between flights,
7:00
they had to be individually inspected
7:02
and often replaced. It was one of the
7:05
major reasons the shuttle never achieved
7:07
the rapid turnaround times NASA had
7:09
originally hoped for.
7:12
The vulnerability of this system was
7:14
demonstrated in the most tragic way
7:18
During the launch of Colombia's final
7:20
mission, a piece of insulating foam
7:23
broke off the external tank and struck
7:25
the reinforced carbon composite panels
7:28
on the leading edge of the left wing.
7:31
During re-entry 16 days later,
7:33
superheated plasma penetrated through
7:35
that breach, destroyed the wing's
7:37
internal structure, and Colombia broke
7:40
apart over the southern United States.
7:43
All seven crew members were lost.
7:47
Colombia is a reminder that thermal
7:49
protection isn't just about material
7:51
science. It's about margins. When the
7:55
protection fails, there is no backup.
7:57
There is no redundancy for the heat
8:01
Today, SpaceX is taking a very different
8:04
approach with Starship. Instead of
8:06
24,000 unique tiles, Starship uses
8:09
mass-roduced hexagonal tiles designed
8:12
for rapid replacement rather than
8:14
obsessive individual maintenance. If one
8:17
is damaged, you pull it off and snap a
8:20
new one on. It's the same basic
8:22
principle, ceramic insulation rather
8:25
than ablation, but with a completely
8:28
different engineering philosophy behind
8:32
Now, tiles and ablatives both rely on
8:35
the same fundamental idea. Let the
8:38
atmosphere do the work. Use aerodynamic
8:41
drag to slow down and manage the heat
8:43
that comes with it. But what if there's
8:46
no atmosphere to work with?
8:50
If there's no atmosphere or not enough
8:53
of one, you're left with the brute force
8:55
option, point your engines in the
8:58
direction you're traveling and fire
8:59
them. Retro propulsion.
9:03
This is the only way to land on the
9:05
moon. There's no air, no drag, no heat
9:08
shield in the world that can help you.
9:11
Every single bit of deceleration has to
9:13
come from the engine. That's why the
9:15
Apollo lunar module looked the way it
9:18
did. spindly, fragile, almost skeletal.
9:22
Every unnecessary gram of structure was
9:25
a gram less of fuel, and you needed
9:28
every drop. The descent engine on the
9:30
lunar module burned for about 12 minutes
9:33
to bring the spacecraft from orbital
9:35
speed down to a gentle touchdown. 12
9:38
minutes of controlled thrust with no
9:43
On Earth, retropulsion has become
9:45
routine, at least for SpaceX.
9:49
The Falcon 9 first stage performs a
9:51
supersonic retropulsion burn on every
9:54
return, firing three of its nine Merlin
9:56
engines into the oncoming supersonic
9:59
airirstream to slow down for landing.
10:01
They've done this hundreds of times now,
10:04
but here's a story that most people
10:06
don't know. In September 2013, SpaceX
10:10
performed the very first supersonic
10:13
retropulsion maneuver on a Falcon 9.
10:17
NASA noticed not because they were
10:19
interested in landing rockets on Earth.
10:21
They were interested in landing things
10:25
In 2014, NASA and SpaceX formed a public
10:29
private partnership specifically to
10:31
study Falcon 9 re-entry data. NASA flew
10:37
highaltitude research aircraft equipped
10:39
with infrared cameras to track Falcon 9
10:42
boosters as they descended through the
10:44
atmosphere. They were particularly
10:47
interested in the altitude range between
10:49
about 40 and 70 km because at that
10:52
altitude and speed, the Falcon 9 first
10:55
stage experiences conditions remarkably
10:58
similar to what a spacecraft would face
11:00
entering the Martian atmosphere.
11:02
NASA was essentially using SpaceX's
11:05
commercial rocket landings as free Mars
11:08
entry research data. And based on that
11:11
work, NASA concluded that the core
11:13
challenge of supersonic retropulsion for
11:15
Mars isn't really a technology problem
11:18
anymore. It's a systems engineering
11:21
problem. The question is how to
11:23
integrate it into a Mars flight system.
11:26
But retropulsion has a fundamental
11:28
limitation and it goes back to the
11:30
tyranny of the rocket equation. Fuel has
11:33
mass. More fuel means more mass to
11:36
decelerate, which means you need more
11:40
This is why atmospheric braking is
11:42
always preferred when an atmosphere
11:44
exists. It's essentially free
11:48
And that's what makes Mars such an
11:50
engineering nightmare. Mars has an
11:52
atmosphere, but it's less than 1% the
11:56
density of Earth's. Thick enough to
11:58
create serious heating during entry.
12:01
Thin enough that it can't slow you down
12:03
nearly enough to land safely. Right now,
12:06
using current technology, NASA can land
12:08
about one metric ton on the Martian
12:11
surface. That's a Perseverance size
12:15
Landing humans and their equipment on
12:17
Mars will require 20 metric tons or
12:20
more. And that brings us to what I think
12:23
is the most exciting technology in this
12:29
Here's something I love. After I posted
12:32
a video about the Aremis heat shield,
12:35
someone in the comments suggested
12:37
something like, "What about a partially
12:39
unfolded umbrella made of heatresistant
12:42
material?" That's a great intuition
12:45
because NASA has been working on exactly
12:47
that concept for over a decade and the
12:51
idea is simple but powerful.
12:53
Traditional heat shields are limited by
12:55
the size of the rocket fairing they have
12:57
to fit inside. The Orion heat shield is
13:00
5 m across, about 16 1/2 ft. And that's
13:04
about as large as you can practically
13:06
build a rigid shield and fit it inside a
13:08
rocket. But what if you could make your
13:11
heat shield much bigger than your
13:13
rocket? Pack an inflatable structure
13:15
inside the fairing, launch it, then
13:18
inflate it in space to a much larger
13:20
diameter before re-entry. More surface
13:23
area means more drag. More drag means
13:27
deceleration starts higher in the
13:28
atmosphere where the air is thinner and
13:31
gentler. You spread the heating load
13:33
over a larger area and you start slowing
13:38
In November 2022, NASA proved this
13:41
works. The low Earth orbit flight test
13:44
of an inflatable decelerator Lofted
13:47
launched as a secondary payload on a
13:49
United Launch Alliance Atlas 5 rocket.
13:53
After the primary satellite separated,
13:55
Lofted inflated its aeros shell to 6 m,
13:59
about 20 feet across.
14:01
At the time, it became the largest blunt
14:04
body ever to re-enter Earth's
14:06
atmosphere. It re-entered at more than
14:12
Temperatures on the heat shield reached
14:16
F, and it slowed to under 80 mph before
14:20
deploying parachutes and splashing down
14:22
in the Pacific Ocean just 8 m from the
14:27
NASA's post-flight assessment, they
14:29
called the performance just flawless.
14:34
The construction is fascinating. The
14:36
inflatable structure is made of
14:38
concentric rings. Think of nested inner
14:41
tubes woven from a synthetic polymer
14:43
that's 10 times stronger than steel by
14:46
weight. Those rings are coated in a high
14:48
temperature silicone adhesive which
14:50
gives the whole structure that
14:52
distinctive orange color you see in the
14:55
Covering the inflatable structure is a
14:58
flexible thermal protection system with
15:00
four layers. The outermost layer is a
15:03
woven ceramic fabric, silicon carbide,
15:06
made into fibers so fine they can be
15:09
spun into yarn and woven on the same
15:12
industrial looms used to make denim.
15:15
Under that are two types of flexible
15:17
insulation and finally a gas barrier to
15:20
keep the structure sealed.
15:22
Lofted was the proof of concept. What
15:25
comes next is where it gets really
15:29
NASA has partnered with United Launch
15:32
Alliance under the TippingPoint program
15:34
to develop the next generation, a 12 m
15:37
Hiad, twice the diameter of Lofted
15:40
designed to recover Vulcan rocket
15:42
engines from orbit for reuse.
15:45
But the real prize is Mars. NASA is
15:48
developing 16 to 20 m versions that
15:51
could land 20 to 40 metric tons on the
15:54
Martian surface. That's the difference
15:56
between landing a rover and landing a
15:59
habitat, between sending robots and
16:02
sending people. It doesn't solve
16:05
everything. You'd still likely need
16:07
supersonic retro propulsion for the
16:09
final descent, but it solves the first
16:12
and most critical piece of the puzzle.
16:14
Getting from interplanetary speed to
16:17
something a rocket engine can handle.
16:22
Now, every method we've talked about so
16:24
far has one thing in common. None of
16:27
them can get you all the way to a safe
16:29
landing speed on their own. At some
16:31
point, you need that final step. And
16:34
more often than not, that final step is
16:38
Parachutes can only deploy at subsonic
16:41
speeds below about 700 mph.
16:45
So, they're never the first line of
16:46
defense. They're the closer, the last
16:49
act. Orion's parachute system is a good
16:53
example of how complex that last act
16:56
actually is. It uses 11 parachutes in
16:59
total deployed in a carefully
17:01
choreographed sequence. First, three
17:04
forward bay cover parachutes pull away
17:06
the capsule's forward heat shield cover.
17:09
Then, two drogue shoots, each 23 ft
17:12
across, deploy to stabilize and begin
17:15
slowing the capsule. Then three small
17:18
pilot shoots pull out the three main
17:20
parachutes. Each main chute is 116 feet
17:24
in diameter and weighs over 300 lb.
17:27
Together they slow Orion from about 325
17:31
mph down to roughly 17 mph for
17:37
All of that 11 shoots a precise sequence
17:40
massive loads on the suspension lines
17:42
has to work every single time. Even the
17:46
materials have been refined through
17:47
testing. The suspension lines were
17:50
originally designed with steel cables.
17:53
Testing showed that a Kevlar nylon
17:55
hybrid worked better, so they switched.
17:58
Parachutes are ancient technology. The
18:01
concept goes back centuries, but getting
18:03
them right for space flight. The
18:05
materials, the sequencing, the
18:07
redundancy, the deployment dynamics is
18:10
anything but simple. It's precision
18:12
engineering applied to fabric and rope.
18:18
So, how do you stop something moving at
18:22
The honest answer is it depends entirely
18:25
on where you're going. If you're coming
18:28
home to Earth from the moon, like the
18:30
Aremis crew, you hit the atmosphere with
18:32
an ablative heat shield, let it char and
18:35
burn and carry the heat away, and then
18:38
deploy parachutes for the final descent.
18:41
The atmosphere does most of the heavy
18:43
lifting. If you're landing on the moon,
18:46
no atmosphere at all, it's pure retro
18:49
propulsion. Engines burning all the way
18:51
down. Every pound of fuel counted, every
18:54
second of burn time critical. And if
18:57
you're going to Mars, that's the hardest
19:00
problem of all. Mars gives you just
19:03
enough atmosphere to create serious
19:04
heating, but not nearly enough to stop
19:07
you. The answer will almost certainly be
19:09
some combination of everything we've
19:11
talked about today. Potentially an
19:14
inflatable heat shield to slow down high
19:16
in the thin atmosphere. Then rocket
19:18
engines taking over to bring you the
19:20
rest of the way to the surface. Multiple
19:23
technologies working in sequence. Each
19:26
one picking up where the last one left
19:28
off. We're still inventing new ways to
19:30
solve this problem. Lofted flew just a
19:34
few years ago. Falcon 9 re-entry data is
19:37
feeding into Mars landing research right
19:39
now. And all of it stands on a discovery
19:42
made 70 years ago that the best way to
19:44
survive re-entry is counterintuitively
19:47
to hit the atmosphere with the bluntest
19:49
shape you can. The engineering only gets
19:52
more interesting from here. If you want
19:55
to come along for it, hit subscribe.