- This is the world's largest 3D metal printer. It was built by Relativity Space, a startup that aims to
print an entire rocket, including fuel tanks and rocket
engines, in just 60 days. I'm like looking inside
a 3D printed rocket that is actually gonna go to space. This giant hunk of
metal, it's unbelievable. This video is sponsored by Omaze. Offering you the chance
to win a trip to space. More about that at the end of the show. - There's a lot of UV
coming off the welds, you can film it, but
don't look directly at it. You get sunburned fast, so
it's like you're suiting up to go in a volcano. All right, we're gonna
go in to the 3D printer, and see how it works. All right, so, yep, just hold this up. Don't look at it. We are in the printer. - I can see it over there. - If we walk around here,
we can get up close. (machinery whirring) (machinery beating) So that's the wire melt team, and the print head moving around. So that's the plasma discharge,
and it's hard to tell, but it's doing things every
couple of milliseconds, it's actually changing
the electric wave form, which is how it's controlling
the deposition so well. - Do you know the temperature of it? Like, is it just above melt temp? - It's just above melting for aluminum. Yeah, probably a few
hundred degrees above. - The melting point of aluminum
is 660 degrees Celsius. So the whole body of the rocket is effectively melted together
one tiny bit at a time. - All the raw metal for the whole rocket
that's printed is this. It's a, you know, we kind of joke it's like Charlotte's web. Like a spider silk, but
this is an aluminum alloy that's on a wire spool. We actually print about
10 inches a second. So this wire is really going super fast, and then the combination of
lasers and plasma arc discharge are working to melt both of
them together at the same time. - So where does the wire come out? - So it's right there, and
then the electric arc discharge happens right at the tip of the wire too. - This is a camera.
- Yeah, that's a camera. - But why would you want
to 3D print a rocket? Is it just because we can? It's funny to me that you had this experience with
3D printing where you're like, oh, 3D printing is clearly the future. Whereas, I feel like a
lot of people's experience with 3D printers as mine has been, it's like incredibly frustrating. I feel like 3D printing is that thing that seems like it should be great, and yet whenever I try it, I don't get a result that I'm happy with. - Yeah, I know. I can tell you, we had
plenty of experiences the first couple of years, where we ended up with a pile
of metal and it didn't work. - But there are actually good
reasons to 3D print a rocket. A rocket has four major systems. Payload, guidance,
structural, and propulsion. The bulk of the rocket is made
up of the propulsion system, including the propellant
tanks and the rocket engines. Cryogenic fuel and oxidizer are pumped through an injector into
the combustion chamber where they react, releasing
an enormous amount of heat. This causes the exhaust gases to expand, exiting the rocket
nozzle at high velocity. The faster this exit velocity and the higher the mass flow rate, the more thrust that can be generated. So rockets are huge, complex
engineering projects, which up to this point have
largely been manufactured using traditional techniques. That means before you
can build the rocket, you first have to build the
tools to build the rocket. For example, to build
NASA's next huge rocket, the Space Launch System or SLS, they first needed to construct the vertical assembly center or VAC. This is a 170 foot tall
tool for welding together the domes, rings, and barrel sections of the rocket's fuel tanks. - They built that like an aerospace thing, and they've had to spin up all
these custom tooling designs and validate that those work before actually starts
building the rocket. And they finally got one
being assembled on the pad after 11 years of development. In contrast, Relativity Space the company, is just five and a half years old, and they plan to launch
their first rocket this year. - I see this as a like
old engineering style versus Silicon valley
style of build something, figure out what's wrong with it, and build another thing
that fixes those, right? The differences, I've always
done that with software. These guys are doing it
with aerospace hardware. - So this is the actual
rocket tank structure of what we're gonna be launching to orbit at the end of this year. So this actual thing
is launching to space. - That will go to space? - This will go to space, and it's by far the largest
3D printed product really, of any type ever made that's gonna fly. I think maybe of any type in the world. - But it still looks 3D printed. Like you can still see the layers. - Yeah, yeah, you can
still see the layers. It only adds an extra
5 to 10% of the mass with the roughness. When you actually cross
section the material and look at the machine parts of it, it looks like normal metal. Like actually at this end,
this is printed as well. We just machine it afterwards. So it looks like normal
metal in the joint sections. - Does the surface roughness
cause any aerodynamic problems? - No, none at all. Yeah, it's actually the
exact same aerodynamically. This whole thing, we simulate
the print before printing, because if you just printed,
you know, the 3D file and said press print, you
would end up with a printer that's warped and like material
falling all over the place that wouldn't actually work. So we've invented software
that reverse warps the whole part before printing it. So the robots are actually doing this really wobbly, weird shape, but then it's actually perfectly straight within a human hair at the entire length. - As it cools
- As it cools. The warpy thing turns into the... - And then we simulate all of that. So it's a big computational solver that simulates it, and there's
many, many other problems we've had to solve to actually get printing a rocket to work. But it's all these little pieces over the last couple of years, and we've really started
to hit some breakthroughs, which is also why now
you see a whole rocket. Yeah, you can step up here actually if you want.
- Can I? - Yeah, yeah. (man laughs) - Hello. I'm like looking inside
a 3D printed rocket that is actually gonna go to space. - Yes. - This giant hunk of metal. It's unbelievable. There's like rings inside. - Those are printed in stiffeners. And so those help prevent the rocket from buckling and crumpling. So if you had a Coke can,
and didn't pop the tab, if you try to step on it,
it's almost impossible because there's pressure inside that keeps it from buckling. But then when you pop the tab, there's no pressure and you
can crunch it super easy. It's not hard at all. So rockets are the same. The 50 PSI of pressure, which is about the same as a car tire, keeps it inflated and
keeps it from crumpling, but then the stiffeners
also help keep it rigid. - Yeah, so believe it or not, a rocket tank is thinner
versus its diameter than a Coke can. So when you look at a Coke can, you know how big it is
and then how thin it is, a rocket tank is actually
thinner than that. So yeah, it's pretty light. It has to be very light. - Sure. Aerospace companies started
using metal 3D printing over a decade ago to
construct small complex parts. For example, the injector. - That is the most important
part of any rocket engine, where you basically gotta
take the liquid propellant, and turn it into a fine mist
that mixes really rapidly. And those have actually been transitioned to 3D printing all over the industry. - Traditionally, something like this, it's a bucket engine injector. So it mixes liquid oxygen and liquid methane propellants together, and this is what actually
produces all the fire and flame that is in a rocket engine. Traditionally, it would be over
a thousand individual pieces and it would take nine months, but here we're 3D printing
the whole thing in one piece. It takes two weeks and
it costs 10 times less. - One of the big benefits of 3D printing is reducing the number of parts. Have you ever thought about how inside a rockets combustion
chamber, it gets really hot. Up to 3,500 Kelvin. That's hot enough to
melt virtually any metal. So how do the combustion chamber
and rocket nozzle not melt? The answer is they're cooled by passing the cryogenic propellants over them. - On the Space Shuttle main engines. I love to talk about them
because inside those engines, it's hot enough to boil iron. On the outside, you can freeze stuff to the exterior of this
because you're running liquid hydrogen through these things. But to make those, you basically had to take
thousands of very small pipes, and then you would form them into the shape of the combustion
chamber and the nozzle, and then you would braze
weld them together. And this was a ridiculously
labor-intensive task. You would have 1,080 individual
pipes running up the side, all having to be weld together to make the combustion
chamber and the nozzle on the Space Shuttle engines. So you can actually 3D print these things. - This is a rocket
nozzle being 3D printed, and you can see the channels
for the cryogenic propellants being printed right into the single part instead of having to add a
thousand pipes on the outside. Smaller parts like these
are typically 3D printed using metal powder and lasers. - So you can see the cooling channels are all being built as the one piece. So this is a nozzle. It really just lays down a layer of powder that's about a 20th the
thickness of a human hair. So it's really, really fine layers just over and over and
over and over relentlessly for probably about a week or so, and then out comes the rocket nozzle all printed as one piece. It's way cheaper than traditional. And this has four lasers going at once. - That's amazing. - I get asked a lot, well,
aren't 3D printed metals not very strong? Or how can it actually work? But the printed materials are stronger than they would be built
traditionally, actually. It's counterintuitive. - It is. - Because we develop our
own custom alloys in house. So we have a whole material science team just developing our own
alloys for 3D printing, and the fact that it melts
and then cools and solidifies very, very quickly, you can take advantage of that physics principle
to get really strong alloys. - Another major benefit of 3D printing is that it allows for rapid iteration. You can build a part quickly, test it, and then redesign rapidly and print again. - So this is a version of the engine that's about three
years old at this point, but what's amazing is
when you actually look at the engine design today, it looks entirely different than this. So each version we build, we
can iterate and make better. So that's the other, you know, when we say software driven manufacturing, that's really what it is. Since you don't have fixed tooling, all the part geometries
are just controlled via the CAD model, and then
the printers just print direct from file essentially. It means you can actually change
the design extremely fast. So building a whole engine
only takes us about a month. So then a month later, you
can do a better version, and a month later, a
better version than that. So this particular one will actually be, I believe one of the first flight engines that's actually launching to
orbit on our first rocket. - So this tubing, not 3D printed, right? - Not today. - Okay. - In the future versions,
we're actually integrating that into the printed housings. And we're gonna have a way
that that's all printed too. - Perhaps the biggest impact of the 3D printing approach
could be to totally transform what a rocket looks like. With 3D printing,
engineers can build parts that would be impractical or impossible with traditional techniques. Smooth, curvy, bio-inspired designs are just as easy to print
as ordinary structures. - This is actually part of
our next rocket, Terran R. So it's even larger. - This is like the base of a tank. - Yeah. Yeah. So it's gonna go out. It's almost done printing. It's gonna go out about to here. So it's 16 foot diameter,
but it's almost like a shell. - I was gonna say like,
this reminds me of suddenly we're in The Little Mermaid or something. - Yeah, yeah. Yeah, it's just for stiffness though. - It's not that you plan
to make it bio inspired. It's like that structure is
actually the optimal structure. - Yeah, yeah. - We're actually designing
many features in the rocket that could not be manufactured
unless it was 3D printed, which is one of the secret sauces of why you had to build a
whole company around it, is because our rocket actually looks entirely different 3D printed
than it does traditionally. Like in my mind, it's been more akin to like gas internal
combustion engine to electric. You know, really, people
are trying to put batteries and electric motors into
existing products for decades. Like everyone knew electric
vehicles were the future, but Nissan and Ford had
really not compelling products for a long time. It wasn't until a company
came along called Tesla that decided, well, actually
the shift to electrification means the batteries, the
electric motors, the factory, the design of the product, how we're actually
gonna scale the company, the supply chain, all of it's different because of electrification. I mean, that's in some ways, the dirty secret of electric cars and why they're able to
be automated in production because the part count is so much lower. So for a fully 3D printed rocket, we have a hundred times fewer parts, which is what we're guiding to. There's no fixed tooling
in our factory at all. Unlike the rest of aerospace that's still really, 60 years later, even since Apollo building products one at a time by hand
with hundreds of thousands to millions of individual parts. And no one's really changed that paradigm of how an aerospace factory
actually fundamentally works. - Yeah, this is the new
fully 3D printed rocket. So yeah, we'll have dragon
fly wing type structures and we're building it so. But that's the first one, and then that's that one for scale. So yeah, it is definitely bigger. Yeah, so our rocket is named
Terran One and Terran R, and then our 3D printer's Stargate. So all the things here at Relativity are named after StarCraft. So yeah, of course, the Stargate printer was with the Protoss
used to warp in spaceships. And so, that's what's warping
in spaceships at Relativity. We have a system in our
avionics called the Pylon that we have to build a lot of. So we always joke, we have to
construct additional Pylons. Most people don't know how rockets are built traditionally at all anyway. And I think a lot of
people assume it's rockets, so shouldn't it be already very advanced and robots everywhere and you know, Elon's got Space X and Tesla, so doesn't Space X just look like Tesla with all these robots and automation? But that's really not true. I mean, aerospace hasn't
adopted automation at all. - One of the issues, right,
is that you're not making a lot of rockets? - Right? - So there's no incentive
to like figure out how to tool up a factory
to like pump out rockets like a hundred a day or something. - Exactly. - Like you would for cars. - Exactly, you're not making a lot. Even with commercial aircraft, you're not making nearly as many and there's orders of magnitude
more parts and complexity. A commercial aircraft has
several million individual parts. So to have robots assemble
several million parts when an automobile has tens of thousands is completely different. It's a much harder problem. So that's where 3D printing
is automation for aerospace because you're not assembling
all those parts with robots like you would with a car, you're assembling them in the 3D file and then the printer just
prints them assembled. - The plan for Relativity Space, is it low-earth orbit or is
it going further than that? - So for Terran One, it's
mostly low earth orbit. The first rocket. Terran R can actually send
payload to the moon to Mars. I mean, it's pretty, pretty huge. I founded the company
because I really thought that there needed to be, you know, dozens of hundreds of
companies making Mars happen. We're focused on taking
this 3D printing tech and what we call the
factory of the future, and one day shrinking it down to something we'll actually launch to Mars
and build an industrial base. So that's the long-term
vision of the company, is build the industrial base on Mars. In many ways, this factory
is just a prototype. It's still far smaller
than a traditional factory. It's far lighter. And I think it's inevitable someone has to build this company. - I don't know that in 10, 20 years that you will be 3D printing
rockets all the time. Because if you are flying lots of rockets, it becomes cheaper to have
a dedicated machine for it. I do think that as a
company, they are well-placed because even if Terran fails
to capitalize on the market, even if nobody wants to
use it as a launch vehicle, they are clearly now the world experts on 3D printing rocket hardware, 'cause they've done everything, right? They've tried to apply
3D printing to places where a lot of people dismissed it. So I think they're sort
of secure as a company. Whether we will see rockets
being 3D printed all the time? That's a good question. - There've been a lot of talk recently about billionaires going to space. - Yeah. - Will a 3D printed
rocket make it possible and a lot cheaper for me to go to space? - Yes, I mean, certainly what we're doing is lowering the cost. So our rockets are
costing about five times to, you know, I believe we can get to 10 or even a hundred times cheaper with a fully reusable rocket
than what we have today. So it can definitely
climb down the cost curve. But I also think, you know, going to Mars and the first people that are going, it really is about what is the
point of being a human being? Like for me, why go to Mars? If we were having this conversation and a million people were
living on another planet, I think it would expand the possibilities of human experience and what
it means to be a person. Like we'd have YouTube channels on Mars and people sharing what life
on Mars is like versus earth. And there'd be long distance
Amelie, like love stories. Like I think there's
just a lot of richness in what human culture
and society can be about. Yes, I think there's criticism about, you know, billionaires going to space and I don't agree with. You know, all of the projects need to actually add up to
some vision that is meaningful. I think that's really important. But I do think going to
Mars is really just about, you know, we've lived
for generations on earth, so what's it all about like, why do we want to keep
improving and getting better and furthering society on earth? So for me, it's pretty existential. What it means to be a human being. (techy sound effect) - Hey, this video is sponsored by Omaze. Offering you the chance
to win a trip to space. The winner will get two seats
on one of Virgin Galactic's first commercial space flights. Meaning you and a guest
will travel 80 kilometers up in the sky. You'll get to see the
curve of planet earth, experience weightlessness, and become one of the
few people in the world to have gone to space. And you also get to
join Sir Richard Branson for a personal VIP tour
of Spaceport America. The flight is estimated to
take place in early 2022. So enter at omaze.com/veritasium
for your chance to win. Now, Omaze partners with charities, in this case Space for Humanity, an organization whose mission
is to expand access to space and train global leaders
in the eventual goal of creating a sustainable future. So part of your contribution
supports this cause. To potentially win a trip to space and support Space for
Humanity, a great cause, go to omaze.com/veritasium. I will put that link
down in the description. So I wanna thank Omaze
for sponsoring this video, and I wanna thank you for watching.
pumped to see what these guys end up launching. I still don't think the tech scales well to larger launchers or high cadence build/launch but on a component level it's undeniable how useful 3d printing is. I like what Scot Manley says, pretty skeptical fully 3d printed rockets will be flying very much, but these guys will be the experts at overall 3d printing of components and that places them in a decent spot as a company at least IP wise
I don't get why they print the barrel section of the fuel tanks. It just seems slow and inefficient compared to the rings spacex makes. Printing the other parts seems smart though. The giant printer is really cool to watch!
Isn't this basically creating a rocket out of one big layered weld?
A hybrid approach would definitely be more optimal for the rocket but I guess the trade off is no buying extra tooling, way less personnel, fast iteration time and no getting locked into a design. It's interesting, I'm not sold on the concept but I don't think they are doomed to failure either. I hope they succeed.
Also could definitely see them building these on mars. Even if their rockets are sub par they could easily set up shop there and build rockets locally for specific applications. Also like they said not just for rockets but for the industrial base. I'm skeptical but not pessimistic.
Side note: I hate the bumpy texture.
10-15% extra weight seems pretty significant. Just seems like they are going way out of their way to 3D print everything.
I recommend this long Interview with Relativity CEO Tim Ellis. It's is quite old (pre Terran R announcement) but i feel that a lot of it is still very relevant. Very in depth questions.
https://mainenginecutoff.com/podcast/154
Okay the funniest thing in this video, is when he stitched himself nodding in agreance with Scott Manleys video, as if he was in the same room with him.
I'm a bit less convinced by this company now. I think they've over automated to a degree. Printing the tanks seems like it is unnecessary, and he said it was adding 5-10% more mass to the tank. That's a huge amount when you are dealing with really slim margins. Also if their engine takes 1 month to print, they probably are going to need to find a way to cut that time significantly if they really want to rapidly iterate. SpaceX is looking at making 10 Raptor engines every week for comparison. The video didn't indicate how many printers they have, but they'll need dozens to print as fast as SpaceX is making Raptors if each one sits on the printer for a month at a time.
On the otherhand, their expertise in complex 3D printed parts is probably better than anyone else in the industry at this point, and like Scott Manley points out, that has positioned them to be a parts supplier even if their launch business fails.
ULA should give these guys a call and see if they can 3d print an engine compatible with vulcan in a year. It could save the company.