Hacker News Reader

Comment by yholio | Hacker News Reader

It's nothing yet except for a statement of intention: we aspire to become the next SpaceX and hope to raise the interest of investors that can't access SpaceX stock. There is no talk about the engines, it's quite clear that the Rutherford won't cut it (Electron already uses 9 of them). So they have to either buy or develop an engine, expensive in time or money.

The mega-constellation sweet spot argument is not really convincing, if you have a reusable mega rocket with a relightable second stage (like SpaceX plans), you can hit multiple planes in a single launch. The 8 ton class is probably just the largest they can afford to plan at this stage.

View on HN

I disagree. For mega constellations it’s about cost per kg to orbit, and for reusable it’s about launch rate. If you build too big, your launch rate isn’t high enough to get the full advantage of reuse.

I think Starship is awesome, but even Falcon 9 is so big it leaves some room for something a bit smaller.

8 tons is in the Soyuz/R7 range, which is what the other megaconstellation is using. And it has had well over 1000 launches and still holds the record (in the 70s and 80s) of the most number of launches per year for a single rocket type, which makes it a good target class for size of a reusable rocket.

8 tons is about where Falcon 9 v1.0 started out, by the way. How many customers wish they had that built the smaller falcon 5.

Additionally, a smaller rocket is useful for some human spaceflight applications. If you only need to launch two astronauts in orbit to fix a satellite, it’d be nice to have a smaller rocket option. An 8 ton first stage reusable rocket is also big enough to eventually have a reusable upper stage and still carry a a couple astronauts, their space suits, a small (Gemini-sized) vehicle (possibly integrated with the reusable upper stage), and some tools and parts to repair the satellite. Launch costs could in principle be less than $1 million, which is about the propellant costs alone for Starship.

Starship is awesome. But we don’t have JUST 757s and 777s. We also have smaller utility aircraft. I’ve long thought that something just under 10 tons, partially or fully reusable, is a good market opportunity. Makes a lot more sense than tiny rockets for megaconstellations.

> Starship is awesome. But we don’t have JUST 757s and 777s. We also have smaller utility aircraft.

This, and the rest of your comment, is quite reasonable.

What it mostly comes down to is $/kg -> orbit. I know Musk constantly over-promises, but there's at least a good chance that Starship/Superheavy will get their total per launch cost below a million dollars. Perhaps well below that. Gwynne Shotwell, someone known for being almost infinitely more realistic than Musk, has stated multiple times that she firmly believes Starship will become a viable on-Earth passenger/cargo carrier.

And that can only happen if a Starship/Superheavy total launch/mission cost is on the order of a long-haul aircraft. That is, well less than a million dollars.

It's also possible that the fundamental physics of Earth surface -> LEO in a fully re-usable way can only be done with large machines. As an analogy: I'm not aware of any 'not large' machines that can transport anything economically across Earth's oceans. The physics, as I understand it, just don't allow it.

Having said all that, I love the hell out of all of the innovation and potential competition coming around in this area. That's the best way to really test the assumptions (many of which I've just stated) that might be holding us back.

An Airbus A380 is the closest comparable aircraft to Starship. It can handle about 82,000 liters of fuel. Until the last 5 years or so, jet fuel fuel was $3/gallon (peaking at $4/gallon) in the US, higher in Europe and Asia. It’d need about 3 or 4 refuelings to travel to the other side of the world and back with about the same payload as Starship. That’s about $1 million worth of fuel.

The rental price to charter an A380 to the other side of the world and back is about $1.3million not counting fuel or the time spent on the ground fueling and going to/from charter location.

So I’d say a couple million per Starship launch would be on the order of a long-haul aircraft price. Still has plenty of room for smaller and cheaper per-launch fully reusable rockets.

Thanks for the specific numbers.

> Still has plenty of room for smaller and cheaper per-launch fully reusable rockets.

Upon further reflection, I'm getting more on board with this, with two big relevant factors: first, how quickly can these new companies work out the fully reusable mojo. SpaceX is clearly many years ahead, but having a predecessor company actually demonstrating a technology surely makes it somewhat easier to re-implement. Second: back to the physics. I don't have a good intuition for this, but I do hope that it's physically possible to efficiently do small scale orbital transport.

One way or another, I'm happy as can be that there are smart people (outside of SpaceX/Blue Origin) really pressing into this problem.

Suborbital point-to-point transport using a StarShip does not require a booster. You only need a booster stage if you need to accelerate to orbital velocities. That brings the cost down dramatically.

A major reason the StarShip is going to be more affordable per launch than a Falcon 9 is that the StarShip is (planned to be) 100% reusable. Falcon 9 (and Electron) always disposes of its 2nd stage.

Hopefully other launch companies can provide 100% reusable alternatives to StarShip just in the name of diversity and planetary capacity. Musk points out that larger ships have an advantage here due to volume vs surface area and the fixed weight cost of avionics.

Ninety nine times out of ten, launching two humans into orbit to fix a satellite is going to cost you a lot more than just launching a replacement satellite.

The number of one-off, irreplaceable, you-have-to-fix-them-if-they-break satellites currently in orbit can be counted on one finger.

As launch costs go down, this is not going to change.

Yes it will. When costs change by orders of magnitude, trades change. Space hardware will always have some non-trivial cost, just like how equipment on the ground still has significant cost in spite of low logistics costs. As reuse lowers cost of space access by orders of magnitude, the relative cost of trashing vs fixing changes dramatically. When the space hardware costs 100x more than the launch, then it makes sense to fix than to trash.

Also, astronauts are MUCH faster at assembly than robots.

Here's a few problems with fixing satellites.

1. From the ground, you have no idea what is wrong with the satellite, whether or not the problem is fixable, and even if it is, what exact tooling/replacement parts need to be brought up.

2. Satellites are not designed to be easily taken apart.

3. Doing any kind of technical work in a spacesuit is incredibly difficult.

The extra cost of building your satellite with sufficient self-diagnostics, with a design that lets someone in a spacesuit take it apart and conduct meaningful repairs on it is going to be paid every single time you launch that satellite. This cost, in both design, and engineering is not trivial.

And the only time you'll ever get any value from it is if you go ahead and do another, even more expensive space launch - this time human-rated. This launch is likely to head on up, and discover that no, they don't have the right tools, replacement parts, or even ability to access the broken parts of the satellite.

The overwhelming cost of a satellite is the R&D that goes into designing it. If space launches are cheap, the correct economics to follow are "Just build two of them, and if the first one breaks, launch the second one."

> Ninety nine times out of ten

I know this was an accident but I'm stealing this.

I do not believe this was an accident. Or am I accidentally misreading..? :p

I really wanted to drive that point home.

99 / 10

> Additionally, a smaller rocket is useful for some human spaceflight applications. If you only need to launch two astronauts in orbit to fix a satellite, it’d be nice to have a smaller rocket option.

That's a nice theory. But that requires an actual capsule that is capable of all of these things. If you want to repair a sat, you need a EVA suit and so on.

Its very unlikely to be cheaper to repair a sat with human, then to simply launch another one, or launch a human controlled robot.

Also, such a capsule does not exist, and if SpaceX is any indication, development of such a capsule would cost 1-2 billion at least, do do it according to NASA spec.

RocketLab does not have that kind of money and I don't see NASA or private financing come up with that.

> Starship is awesome. But we don’t have JUST 757s and 777s. We also have smaller utility aircraft.

While that is true, its not an analog. From Starship once you are in Orbit you can do all that stuff too.

There are already companies like Momentus and others who whole business is to make connecting flights.

RocketLab might be in a good position to be 'second best' and launch stuff that people don't want to put on SpaceX rockets. They will have to compete with New Glenn for that.

I wish them luck.

Where they will get an engine that capable, in that short a time is also really questionable to me. Maybe they are already deep in development, but it doesn't seem like it.

But we do have just airplanes that are 100% reusable. So far, Rocket Lab only talks about reusing the first stage; second-stage reuse is hard, and a huge cost savings.

But maybe until Starship gets the launch rate to reach its potential, there's room for smaller rockets with disposable upper stages. SpaceX is banking on the space industry expanding well beyond megaconstellations, but that will take a few years. By the time we get there, maybe other companies will manage reusable upper stages too.

Right. Having a decent sized reusable first stage on Neutron gives RocketLab the option of a reusable upper stage down the road. It’s pretty tough to justify going for full reuse before you have even accomplished partial reuse. Something similar to the recovery method of Electron’s first stage would be appropriate for a reusable upper stage. (And at one time, SpaceX was planning it for Falcon 9’s upper stage... ballute/heatshield and parachute for Falcon 9 upper stage recovery).

> If you build too big, your launch rate isn’t high enough to get the full advantage of reuse.

But if you throw away second stages, your rate of reuse goes waaay down.

I don't think an 8-ton lift vehicle is going to achieve second stage reuse.

On the contrary, there are MORE viable methods for reuse with a smaller vehicle. They can do mid-air recovery with such an upper stage, for instance, but no hope of that with a Starship-sized upper stage.

More ways to land, but the performance penalty of heat shielding will be crippling on a smaller stage.

I doubt that. SpaceX was going to make Falcon 9 fully reusable, and it wouldn’t have been crippling. Might be too small to launch a full crewed Dragon or 60 Starlinks, but I doubt if its payload would’ve even been halved compared to partial reuse.

>> The mega-constellation sweet spot ... multiple planes in a single launch.

Mega-constellations are exactly the type of project that doesn't need multiple planes. They require multiple sat in each plane, meaning one rocket full of small sats can go to one plane and dump them all. I suspect something has been lost in translation. I think they mean to say that the relightable engines will allow access to different altitudes on a single launch, multiple orbits within a single plane.

Depends on the plane.

Inclination changes are expensive, so if you want different inclinations, it almost always makes sense to use a separate launch vehicle for each inclination you want to target.

However, the longitude of the ascending node (Ω) is perturbed by the earth's oblateness, and the rate of Ω precession is a function of the semi-latus rectum (p) and the inclination (i). If you adjust the orbit altitude to tweak p, you can adjust the precession rate to swing the plane around to where you want it.

This takes a while (months), and you of course need some sort of propulsion on the spacecraft to return you to your target altitude after the maneuver, but if the launch vehicle can drop you into the higher/lower altitude above/below the target plane rather than doing that on-board the spacecraft, the delta-v cost is halved.

I would call that on-orbit maneuvers rather than part of the launch process. This rocket will drop the sat off at the contracted orbit and fly away. Multiple engine firings will mean separate customers can get at best separate altitudes.

Rutherford is really reliable. Why not use like 27 of them? Is there some reason that this is inherently bad?

It would probably be less efficient to use lots of small engines, but perhaps the reduced cost of using an already existing engine that you can mass produce would make up for it.

Rutherford is only a 26 KN engine. An original Merlin 1 engine was 340 KN, so you would need 15 Rutherfords for each Merlin.

A total of 135 engines for a rocket comparable to the original Falcon 9 which they are targeting.

Yeah I guess we tried lots of engines once, though that was a long time ago:

https://en.wikipedia.org/wiki/N1_(rocket)

The Falcon Heavy has 27 engines and works fine.

Its not the number of engines that is the problem.

The problem is how efficient is your engine, how is the Trust-to-Weight ratio.

To do a reusable launcher you need to be really damn efficient on every part of the rocket. Elon explains this pretty well in the talks he gave around 2014.

Tangent: a lot of the rocky planets we have found in the universe are larger than Earth.

With Earth's gravity is seems like we are just barely able to make a reusable launch system by pushing the physics of chemical propulsion to its limits. Better technology can only help so much. We are already close to the limits of physics for things like rocket efficiency, material masses, etc.

Make Earth even say 10% bigger than we'd never be able to build reusable chemical rockets. Make it even bigger still and disposable chemical rockets might be infeasible, making nuclear propulsion the only viable way to get out of your gravity well. A nuclear first stage launching from inside your biosphere is not a great idea, and making nuclear rockets reusable is harder than making chemical rockets reusable. Keep increasing gravity and it only gets exponentially harder still. At some point any kind of space flight become impractical even if you're willing to irradiate yourself.

But... make a planet smaller than Earth and pretty soon you start to have trouble holding onto an atmosphere! Look at what happened to Mars.

This could be yet another Fermi paradox answer. Earth may be right on a knife edge between too small to have a long term stable biosphere and too big to get off the thing!

Of course maybe an intelligence would eventually develop a propulsion system that could crawl out of a super-Earth gravity well using something like nuclear fusion or microwave beamed power, so who knows... it's just another possibility.

> A nuclear first stage launching from inside your biosphere is not a great idea

I disagree. There are nuclear rocket designs that don't have radioactive exhaust.

Sure, its a little bit of a problem if it explodes, but not to bad.

And actually if you do that, then you are much better set up for deep space exploration.

> and making nuclear rockets reusable is harder than making chemical rockets reusable

Not sure that is actually true. I don't see a reason why something like a NERVA engine should be just as reusable as any chemical engine.

We have way to much fear of radiation. We should be already living in the nuclear age.

Given that NERVA was a 1960 engine I don't think that answer works for Fermi, specially given how many more smaller planets there are (we just have not found them).

Read the launch history most of it the engines weren't the problem. *edit control was the problem, it just happens to be connected to parts of the engines.

The N1's first stage had 30 engines, not much more than the Falcon Heavy which has 27

Batteries. Imagine all the weight.

interesting if the Neutron will use electric-pump-fed engines, is this a concept that is scalable to bigger engines?

I read a paper that indicated you can build them to any arbitrary size. Electric feed pump designs don't get worse as they get bigger. But turbo pumps have better performance at large sizes than electric feed pumps.

However the paper also assumed a fixed weigh battery pack. On the second stage electron tosses one of it's battery packs overboard. Also raw performance isn't as critical for the first stage.

One other thing, electric feed has much simpler plumbing than a turbo pump feed.

Ans: Maybe, but I have no idea.

Interesting, do you have a link to that paper? I guess the regular engines that are running inefficient fuel cycles to power the fuel pumps aren't too weight efficient either, due to the inefficiencies of these fuel cycles. Interesting why this approach wasn't tried earlier.

Took a bit of searching but I found it again.

https://www.aacademica.org/hernan.emilio.tacca/9.pdf

If I remember correctly simple turbo pump feed engines have to run using a lot of excess fuel. Simple designs just dump the turbine exhaust overboard. That represents a fair amount of wasted energy. Advanced designs feed the exhaust back into the engine. I think they also scavenge some energy from cooling the nozzle.

thanks!

Reply on news.ycombinator.com