Hacker News Reader

Comment by GuB-42 | Hacker News Reader

This is great but... what are they goig to do with so many Raptors?

SpaceX is following a "fully reusable" strategy, I am not even sure if the BFR booster has an expendable option, Starship may be expendable but it is clearly not the configuration they are pushing for. And yet, SpaceX is producing these engines as if they were consumables.

We only need so many launches, especially super heavy rocket launches before we run out of payload. And if they build as many rockets as they have engines for, where are they going to store them? What about the launch pads? Etc...

I am not criticizing SpaceX strategy, I know that Musk is a fan of assembly lines, for a lot of good reasons, so maybe they are just making so many engines just because they can, or so that they can blow them up later during testing (another SpaceX hallmark), but I am curious about their medium term strategy (i.e. after Starship is operational but before crazy projects like Mars settlements).

View on HN

> We only need so many launches, especially super heavy rocket launches before we run out of payload.

Why do you believe this is true? It's entirely possible that there's a massive amount of unmet demand that's only economically feasible at lower price points that SpaceX is beginning to unlock with reusability and economies of scale.

Even SpaceX themselves is a huge customer with Starlink. An extremely low cost of launch not only means that you can put drastically more satellites up, but you can put them in lower orbit where they will have a shorter lifespan, because they decay orbit and burn up exponentially faster, mitigating the risk of Kessler Syndrome. You can also iterate the satellites much faster as technology advances and waste less time making sure that the satellite is 'perfect' because you only have one shot at launching a JWST after 20 years of work.

On that note, why is one Hubble or JWST enough? We probably need hundreds if not thousands of them if we want a sufficient observation coverage to be fully away of all extraterrestrial (asteroids and commets) threats that might be coming our way.

Starship will certainly enable some new things we could not do before, or that were simply not economically feasible. But even then there is likely still a severe limit to the demand for launching payloads. All those payloads cost serious money, and while you can save a significant amount by removing the necessity for very lightweight construction due to the enormous capacity of Starship, this might only move stuff from "extremely expensive" to "very expensive".

> But even then there is likely still a severe limit to the demand for launching payloads.

The Starship is about changing the approach, not applying state of the art to a changed parameter. Currently there is a limit. After Starship is offered on the market, that limit may significantly change, as its main reason to exist is the high price of the launch services.

> All those payloads cost serious money

Right, but if you, say, are going to send off-the shelf - or, ok, almost off-the-shelf :) - Caterpillar to the Moon, the payload cost suddenly drops. This is exaggerration, of course, but the idea is that payloads may become much - orders of magnitude - cheaper to produce.

If you can drive the price ($ per ton to orbit) down, more demand will be created. Historically and right now the payload cost dwarfs the rocket costs. That will change and be reversed over time.

Some very sophisticated stuff like space telescopes is much more expensive than their launch costs. But indeed, try launching stuff like propellant or shielding and the difference currently is so high its not funny.

> Some very sophisticated stuff like space telescopes is much more expensive than their launch costs.

Today. Space telescopes are much more expensive than their Earth-bound cousins. In a large part because you can't afford failure. With cheap launches you'll send much cheaper telescopes, which will still do the same science - and they are cheaper because they are not made not to fail with very high probability.

Agreed, still even in the future instruments will be likely more expensive than say bunch of liquid H2 for your NERVA engine & which would influence how they are being launched and moved around.

I think the whole long-term point is to be able to "build" fuel in space somehow, either on the moon or in orbit, because then the rocket equation has been effectively solved.

Starships bringing fuel up is a stop-gap.

Maybe, maybe not. Historically speaking, every time we make doing some expensive and rare thing dramatically cheaper or faster, somebody will dream up a way to make use of that new capacity. I'd be more than happy to see what people dream up to do if we manage to make access to space 10x cheaper.

> waste less time making sure that the satellite is 'perfect' because you only have one shot at launching a JWST after 20 years of work.

Yes, this one shot mentality is extremely expensive and limiting.

Blast those probes out there in quantity, and iteratively fix them.

Move fast and break things

Let's not when it comes to human-rated spaceflight, though :-)

Why not? That's exactly what we did when we ventured onto the high seas and across unknown continents. The spirit of adventure and all that. I'm sure you'd have many thousands of volunteers.

Because the risks are different?

Do you have the same attitudes towards helmets whether you're walking on the street or riding a motorcycle? I'm assuming not, because you understand they are different risk profiles.

When risk profiles differ, so do the mitigation strategies. One of the ways spaceflight manages different risk profiles is by implementing a risk-based approach. For example, NASA classifies their software risk based on probability and severity. When risk is high, they leverage more requirements and more oversight as a way of mitigating the risk to an acceptable level.

What you seem to be insinuating is we should be willing to accept more risk. While I agree in principle, the hard part is getting stakeholders (from astronauts to contractors, administrators, politicians, and taxpayers) to agree. They get a say, too.

Agreed, and even if there's not enough actual demand, think of all the extra resources that they can start to stockpile in space for a trip to Mars, etc. by hitching a ride on the same launch as customer payloads.

A full Starship will require 39 engines, and they will likely lose some during development. Even with a raptor a day that is less than 10 Starships a year. Of course if it is fully reusable and doesn't explode too often even 10 per year will be too much production (unless large-scale Mars colonization missions actually happen ). But it will take time, and for the meantime SpaceX does need to build up a fleet of Starships, and that takes a lot of engines.

And that’s exactly the plan for SpaceX, a whole fleet of Starships. Mentioned by SpaceX and Elon Musk.

“Fully reusable” is still a little wiggly. Cars are fully reusable and they still get replaced after 150k miles. Are we going to begrudge someone calling it reusable if they have a MTBF of 10 flights? Or at least, as long as partial failure results in a successful mission but a decommissioned engine or unibody?

You’d use your whole manufacturing capacity then to do 90 missions a year.

There's not a huge number of datapoints yet, but there are several Falcon 9 boosters that have flown >10 times (including two that have flown 14 times.) I'm quite confident in Super Heavy at least handily exceeding 10 flights MTBF.

MTBF is mean time. So “a few over 14” isn’t anomalous if the mean is ~10. In fact it’s a bit low.

"Fully reusable" in all uses I've seen of with respect to aerospace it seems pretty fixed to me. We don't talk about "fully reusable" cars because there isn't an example of a non-reusable car. The "fully" part doesn't refer to how many times it lasts, just that the vehicle is reusable in whole instead of in part. "Wholly reusable" might have been a more semantically correct term, but "fully reusable" is what has stuck.

The Mars plans are medium-term, though, and they need many thousands of Starships+Boosters for that plan to work. In the short-term, SpaceX is 'behind' on launching Starlink, especially v2 which is needed to fulfill the promises of the network and be profitable (and v2 can only launch on Starship).

In the short term, they need hundreds of full stage rockets, and in the medium term, thousands. That's tens of thousands of engines needed in the next ~10 years. I don't think 1/day is fast enough and I'm sure they will increase that rate as soon as possible.

Also for testing, SpaceX doesn't want to be afraid of blowing up rockets and losing engines, so the first few hundred engines will likely not be reused.

That plan will not work, as there is nothing that could possibly be gained by flying thousands of rockets to Mars: there is nothing worth having on Mars.

This is what baffles me about the current space-travel fetish. We know what's up there - basically nothing at all. Even if you go to another planet and mine for raw materials, bringing material back isn't even remotely economically feasible and is unlikely to become so in any of our lifetimes.

Going to space in the 1960s made sense from the perspective of scientific advancement, and let's be honest, showing off. Now, we realize that the world has serious problems - but the people with influence are decidedly un-serious and so would rather continue to show off instead of fixing the mess they helped to create.

> even remotely economically feasible

Depends on what exactly are you mining and whether it is possible to make fuel out there. And with launch cost falling drastically I expect to see this in my lifetime.

Why would you want to bring it back? There are literal gigatons of useful material 'up there', perfectly usable in-place (or anywhere that's not down a well) with low effort. Even with Starship, it's easier to use resources at the destination rather than boost them up.

Because we only care about resources to the extent that they are useful to us, and by definition we are down here and not up there.

Now, it's important to note that there are some wonderful research insights that could be gleaned by landing things on Mars, and also perhaps by using martian resources to launch things further away from Earth. None of these require 1000 rockets landing on Mars though.

Demand for manufactured goods on Mars is at an all-time high, but I fear still not enough to make the economics work.

This kind of thinking is so defeatist - why do anything at all then? And why is it the pro space people who need to stay on earth and save it, what about everyone else?

Human history is progress and this is progress.

“As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas, and land on barbarous coasts.”

― Herman Melville, Moby-Dick or, the Whale

Defeatist? Elon is the one wanting to give up on this planet and start from scratch.

> And why is it the pro space people who need to stay on earth and save it, what about everyone else?

If we limit "pro-space people" to the billionaires sucking up all the oxygen in the room with their dick-waving, the "why do they need to save it" is obvious - they have the money to make a positive difference. Failing that though, if Papa Musk wants to blast himself off to Mars and die of radiation sickness, he's welcome to do so.

Elon's main argument is to become a multi planetary species, he's not arguing leaving earth for good at all.

Money does not magically fix people however much you want to believe that.

People were talking about "the world's problems" in the 60s too. Fortunately we still went to the moon and inspired generations of people to quite literally reach for the stars. The idea that we could simply reallocate the relatively small number of resources currently used for spaceflight, and that would make some meaningful impact in the world's problems is silly. There will always be new problems on Earth to solve.

The resources are symbolic, but the messaging is terrible - "We fucked up this planet, but don't worry because Amazon and Tesla are going to ship us "lucky" people off to their interplanetary labour camps, for the survival of the species."

Only the naysayers message it that way though. And it’s pretty disingenuous messaging at that.

But I guess you made my point for me… your argument has nothing to do about actually fixing anything, it’s entirely about optics.

It's an off-planet ball of rock.

That's tremendously valuable, because it's off planet and out of our gravity well, so we can launch much bigger rockets (measured in interplanetary payload capacity) than we can from earth (though the moon itself may be better, depending on what we can get there).

Launch rockets where, and for what?

If you want to do science, then I wager it is better to do more of the work on Earth and head straight for your destination, avoiding the Mars gravity well entirely.

Huh? We do not know what is up there. That’s the whole point of exploration. We don’t even fully know what our planet has.

The world is and will always be a mess. That’s the nature of human being, we have different ideas, priorities, mindset. We can’t stop progress in an area just to fix some other areas. Besides, it is baffling that space is always singled out when it comes to fixing earth’s issues. Why not says, divert all resources of the video gaming industry to fixing climate; stopping all sporting competition, especially professional ones, for the same goal? That’s of course straw man and unrealistic argument, but space skeptics repeat the same argument on every single space discussion.

This is a matter of perspective.

Elon has said that his motivation is to make humanity a multi planetary species thereby lowering the risk of extinction (asteroid impact, nuclear war, pandemic, biological warfare, catastrophic climate change, what have you).

Personally, I suspect he might just be doing it because he's nerdy and likes the idea.

For all of us who may not agree with his motives, the technological innovation required to make such a project a reality will benefit all of us anyway; so, for my part, I hope that SpaceX/EM succeed in their goals.

> Elon has said that his motivation is to make humanity a multi planetary species thereby lowering the risk of extinction (asteroid impact, nuclear war, pandemic, biological warfare, catastrophic climate change, what have you).

None of those will leave the Earth even a millionth as inhospitable to life as Mars already is - so this motivation makes no logical sense.

Not to mention, with anything resembling current technology, any Mars colony will be highly dependent on Earth for a good few hundred years - there is no way to bootstrap a fully functioning high tech economy any quicker than that, and life on Mars would depend critically on the highest of high-tech (there is no way to live on Mars without computers, for example - so you would need to be able to build, supply and man advanced chip fabs on Mars before you could have a self-sufficient colony).

It also can't be forgotten that there is no proof that a human population can survive on Mars at all, given our current technology. For example, we have no way of knowing yet if humans could successfully reproduce and bear children healthy enough to reproduce themselves in such a low-G environment. We may require genetic engineering to actually be able to live on Mars (or at least human settlements might require huge centrifuges of some kind simulating high-G, which is way beyond our current abilities to do on a planet, or at least increases the costs thousand-fold).

If no one ever starts working on it, it will never be possible. Anything we can do beyond what our bodies allow on their own, someone had to work on.

Building 1000 rockets is not going to solve this problem though; and there are far easier experiments that can be done on mammals in low-G settings before "starting a colony" on Mars.

Not to mention, there are good reasons why the ancient Egyptians did not start work on a rocket program, even though they did want to fly.

Modern computer hardware doesn’t bring us the internet, among other technology advance, but for sure if nobody was working on it in the 1950s, 1960s there wouldn’t be the Internet we all so rely on today. Starships, a lot of them, are prerequisite if we want to have any significant presence in another planet.

> there are good reasons why the ancient Egyptians did not start work on a rocket program.

They didn’t have the required knowledge to do so. The rocket equation, after all, was only established a hundred years ago.

The problem of going to Mars is more or less solved. Advancing on the rocket front is not going to bring us one iota closer to knowing if humans can mate in very low gravity.

Also, computers are useful even if we didn't think of the Internet. 1000 space ships on their way to Mars are not very useful if we can't actually live on Mars.

I think NASA's point is that it's looking for reliability in manufacturing and reliability in the field. Being able to produce AND TEST the engines rapidly will drastically improve both of those fronts.

If you're only able to make 4 a year (as does Rocketdyne, as mentioned in the article) then your ability to respond to failures during testing is abysmal at best. Also your ability to optimize and improve, simplify, harden, etc the manufacturing process and drive costs down also sucks.

Slow iteration for testing and enhancing rocket engines has been the bane of the industry since inception.

As a kid, I read about Goddard and the early days of rocketry, which was a bit wild. Does anyone know of a good history that touches on the factors that contributed to the increasingly slow iterations in the field prior to SpaceX? My fuzzy knowledge of Mercury, Gemini, and Apollo missions suggest that they were fairly fast moving. Was the Space Shuttle era when iteration slowed down?

A lot of things happened. Apollo 1 disaster, end of space race, space shuttle not hitting enough launch volume and having unclear requirements, Challenger and Columbia incidents which skyrocketed shuttle between-flight maintenance, continued failure of NLS>Constellation>SLS.

This seems correct.

Things move fast until something bad happens; then layers and layers of oversight are added to avoid bad things happening again. Years go by without bad things happening and people begin to wonder why we need so much oversight so they begin to peel them back. Rinse and repeat.

In that context, a major mishap is a very big risk to SpaceX because it can erode their efficiency. One of the advantages of the CCP approach is that there is much less oversight, so things can move quicker. I'm not so sure NASA will be okay with that if they lose an astronaut.

If manned space travel becomes as routine as Musk seems to desire, we're going to lose some astronauts.

Read "Ignition!" by John Clark. https://archive.org/details/ignition_201612 . Mostly about search for liquid rocket fuels, but talks about the industry as a whole to some extent.

I have a lingering suspicion it was the astronauts. All of those pilots wanting to be in charge of the ship, in an age where is was hard enough to get a computer to do the work without having to cooperate with a human.

A lot of the stuff SpaceX has been able to do is down to automated systems. Maybe we would have had rockets that could land in the late 80’s if that turf war had gone the other way.

IIRC - Shuttle autoland capability was vetoed by the Astronaut Office. There was some capability within the guidance software for it, but it was never demonstrated (some early Shuttle flights got close to demonstrating it though).

The bigger rockets use many engines, and for NASA applications, may not be coming home. So, deep space / Mars will be littered with raptors in 20 years, I bet.

I doubt it. Raptor isn't a great deep space engine. It's ISP is a good bit lower than hyrdalox engines.

Hydralox is worst of both worlds. To get out of gravity well thrust/weight is key, and Raptor is king at that. In deep space ISP is key, and ion thrusters are way more efficient than hydrolox at that. People don't realize that SpaceX is perhaps the world's largest manufacturer of ion thrusters since they use them on Starlink.

Storage time is also important, and it's likely easier to store long term (Earth-Mars flight time is months) mid-cryogenic LOX and liquid methane rather than deep-cryogenic LH2.

"To get out of gravity well thrust/weight is key" this isn't really true. TWR matters a lot for the first stage, but by the seccond stage, higher ISP and lower TWR is worth it as long as thrust is reasonable. ion thrusters are great for light unmanned probes but anything heavy needs higher thrust and hydralox is key for that. Ion propulsion is also pretty much worthless past jupiter since you can't get enough electricity.

Past Jupiter hydrolox is unusable because of boil-off issues. The best option past Jupiter is nuclear electric, aka ion propulsion.

If you need both thrust & efficiency, a large number of ion thrusters is often a good option.

ISP is like fuel milage, and the economics change drastically when there's infrastructure in place. Having an abundant source of propellant near a celestial body with a much smaller gravity well (Mars) could easily change the economics and practicality of which fuels make sense for different missions.

Put another way, we might one day look at Starship as a big rig semi truck (though the ISP of Raptor is actually really good while a big rig's is very much not). The fuel efficiency isn't really that important.

I think (or hope at least) we're far away from being able to really see how much space access is going to change if Starship actually pulls off what they're trying.

The big difference between ISP and millage is that ISP has exponential effects.

Yes, but the economy of scale makes it much better than alternatives for now. (Cost wise).

Elon keeps hinting about competing with airliners for trans-oceanic flights. My bet is that Starship is going to become the suborbital trans-oceanic machine that shrinks the world in a way not seen since the airliner was invented.

Everyone be nice to Grandma!

She pulled 9Gs out of Khazakstan just so she could see you on Christmas day.

And she has to cruise back in the Morning!

Talk about rocket lag

If there is one thing we desperately need to avoid as a society, it's developing new technology that burns massive amounts of hydrocarbons to make rapid travel anywhere (somewhat) cheap and ubiquitous.

So, basically Concorde, but worse and more restricted. Needs a launch site remote enough to not drive everybody crazy with the noise (Shuttle launches could have been heard dozens of miles away), and a secure trajectory to not kill anybody on the ground if things go bad. Not even talking about reliability, g-loads, and many other things.

You can't jump from Mayfair, London to Manhattan with this system, at which point any flight time improvement becomes negligible.

High speed ferry to offshore platform? That should address both noise and safety.

All the check-in gets handled on shore, then a fifteen minute ride out to the platform on a high speed ferry, 20 minutes to onboard, then 30 minutes to any other offshore platform in the world, followed by a 20 minute offboarding, and another fifteen minute ferry ride.

The actual travel time would be less than the time you're expected to wait in the airport prior to boarding an international flight.

Are there any ferries that can hit 200 km/h? Because that's what it would take to get reasonably far from the city.

Only half of it:

https://en.wikipedia.org/wiki/SR.N4

https://en.wikipedia.org/wiki/HSC_Francisco

Maybe a future revival of https://en.wikipedia.org/wiki/Ground-effect_vehicle ?

Or some shiny airtaxi vtol running on renewables, electricity, handwavium?

Or just a hybrid HTHL vehicle, think SpaceShipTwo with an additional airbreathing engine and some thermal protection instead of a fancy feathered system (not feasible at high velocities). Take off on a quiet engine, get away from cities, make a rocket-powered suborbital jump, vomit in between, re-enter, land quietly across the pond.

(still sounds handwavy and not very realistic, though)

That sounds like Reaction Engines Skylon / Lapcat A2.

Why not aim for the moon? Honestly wish we'd invest in moving a lot of toxic manufacturing processes over to the moon (I'm sure way smarter people than me can figure out a way to make those processes work in that different environment) and slowly turn Earth back into a green, beautiful planet.

Need thousands of Raptor engines over the next 5-10 years.

Payload will not run out, demand is there (Starlink, military etc.)

So Elon Musk is aiming for 1000 starships towards Mars before 2050. 100/year over 10 year period (or something like that).

Each booster uses about 33 Raptors, and each Starship uses 6. I assume the boosters can be reused between launches to some extend. So they might not need 100 boosters.

But you'll need more than 100 ships active; I assume they wouldn't launch 100/year, but rather 200 every 2 years (due to the 26 month launch windows to Mars). Not sure about the return windows and strategy.

That's at least 200 ships (1200 engines) you need active at any given time + boosters + refueling vessels (Each launch will require 6-9 refuels in orbit).

Additionally, there is the whole Earth-to-Earth transport idea (i.e. New York to Tokyo in 50 minutes).

> We only need so many launches, especially super heavy rocket launches before we run out of payload. And if they build as many rockets as they have engines for, where are they going to store them? What about the launch pads? Etc...

The explicit goal, as has been stated several times by SpaceX leadership in the past, is for Starship to eventually cannabalize Falcon 9 and Falcon Heavy launches. They've said they'll keep launching them as long as there is demand for them, but there will be more and more incentive for companies/governments to stop buying Falcon 9 launches. Already SpaceX sells some launch contracts for "arbitrary vehicle" launches where SpaceX can elect to launch the spacecraft on Starship instead of Falcon 9/Falcon Heavy if SpaceX so wishes.

For one thing, they are starting with expendable starships, so they can get Starlink V2 up faster.

My opinion... they are going to drive down costs through economies of scale. Musk and Shotwell want to continue to drive donw the cost of getting payload to orbit.

Asteroid mining is another very profitable and useful option. We are talking about many trillions to be made, many things to be built, and people who get up there first could end up controlling lots of things for the foreseeable centuries.

Think of British East India Company, except the scope is way more than anything we can imagine.

If you are a super power, a project like this is something you must absolutely consider.

What are they going to do with so many Raptors? Launch fleets of hundreds of starships at a time to colonize mars.

In practice, I expect much of the answer will end up being "militarize space a whole bunch more".

I don't think anyone's expecting anyone else to actually abide by all those space-related treaties if there's another Great Powers war. Which, this century, means only one pairing: the US and China.

More surveillance, satellite kill vehicles with perhaps a little plausible deniability, dark, redundant GPS and comm sats, maybe some new, secret constellations at novel orbital heights for when all that debris gets spread around the lower orbits and screws everything up... maybe not actual space-based ground-targeting or anti-ICBM weapons, but then again, countries do cheat on treaties.

I'd encourage anyone bothered by this post (evidently, someone was) to check out the stats on past and upcoming SpaceX contracts. The overwhelming majority are the US military and NASA, and the US military's share isn't far behind NASA. Seems clear to me the smart money is on "same stuff, but more of it" if they bring costs down even farther. And I'd say the US military's got more room in the budget to quickly put together useful payloads than NASA does, if the prices drop, which narrows it down to "mostly more military stuff". And especially if Starship gives them the ability to put large payloads in higher orbits on the cheap. I bet they've got filing cabinets and skunkworks warehouses full of shit they'd love to build/finish and launch as soon as the costs work out.

But maybe I'll be surprised and military launches will remain about the same while StarLink competitors and elementary school students launching cubesats will eat up all that new, cheaper capacity.

Give some to ULA just to taunt Bezos? But seriously, I'm not sure why they aren't looking to sell them to other US rocket companies.

Reply on news.ycombinator.com