SpaceX’s monstrous, dirt-cheap Starship may transform space travel

She worked really hard on a proposal about how they would go about selling the Falcon rocket to potential customers. Elon took one look and said "That's great. I don't want you to write reports and proposals, I want you to do it.".

My god, I would give anything for a boss like that.

Pretty heavy statement. Which part is the fraud? The timeline or the ability?

I think it's fair to say far-fetched, but the label fraud or even "sounding like fraud" sounds too heavy.

Why some refuse to hold him responsible for his statements is just weird.

I'm not sure if that is a reliable source. (Please don't read that comment as a general commentary about Wiki!)

But how long does this actually last? The amount of people that can afford to pay that is minimal, I dont see space tourism driving any serious markets for an extremely long time (if ever).

And there is a reason point to point transfer isnt being mentioned anymore. The logistics and regulatory hurdles are going to be massive, and I seriously doubt we see that outside of maybe military tests for cargo transport.

But for just $10, we can't really be using fuel in the first place.

Interestingly there are proposed launch systems that, while not being $10, would potentially be under $1000 and transform passenger and freight movement on Earth too. I am of course talking about orbital rings [1].

[1]: https://www.youtube.com/watch?v=LMbI6sk-62E

Seems to me as if Starlink without Starship isn't all that great.

Also, OneWeb - despite being smaller in scope - managed to do a similar thing without reusable rockets. Now that doesn't make F9 bad - it's still a great machine - it's just not the game changer many hoped it to be.

Falcon Heavy is again evidence of this. We're not seeing ride-sharing on FH. We're seeing large (primarily military) payloads that F9 just can't do.

The logistics of such system would be hell, too - Concorde only had limited routes due to noise. A rocket of Starship's size is another thing entirely. You'd basically have to launch and land it offshore so the logistics getting to and from the launch site and into the vehicle would kill all benefits.

Not to mention the simple things like weather, closing airspace, and the lack of cross-range capabilities, which would make emergency landings or evading bad conditions on the landing site pretty difficult.

LOL, not even in the same order of magnitude ...

Unlike planes, they carry oxidiser and fuel, so they do not burn, they detonate. Minimal safe exclusion zone is measured in square miles!

Simpler? The fuel is chilled 2/3 of the way to absolute zero, and the engines are at several thousand degrees. Tiny manufacturing error and everyone dies.

Here's another way of putting that: SpaceX has designed it so F9 and Starship payloads will be interchangeable. That's not quite the same thing. You can't just take an F9 payload and bolt it on top of an Ariane rocket.

> Starship is being designed with the intent that it will cannibalize all Falcon 9 launches eventually.

Sure but that's just going to take a long time, particularly if F9 is, say, sub-$15m per launch. There's a lot that still has to be done for Starship (eg the landing system, the crew component).

The limitation there is capable RTGs, not launch vehicles.

> 2. a much cheaper Webb telescope

Webb couldn't have fit into a Starship fairing either.

> 3. build a base on the moon

For whom? Launching stuff is one thing, but you'd need to design, build, maintain, and supply such station. If you think the ISS was expensive, a Moon base would dwarf the ISS costs. Even if launches would be cheaper, there's still a lot of logistics and support behind such station, which is one of the reasons no one did it yet.

> 4. make a mission to Mars possible

That could be done without Starship - as much as I think Zubrin has drifted into the nut job corner over the past decade, he still has a point: crewed Mars missions would've been possible since the 1980s.

One of the things I'd dearly love to see in lifetime is a return to the ice giants (ie Uranus and Neptune). We've only visited once and it was in the 1980s with the limited tech of Voyager 1 and 2. Think of what we could learn with a Cassini-level mission and Cassini itself is now ~14 years old.

But these missions take a long time to develop. We actually have a launch window coming up in the early to mid 2030s where we could feasibly reach the ice giants but the development period is so long and no proposals are at an advanced enough stage that we've likely missed it (which makes me sad).

> 2. a much cheaper Webb telescope

Would we make a cheaper version? Or would we take the opportunity to make an even bigger version? I suspect the latter. And JWST took 20 years to develop.

> 3. build a base on the moon

I actually believe we'll see a man on the Moon in the next 20 years. It could be much sooner but these are still government programs subject to change at any moment. I mean think where we could be if the SSC had been completed.

Apollo was driven by Cold War era competition. Our appetite for risk has changed. This means we may not see a Moon mission until rescue in the event of a catastrophe is actually possible. That will delay things.

> 4. make a mission to Mars possible

There's a thread about this from earlier today. IMHO a manned mission to Mars makes almost zero sense. Despite Musk's wildly infeasible and impractical predictions about Mars colonization, I wouldn't be surprised if we don't see men on Mars before 2050.

All of the things you've listed here though aren't really commercial enterprises. Or at least they're far from it. They're flagship missions by governments because they're so wildly expensive. This creates some demand but is it sufficient to justify the program? I don't know.

I do suspect we need to find commercial enterprises to justify the Starship economics and I'm just not sure what that looks like yet.

Why? If you have to spend almost 10M$~ on an expensive high tech upper stage you your just gone throw in the ocean its clearly give you a price lower bound.

> it's not as if 100t to LEO is novel or anything

Starship does 150t in its early version while being fully reusable.

> The Soviets and later Roskosmos abandoned the vehicle, because they couldn't find a use for it either

The Soviet union collapsed and they didn't have the money for it.

https://caseyhandmer.wordpress.com/2019/08/27/there-are-no-k...

The equivalent rocket is the Starship, which clearly can be built.

Technologies like 3D printing, new alloys, better fuels, modern computers, etc... have made dramatic differences to the design and construction of these rockets.

Have a look at the circuitry used on the Saturn V: http://www.righto.com/2020/04/a-circuit-board-from-saturn-v-...

Especially the "instrument unit": http://static.righto.com/images/lvda/IU-labeled.jpg

It's huuuuuuge! The computer power of that entire thing could fit into a single chip these days. Trying to reproduce something like that 1:1 in the modern age would be just absurd.

Note that the Saturn V design isn't just cost-inefficient because technology has moved on. It wasn't cost effective back then either! The per-launch costs were over $1B in inflation adjusted dollars, putting it squarely into the same ballpark has the SLS, which is a monstrously expensive pork-barrel project designed primarily to burn money, not rocket fuel.

We can't build an authentic Great Pyramid, 1950s supercar, or 1970's minicomputer, either. At least, not for anything resembling cost-effectiveness compared to the current way to do things.

Who are "they", and how can that opinion hold true when we are in fact building something comparable to it?

That doesn’t mean we can build a 100% copy, though. Apparently, some of the welding requires techniques that have been replaced by better methods, and welders who can replicate the original do not exist. I guess that’s solvable by using robots, but if, e.g., 3D printing or even glueing works better, and the goal would be to get a rocket that size again, why bother?

There's a planetarium nearby with a cool history made from the dome of an oxygen tank. UAH has a special collections room dedicated to the Apollo program. There are some other rockets on the side of the interstate near the outskirts to let you know you're entering rocket city.

If you plan carefully, you can see everything in a day.

So there was no barrier between the vehicle and visitors. You could walk up and touch it, which I did. Then I walked to the engine end of the first stage and climbed up between two of the F1 engines. I was able to climb into the center F1 and stand up at the end of the expansion bell.

I didn't even come up to the center of the bell. It drove home the titanic size of the machine in a way that nothing else had.

I had to climb down when a moderately pissed-off guard saw me after a few minutes. It was a wonderful experience.

They have nothing near the size of a Saturn V. NY HoS has:

Gemini-Titan II

Mercury-Atlas D Rocket

Mercury-Atlas D “Friendship 7” Replica Capsule

Mercury Capsule

Saturn V F-1 Engine

For perspective, the Saturn V is about 20 times the size of Titan II (by mass), and could almost deliver the whole Titan II into low Earth orbit (LEO).

Anybody that has watched the Russian space program knows this is nonsense. They have barley enough money to continue to fly their 50 year old rockets. Their supposed next generation rocket has barley ever flown.

Since that article they have already announced like 3 other rockets as well.

Russian space program is 50 year old soviet tech and a lot of political propaganda.

They always have the option of designing another larger engine later.

It also gives you engine out capability and also the ability to effectively throttle the rocket in a much more fine grained way by shutting off engines.

No, that's completely wrong. There are challenges in running a bunch of engines, but it's critical to pull off for reusability. All rocket engines have a strict floor on how low it's possible to throttle them. Even 40-50% is quite hard, and I believe the Raptor is 40% though with the goal of getting somewhat lower. And for old space expendable rockets that was plenty.

But to land a rocket, you necessarily need to be dealing with bringing it down when it's nearly empty. Rockets are mostly fuel by mass, and when that's expended are very light. The thrust necessary to lift a fully fueled rocket up is massively ginormously more then would be needed for its empty weight. So if you light up at full power even for a moment it's going to fly right back up again. You want a low enough minimum thrust/weight ratio (TWR), ideally below 1 (so you can gradually descend/hover), though a rocket can get by with a bit above with a "hover slam approach" approach like the F9 uses. But it can't be that far above. But you can't just throttle a rocket engine down to 1% either. What to do?

The answer is multiple engines. Throttle is limited, but you can always turn an engine off. So if you have 10 engines which each can throttle to 50%, your effective ability to throttle from max thrust is .5/10 or 5%. With 30 engines at 40% would be .4/30 = 1.3%. Very helpful. Also means more potential redundancy for landing.

There are challenges with plumbing, harmonics, control etc which are real, so as always with rockets complex tradeoffs, and big engines have their own issues. Small engines are also easier gimble, another valuable thing for landing, and to mass manufacturer which is a whole extra aspect for SpaceX, Starship is about economics remember. But there are many solid technical reasons for the choice.

The only downside is that if any of the engines explodes catastrophically, it will take the entire rocket down. This used to be a big problem in the Soviet N-1 lunar rocket, but contemporary technology can mitigate it quite a bit: with enough telemetry and computer power, you can determine that the engine is acting funny before it explodes, and you can switch it off. Losing one engine out of 30 isn't fatal.

Also, with computer modeling, a lot of the engine design and testing process takes place on screen, so many failure modes are debugged early.

SpaceX haven't, yet.

> As I recall, the Soviet moon rocket failed because it had a cluster of engines that they could never get to work together. The Saturn V had a cluster of only 5, and barely were able to get that to work.

N-1 rocket had many troubles with the engines, but it wasn't specifically due to the number of engines: the NK-15 engine couldn't be test fired, vs the Raptor definitely can. You can read more about the four failures here[1].

Similarly, the biggest issues with the F1 engine (powered the first stage of the rocket) was with the injector plates: there was a lot of combustion instability due to uneven mixing in the giant combustion chamber. The Raptor, by virtue of injecting both fuel and oxidizer in gaseous form, will mix better (and it has much smaller combustion chamber anyway)

Having two launch pads (they're starting work on a second tower at the Cape) is also a big help, as they would have backup launch site in case there's an explosion.

[1] https://en.wikipedia.org/wiki/N1_%28rocket%29#Launch_history

Soviets were desperate by then, no money, no time, and political pressure.

> How did Musk solve this problem?

Raptor 2 apparently has melting issues, you can only imagine what will happen when you have whole array of them firing.

The N-1 failed partly because the engine were untested and had high failure rates. The could not be tested on the ground and had a lot of manufacturing issues.

The N-1 failed because the on board computers couldn't handle engine shut downs very well. They basically just shut down the opposite engine.

In some flights that problem lead to complex piping being broken as well, but its by far not the reason the rocket failed.

And SpaceX has already done the Falcon Heavy with 27 engine and its perfectly fine.

What matters is that your engine work well and that if something happens to one engine the computer can figure that out and perfectly adjust all other engines to that new reality.

Would be great to learn about the questions they ask.

SpaceX in its whole history has shown very, very little CGI. Specially compared to many other traditional rocket companies.

> Orbital refueling - no technical details, just cgi images

We know for a fact that NASA was very, very impressed with SpaceX work on orbital refueling. In their bid for the moon mission they got a special shout out from NASA because their approach to it was so much further developed then anybody else.

They also have a contract with NASA about demonstrating cryo fuel transfer in space.

Have you ever consider that just because they are sharing everything with you personally doesn't mean they haven't thought about it?

> crew capacity - no details > ship layout - no details

Why would they provide details? The crew capacity is not a fixed thing. Depending on the mission the internals will be different. We know what NASA is planning for the moon. In that version there will be very tiny crew.

Likely a somewhat larger crew for the Polaris program. And probably somewhat more for the DearMoon project.

Compared to making a fully reusable super heavy this is peanuts.

> life support - ummm scaled up dargon??

Do you have anything to say other then '??'? Yes, those are machines exist and if you have a larger inside you need to add more of them. This is totally possible and not a '??'. If you are not constraint by weight then doing that is totally reasonable.

And NASA own evaluation of the HLS bid also thought it was reasonable.

Its not reasonable for a Mars flight, but that by far not one of the first missions, they have lots of time to continue to work on that technology.

> raptor 2 - 'its stripped raptor 1 and much better' (still in tests, not proven tech)

They have produced many more of them that many other rocket engine program produced in their hole history. They have test-fired them many, many, many times and they have done full duration test fires as well.

> refuealing and relaunching withing few hours - Falcon 9 takes 30days!

You can't just look at Falcon 9 and say 'this proves it'. That's nonsense. They have many Falcon 9s, why would they try to rush a landed booster into service when they have 10 others waiting.

Falcon 9 often lands in the ocean and it takes a week+ until they even get it back.

> Also someone made a super cut of his 2022 update talk and its beat for beat the same thing he said in update from 2 years ago. Its not a good sign if you havent done anything in 2 years for a project that promises so much in few years time.

Of course and he talked about the same thing in 2016 as well. Because the talks are not primarily about technical details.

Also, your comment is just plane nonsense. We have cameras on their production and launch site literally every day. If anybody thinks that they have not made progress in 2 years then they have not been paying attention and are just deliberately making disingenuous argument.

That said, I don't see passenger rockets being a thing for quite some time.

I first flew across the Atlantic as a baby in a propeller airplane. The last time I crossed, it was in a 757, the airplane I worked on 40 years previously.

The big advances in my lifetime were pretty much all in computers. Ironically, that was the one thing not predicted by any of the endless scifi books I read.

If it's so easy why doesn't everyone do it?

I know it’s the trend in the world, the established media, social media, etc. but it’s possible have have a more nuanced view on things than ALL VERY BAD or AMAZING AWESOME.

For example, one can reasonably criticise Musk for his stupid outbursts on Twitter and elsewhere (and of course, also criticise the media for breathlessly reporting every bit of nonsense he spouts for clicks.) He is notoriously bad at predicting the timing of when things will happen. Robotaxis are probably 5+ years away, if ever. Tesla delivers their new models much later than promised, and has poor customer service. Hyperloop was a concept that would probably have been ignored were it not for the outsize attention that Musk gets. Etc.

But… these views can reasonably coexist with also acknowledging that two of his companies have revolutionised their parts of our world within a very short space of time. SpaceX has revolutionised the commercial rocket space, and Tesla has driven the whole automotive industry towards building EVs by leading from the front. And it’s almost guaranteed that neither of those would have happened without them being led by Musk (and Shotwell).

> Hyperloop

What lie? Have you actually done any research.

Musk said he worked on something, he is not gone develop it as he has no time, but if somebody else wants to, they can. And to help that he opened up this BluePaper:

https://www.tesla.com/sites/default/files/blog_images/hyperl...

That's literally it. No produce was ever promised. The only thing that was promised was to deliver a BluePaper and they did.

> Semi

A produce they are delivering in slow numbers and there is no question that they will deliver more. Its a question of battery availability and business strategy.

> VTOL with pressurized air!

Again not true. You keep mixing up things talked about and things promised.

For the new Roadster they said they would work on a SpaceX package that would be able to for a small time hover a vehicle. They have neither sold that package only talked about that they are working on it.

> Pedo guy!

So guy was an asshole to him, and then he was an asshole back. They are both not the nicest people. So what.

> FSD! Robotaxis!

Like many others he was wrong in their prediction. Like many others were on the same topic. Ok, fair. But the rest of your comment is nonsense.

> And abusing their workers?

More nonsense. Tesla workers with stock options have done far better then any autoworkers. Many SpaceX engineers retired as millionaires. Overall working at SpaceX its highly likely that you would have made more then comparable engineers specially of comparable seniority.

Go look at any analysis of launch prices over the last 20 years, they have gone down.

SpaceX came in with a very aggressive price and they have lowered it a bit since. However since they have no competition they have increased their own margin rather then lower the price further.

That has made them a highly valuable company capable of investing in next generation.

> And reusability has its own problems, like lowering the amount of payload by 25% (you need to save fuel for re-entry).

That argument is equally dumb for all other types of transportation as it is for rockets.

> You can maybe get 10-20% savings if you ever get it working (see previous failed attempts...).

Complete nonsense. Hardware on a large rockets is above 80% of the cost and large fixed cost can be improved also by higher flight rates that re-usability make possible.

Anybody that has studied that space seriously understand that massively lower cost are possible with full rapid re-usability.

I'm pretty sure any organization that's able to figure out such launch vehicle will be able to do their own due diligence.

Shuttle launches were about $450M per launch, with 7 people and 50,000lbs of cargo. Came to about $10,000 per kg, vs SpaceX Falcon 9 at $27,000 [edit] and the Dragon around $9,000.

[1] https://en.wikipedia.org/wiki/Soyuz-2

Cargo flights are different, and you should also look at total program costs from the deal with NASA. The NASA contract [1] was for ~12 launchs, minimum 20T of cargo for $1.6B. That puts the total program cost at $80,000 per kg. The space shuttle program costs were about $60,000 per kg.

But don't take my word for it, here's a quote from a NASA scientist saying their payload cost per pound went up. [2]

  "My cost per pound went up with these rockets," Margasahayam told Tech Insider. "On the shuttle, it would be much less."

[1] https://www.nasa.gov/offices/c3po/home/CRS-Announcement-Dec-...

[2] https://www.businessinsider.com/spacex-rocket-cargo-price-by...

> So far, the average costs of launching cargo remain on par with the space shuttle at about $US30,000 per lb. (The space shuttle cost about $US1.5 billion per mission, including development, and could carry up to 50,000 lbs of cargo.)

Also not to mention that the Wikipedia table is probably irrelevant, because for the Space Shuttle launch cost is just the total cost of the entire program (including R&D and everything else) divided by how much it actually carried, while SpaceX figures are literally just Elon's marketing material. Look at the citations.

> The development of commercial launch systems has substantially reduced the cost of space launch. NASA’s space shuttle had a cost of about $1.5 billion to launch 27,500 kg to Low Earth Orbit (LEO), $54,500/kg. SpaceX’s Falcon 9 now advertises a cost of $62 million to launch 22,800 kg to LEO, $2,720/kg.

The fact that SpaceX advertises that figure doesn't make it true.

There is usually 1-2 order of magnitude cost difference between vehicle price and the gas in the tank.

There are tons of possibilities; James Web 2 with 10x the diameter, a probe to catch up with Oumaumau, real asteroid prospecting. Oh and orbital hotels. Does anyone know of any concrete projects?

Toilet research.

Soyuz is the only rocket that has flown much more than 100 flights, and it has a reliability of 97%.

Several other rockets, including Falcon 9, have flown enough successful flights that statistically their reliability is unlikely to be less than 99%, but we cannot know that it's any higher than that.

NASA figures that the failure rate of a Dragon flight is about 1 in 500, but that's based on analysis rather than statistics and is likely quite wrong. Note that the highest chance of failure in their analysis is a micrometeorite collision rather than a rocket failure.