46 thoughts on “Starship”

  1. When Starship was first proposed, I said something that’s still not widely understood: The instant it puts its upper stage (Starship prototype) into low earth orbit, that’s the end of everything else. Starship can distigrate on reentry. SuperHeavy can explode right after the second stage begins independent flight. It doesn’t matter, because SpaceX has just demonstrated an expendable launch vehicle that can put upwards of 300 tons in low earth orbit, and which costs no more than $50mln to build, per vehicle. Likely a whole lot less, once Raptor is in mass production. And if reusablility succeeds? Just imagine.

    1. I guess I am a DOS-using flip-phone-wearing late-1990’s car driving dinosaur.

      Is Starship a better path than mass producing Falcon Heavy and developing some manner of on-orbit tanking?

      What is the SpaceX Special Sauce that they can build a vehicle that size, expendable or otherwise, for 50mln/launch?

      Do they employ Willy Wonka’s Oompa Loompas as low-wage engineering labor? Do they use whips?

      Whenever you saw a chart with a vehicle that appears larger than a Saturn V, the accepted wisdom has always been, “Ooooo, that looks expensive!”

      If someone says, “Look what Elon has accomplished with Tesla”, yes, what has Tesla accomplished? Apart from government mandates and the desire to engage in green virtue signaling, what does a Tesla automobile accomplish that everyone is going to quit driving their clunkers and get one of those puppies?

      If they are really going to put 300 tons in low earth orbit for a cool 50mln — less than $100/lb, bay-bee — that accomplishment would make the Tesla automobile seem like a toy.

      1. Don’t forget, fifty million might be what they cost to make but that isn’t the price. It might make sense to do some expendable missions to retire early iterations of the vehicle but they are probably amortizing profits and not just costs, meaning that an expendable launch would have to realize some percentage of expected lifetime profits.

        Just some random numbers. It is like making $100 million off something instead of $1 billion. They might accept getting money quicker as a good tradeoff over letting the vehicle go its full life but likely price it accordingly, time value of money and all that. All of these tradeoffs will change as the launcher matures.

        1. So this thing is like the California High Speed Rail, that was supposed to cost 10 billion, give or take at some point in history, and they will be lucky to get ‘er done for less than 100?

          It is also like California High Speed Rail in that its proponents talk about a 220 MPH top speed, which by itself is at the upper range of what countries with a lot of experience with this kind of thing achieve and even China is backing off on speed because reasons? And then they project travel times that don’t account for acceleration and braking, multiple stations stops, reduced speeds approaching city centers?

          1. There was a college math professor that I knew who, supposedly, went up to the University’s first computer bank in the punch-card days and…
            …wanted to see what answer it would give if you “divided by zero”. The result was the Magic Blue Smoke™ and a call to the company tech representative, or so the story goes.

            That is what I feel about the SLS — if it works it will be a kludged-up hot mess but only IF it works. Oh, the tech rep ‘patched’ the computer code so it would automatically abort “divide by zero” calls in the future. A fix far simpler and cheaper than the SLS would need.

          2. @ Paul Milenkovic;

            California’s High Speed Rail is actually quite realistic in its speed projections as long as a few minor operational assumptions (based on regular rail) are changed.

            The first, as you point out, is the planned speed does not take into account the time of slowing to approach a station, time at the platform, and acceleration time. The obvious answer is to not do that. The planned speed can easily be obtained via not slowinhg down or stopping for stations; simply pass through at 220mph, ejecting the passengers as you do so. Embarking passengers can be swept up in scoops and pumped into the passenger compartment of the train.

            Likewise, maintaining 220 mph on the many stretches of the route winding through mountains is certainly achievable, so long as you use a track design that allows for the train pulling upwards of 20 Gs in the turns.

            If you simply make the above operational assumptions, the projected timetable makes perfect sense.

          3. California High Speed Rail (AKA the Fresno Municipal Subway) is a government project made super-extra-double-secret inflexible by mandating an impossible and mutually contradictory set of goals into law by referendum. SuperHeavy/Starship is a private project with a single decision-maker and absolutely no constraints except his own preferences. He has made substantial changes and implemented them overnight. SpaceX has delivered on every capability and price goal to date. The two projects are pretty much at opposite ends of the engineering project management spectrum.

    2. Starship removes the launch bottleneck for now. Imagine the possibilities?

      For one, anyone who continues to focus on the launch bottleneck will be spinning their wheels while anyone who goes through the USA will be able to pursue their desires in space. Most countries, and many many companies, are capable of doing things in space and doing things in space is more important than how you get there.

      There is a role for government to play but they will likely mess that up and focus on things they shouldn’t while ignoring things they should, which could still work out OK assuming Americans retain American culture.

  2. The special sauce is economy of scale.

    When you build just one copy, and it has to work perfectly the first time, and the operating margins are low so you’ve got to finess the launch, it’s always expensive.

    When you mass-produce the item, you get savings. You get multiple chances to look for defects and shortcomings, and fix them.

    And when you go big, mass forgives a lot of things. A half-scale Starship wouldn’t work (look at the abandoned efforts to recover the Falcon 9 second stage).

    Anyone could have done this. The key is willingness to be bold and go large — production and size.

    1. Except that BO could mass produce New Glenn and not see any savings. SpaceX had production in mind from the beginning and set up their system to support that. BO would have to start from scratch.

      1. Jarvis is a big step in that direction, a stainess steel methalox second stage for New Glenn to replace the “existing” hydrolox expendable. If Jarvis works, expect a stainless steel first stage.

  3. As Elon says about the activity at Boca Chica: “We are not building Starships, we are building a *production line* for Starships”.

    2. I truly hope that resilience against hurricanes was factored into the infrastructure at Boca Chica. Lots of resilience.

        1. The cheapest buildings at Boca Chica are almost certainly the Sprung Structures “tents” used for component fabrication. Those all meet the Dade County, FL hurricane-resistant building code. “Tents” exactly like those in use at Boca Chica have handily survived actual hurricanes. Contrast this with the significant hurricane damage suffered by more conventional factory buildings such as those at Michoud.

          But I think NavyNuke is likely right about the relative paucity of hurricanes in the area. The homes in Boca Chica Village and a few other pre-existing structures have been standing for decades and don’t look to be especially robustly constructed compared to the SpaceX production and launch infrastructure.

          1. Yes, those tents have a much better aerodynamic shape presented to the wind, and if properly anchored will ride out a storm very nicely. Overhanging eaves and slabsided structures tend to go a-roaming when the anemometers die.

  4. “we could actually build and launch a base that could house 1000 people in a year or two. We probably won’t, but we could.”

    We shouldn’t because that would put a lot of resources in places that we might not want them a month or year later. It would be great to roll out a prospecting suite at many different sites that would lead to larger bases sometime later.

    “Even if the space industry fully understood Starship, I think it would be very difficult for them to plan and adapt rapidly enough to match the coming explosion in launch capacity.”

    The concept the author notes but doesn’t name is Min/Max. It isn’t that min maxing would disappear with Starship but that it becomes less critical and what you are min maxing would change. I think JPL would look at the capability and min/max the f out of it just like they do now and that would be just fine for them as specialists as long as everything that comes out of government isn’t the same way.

    I think that people do grasp Starship’s capabilities but understanding them isn’t the same as building up just in time payloads to use those capabilities to its fullest prior to seeing the vehicle in action with near term payloads. For example, John Deere wouldn’t make a Moon Tractor before anyone has returned to the Moon.

    It will be exciting to see what people come up with for payloads and while the author worries about NASA and the government not being a controlling partner in everything, I think that is a bonus. The government is incapable of the creativity and knowledge needed to plan the future.

  5. I think Starship will create competition.
    {and Musk and SpaceX will do better with competition}
    But we still have the question of whether Starship works.
    And we still have the question of whether the Moon has mineable
    water. If the Moon has mineable water, it would be good news for
    SpaceX, but it’s also competition.
    Other than the reality of what on the Moon, the major problem with lunar water mining remains the lack of market for lunar rocket fuel.
    If one can mine lunar water, the moon is viable. It’s a lot easier to get to than Mars. But minable lunar water doesn’t mean lunar settlements, Mars should be first place to have towns.
    But Mars needs towns on it, the Moon does not need towns.
    We don’t know, yet, if Starship is viable. Nor that the Moon is viable, nor that Mars is viable.
    One thing about viability of Mars, is effect of Mars lower gravity upon on life.
    Another thing, is viability of a town on Mars, is the question, are there better place on Mars for a town. Or same problem as lunar water mining, how can know if there is not a much better site to mine lunar water.
    Is lunar south pole better than north pole?
    What lunar water mining going to do, is cause a lot more exploration of Moon. What Mars settlement going to do is cause a lot Mars exploration.

    Mars settlements WILL use the Venus orbit- using Venus orbit gives twice the launch windows to Mars.
    Both the Moon and Venus orbit, will make Mars more viable. And of course Mars settlements will make the Moon and Venus orbit, more viable.
    Back to Starship, it will generate a larger rocket launch market- and larger rocket market, will cause different rockets to be developed. And if enough market for rocket launches, more spaceports will be made.
    It seems possible, that some country steals Musk.
    Or some countries encourages creation of other Musks.
    What seems obvious is no one {nation} is spending enough money on Space. Probably this will change within next 10 years.
    It’s the lull before the storm.
    Will FAA screw this up. Will starship land more or less intact.
    Either could add years to this.
    It seems a lot momentum has being added, but that space agencies are in stand still is “normal”.
    A problem or question is when does starship prototype launch and how many people are going go there to watch it.
    How many people are going use their boats to watch it?
    Could a huge boat parade madness, delay it?

    1. Since we don’t know how to build structures in low gravity, it would be nice to have some place nearby to figure it out and practice on, before having to “learn on the job” millions of miles away. How do you move kilotons of dirt and rock out of the way for your mine, or for your underground habitat, when the mass and the intertia are different from your 1G excavation experience? Play on the Moon, eventually live on Mars.

      1. When I was thinking about my Martian Homestead idea, I was surprised to find out there’s a full suite of existing electrically powered heavy equipment (backhoes and cranes, etc.) that could be converted for use on Mars. and probably the Moon. Plus the skate from a Tesla semi could be adapted into a bulldozer. The Homestead was based around buried Atlas shelter components, quonset huts on the surface (filled with dry nitrogen to keep the dust at bay) and various modified Tesla vehicles. The biggest problems were ECLSS. Mining rebreather cabinets and marine oxygen generators, plus perchlorate candles were all I could think of.

  6. The money quote: “A dollar spent on mass optimization no longer buys a dollar saved on launch cost. It buys nothing. It is time to raise the scope of our ambition and think much bigger.”

    1. While I don’t recall the specific mission, I recall reading many years ago of a real world example of your point. NASA was developing a deep space mission and the chosen launch vehicle was a Delta (probably a Delta II or even a predecessor). For some reason, the mission was switched to an Atlas Centaur. The savings from not having to optimize so much for the lower capability booster more than made up for the higher launch costs.

      High launch costs drive up the costs of everything. Because launches were so expensive, the payloads had to last as long as possible while being as light as possible. That drove the need for exquisitely engineered components with a lot of redundancy, made from expensive, lightweight materials. Because the payloads were so expensive, that drove up the reliability related costs (including the bureaucratic processes) of both the satellites and the launchers. Just like how it takes propellant to accelerate propellant, cost drives cost.

  7. “SpaceX has just demonstrated an expendable launch vehicle that can put upwards of 300 tons in low earth orbit..”

    Believe it is actually >100 tons in LEO:

    “Starship is designed to be able to launch bulk cargo into LEO in >100 T chunks for <$10m per launch, and up to thousands of launches per year."

    This is most interesting:

    "Starship is intended to be able to transport a million tonnes of cargo to the surface of Mars in just ten launch windows, in addition to serving other incidental destinations, such as maintaining the Starlink constellation or building a big base at the Lunar south pole."

    Suppose if you have "up to thousands of launches per year" you could launch into orbit 24/7 just park payload in orbit until launch window to Mars opens up every ~26 months fuel it in orbit (in orbit refueling) and basically like a caravan to Mars of many ships when said window opens.

    1. Tim, 100 tons is the LEO payload of the fully reusable Starship. The 300ish tons figure I used is for the fully expendable version, something Musk called Starkicker, with all the recovery hardware removed, including some unnecessary engines. My point was, if Starship/Superheavy fails, you’ve still got a gigantic, cheap rocket. More below.

  8. I am 71 years old and grew up in Leave It to Beaver America. While I find OldSpace bleats, in which SpaceX plans are called lies and the fantasies of “fanbois,” the truth is, OldSpace is part of the Military-Industrial Complex Eisenhower warned us about, and there afre a lot of people with much to be ashamed of…

    Secret sauce and oompa-loompas:

    The reason Starship isn’t comparable to “mass-produced Falcon Heavy” is because FH is made from expensive aluminum-lithium alloy machined and shaped using expensive stir-friction welding. Starship is made from cheap commercially sourced stainless steel sheet metal rolls, assembled out of doors with conventional welding techniques. In addition, FH uses 28 Merlin engines that cost twice as much as the 32 Raptor engines on Starship/Superheavy. A single FH costs $200 million and has 20% the payload of expendable Starkicker (as Musk called it). That means for the same throw weight, FH will cost 20x Starkicker. Starting to see the shape of things to come?

    As far as workforce goes, SpaceX culls the bottom 5% of its workforce every year, while hiring only the best engineers and techs it can find. Glass Door is full of bitter complaints from the culled. It also means that all the other aerospace companies are only getting the dregs now, people for whom “work-life balance” is more important than the future of humanity in space. The SpaceX workforce sees itself retiring on Mars. I wish them the best of luck. I’d be happy to die on Mars even if it was just “on impact.” At least I’d get the trip first!

  9. Because of Joe Biden, I hear about idea of states leaving the Union.
    Have not been paying it much attention.
    But maybe Texas leaves, because they want to be most powerful nation on Earth.
    Or what delaying the launch is environment assessment, in which FAA could decide it require environmental impact report- requiring years to complete, rather than just environmental assessment.

    But for Starship to be cheap, rather than just testing it, it will require environmental impact report. Or rather the FAA being in the lead in environmental assessment, the hopeless EPA will be running it, and the amount factors to included in terms environmental impact report could near infinite.
    Or simply impossible to assess.
    Or it’s like the delay of building a nuclear powerplant- endless delays and crazy stuff like involving the unknown Mars environment in decision process.

    1. Native Texan here. For secession to be worthwhile we’ll need to have California annex Austin.

      Just kidding. Secession is a terrible idea, but it’s fun to think about.

      1. Texas started out as a Republic and then was admitted as a State into the Union.

        Is there any substance to the claim that the circumstances and terms by which Texas was admitted allow it to secede?

  10. Magpie,
    Too bad “the tech rep ‘patched’ the computer code so it would automatically abort ‘divide by zero’ calls in the future.” Had it let it run we might have gotten the question to the answer that is 42.

  11. What Star Ship/Super Heavy do is force us to confront all the other unanswered questions of space travel now that our attention isn’t being diverted by trying to build interplanetary travel around what amount to pop bottle rockets.

    Can humans survive 2 years in a low to no gravity environment?

    If they can, will they be able to return to a full g and function in a reasonable fashion?

    What will be the long term effects of radiation exposure outside the protection of a magnetic field?

    Can we procreate and give birth under reduced or no gravity?

    And so on.

    In a rational world, SLS would die as the first Star Ship re-enters the atmosphere.

    1. You’re right about these points, of course, though some of them are more important than others. I’d rate procreation most important, and they all require human experimentation.

    1. In my job, API is Application Program Interface. If that’s what he intended, the meaning I think is that the launch details aren’t really considered by the end-user–the user instead starts with something like “Step 1. Construct science package (est. mass 200-300 tons). Step 2. Science package and transfer stage in LEO. Step 3…”

  13. The agency wants the rocket to become a “sustainable and affordable system.”

    Alas, the system can neither be made sustainable or affordable due to the original statement of requirements that resulted in two SRBs and disposable second stage with 4 main engines wasted.

  14. Regarding the Friday morning update…

    Rand, you’re missing the obvious; NASA can very easily achieve this 50% cost reduction. The clue is their lack of transparency on the cost. Let’s say the real cost of producing and flying (including staff costs, ground support costs, etc) of one SLS per year is 2 billion. In order to achieve this 50% SLS cost reduction, all NASA has to do is allot a billion of the support costs to other programs (thereby likely incurring 200 million in extra overhead and management costs, but that’s okay as it won’t count as SLS costs) thereby “reducing” SLS costs to 1 billion a year (a 50% reduction).

    It’s the art of creative bookkeeping. They’ve proven themselves quite good at it. (building rockets though, not so much…)

  15. Nelson is going to ensure that the pork keeps rolling in, if that means no rockets actually fly, he’s okay with that.

  16. I’m pretty sure you could get SLS-1 down to around the cost of a Delta IV Heavy by upping the production rate to around 4 a year. That would mean a full production line for RD-25E and Delta IV Upper Stage. The result would be twice what Delta IV Heavy could do.

    But there’s also a path to creating an SLS that’s largely reusable and thus cheap enough. I propsed this around 12 years ago on NSF and got hooted at for my trouble: Replace the 5seg RSRMs with 6 Falcon Heavy boosters (which you would recover down range on the deck of a former oil tanker). Put 5 RS-24 engines in a boat tail, to be recovered at Australia, using a discardable heat shield, parachutes and airbags. The result would put an ACES-70 upper stage and around 200 tons of payload in LEO. In 2009, this was fantasy. Now? Falcon booster recovery is routine and ULA bleats about “SMART” engine recovery. The best part is the new name… wait for it…

    Delta V.

    2. I had a piece in The Space Review almost four years ago that spitballed a related revised architecture for SLS including the use of Falcon 9 1st stages as strap-ons. Turns out you can actually get 10 of those around the circumference of an SLS core stage.

      1. I think I remember seeing that when it came out. Back in 2009, anything to do with the as-yet unflown Falcon 9 was considered absurd. But Musk intended Falcon Heavy from the beginning. When he still considered Falcon 5 as well as 9, there was Falcon 9S5 and 9S9. Only 9S9 got built, of course.

        I’ve been working these ideas for a long time. I had a piece in Spaceflight (the BIS magazine), in 2004 I think, called “Off the Shelf and On to Mars” that talked about reviving the old Type 1 SDV idea from 1977, and side-mounting an Ariane 5 cryogenic core, with an Apollo-like spacecraft on top. It would have been a two-launch lunar solution, Farther back, I did a piece in the Nov. 1989 Ad Astra called “Harvesting the Near Earthers,” that I used as the basis for my 1991 novel “Fellow Traveler,” about a Soviet attempt at an asteroid retrieval. Lori Garver was Ad Astra editor at the time, and my phone chats with here probably planted the idea that became ARM (and thus led to Gateway).

        Neither of these pieces is available on line, as far as I can tell. “Harvesting the Near Earthers” is in the back of paperback copies of “Fellow Traveler,” and “Off the Shelf and on to Mars” is in my e-book collection “We are the Hollow Men.”

        1. Well, at the time (2004) it was the best choice. Falcon 1 seemed improbable, Falcon 5/9 unlikely, and Mars Colonial Transport (eventually Starship) was in the “wet dream” category. My article was called “Off the Shelf and On to Mars” for a good reason. You could actually buy RSRMs, ETs, RS-25Ds, and Ariane 5 cryocores in 2004. You could also look at photographs of the 1977 STS test thrust frame and see what your SDV would look like. It was a believable fantasy. All that remained was to figure out the spaceship. You had the Apollo CSM blueprints, and could certainly buy AJ-10s from Aerojet. Developing this, starting in 2005, would have taken 4-5 years at a cost of $4-5bln. It was the right choice at the time. And by 2010, the shape of Falcons and Dragons to come was at hand. Instead, Bush, Obama, and a couple of Senators, made many bad choices, and here we are almost 20 years later.

  17. Felix Schlang had an update on the launch platform for Starship. Brought out some really interesting points on how the stand itself connects to each of the fixed engines in the outer ring for their initial pressurization and such. With the end effect of off-loading that off out of the booster into ground support. The outer engines are only used for lift-off and not descent, so only the inner engines have that extra plumbing, weight and complexity.

    That just illustrates the out-of-the-box thinking that NASA and it’s contractors just can’t conceive of.

    1. Musk refers to the launch infrastructure as “Stage Zero.” Anything that doesn’t have to be on SuperHeavy/Starship, isn’t. The outer engines on SuperHeavy are labeled RBoost. The 3 sealevel engines on Starship are mainly intended for steering and landing, most of the impulse going through the 3 Rvac engines. Starkicker only has Rvacs, and uses RCS thrusters for steering, as it will never land.

Comments are closed.