What if Starship can't work as currently envisaged?

[OTV Booster]

There are several possibilities depending on what part is the killer problem.

I'd argue that Super Heavy is already developed past the point where "it will never be able to launch anything" is plausible. So the simplest, but also most limited, fallback would be a pure Starlink V3 launcher - expendable upper stage on top of a Super Heavy. That would work for Starlink but not for HLS or Mars.

Beyond that, it depends on where the problem is.

The Raptor engine itself seems ok, but there could be fundamental problems with the integration into the vehicle in Starship v2. That could presumably be redesigned.

If the TPS cannot be made reliable enough ... well, HLS could probably be done with expendable tankers, it'd just be more expensive, but that would rule out the Mars plan.

OTOH, a total TPS redesign (transpiration cooling everywhere?) is a possibility.

--

I don't think SpaceX will give up on the general concept/category of "super heavy rocket with some kind of Raptor/Raptor derivative engine". But significant redesign is possible.

Superficially, it appears that the TPS is doing ok except in specific areas. Specifically the fin hinge and root areas. Active cooling needs might be limited to these areas.


IMO rapid turnaround may be one of the stupid requirement that Elon famously hates. Demanding a rocket ship to perform like an airliner always struck me as being a bit silly. Can the system work economically if a ship turnaround eventually settles in to 15 days? Ten days? Five days?

[Kaputnik]

5.5km/s delta-v isn’t even that hard.

Well that's a good thing, then. As currently sized, SS seems to be aiming for significantly more delta V than that. So shrinking the Ship could be a very helpful move, if required.

I mean, you literally just watched 2STO RLV with RTLS on one of the previous starship flights.

Sure. And it had no payload. And we don't know how much refurb it would have needed to fly again.

Imagine it's the 1950s and we're talking about supersonic airliners. "Look", you say, "lots of planes have already broken the sound barrier, why on earth would you even think that we won't have commercially successful SST within a few years?"
Well we know how that panned out. After many years and huge government subsidy a viable product was produced, but it turned out to be very difficult to go from a vehicle that worked in a limited use case to one that worked economically and safely at large scale.

That's a very long winded way of saying that Starship will have failed of it can't fly safely and economically.

I just don’t understand this kind of refusal to acknowledge something can be done unless it’s already being done. Certainly no one who thinks like this has ever attempted to actually invent or build something new.

I'm not refusing to acknowledge that it can be done. You are confusing a question with an assertion.
*IF* the current plan can't be made to work, what are the steps that could be taken to get the design over the line?

[Robotbeat]

Concorde was arguably a success. It operated for decades.

It’s sort of obvious to me that Starship might not succeed like is hoped. But that doesn’t mean there are obvious technical showstoppers.

[SpaceLizard]

Classic Mr. Scott post: Loads of referential gibberish, no productive contribution... 

[Twark_Main]

Doing a thing differently and expecting the same or different result is called RESEARCH.

Doing the same thing and expecting a different result is referred to as INSANITY

No two Starship launches have been the same hardware, and no two Starships have failed for the same reason. So SpaceX is doing good on this front.

Let’s say there are 50 critical parts that are all dependent on each other.  The number of questions needed to be asked can be estimated as n!

This is why in any real system you use modularity so you don't have everything dependent on everything else, and redundancy so you minimize the number of critical parts.

Modules talk across well-defined interfaces, which manages complexity and risk. This was the key lesson of Apollo.

https://www.youtube.com/watch?v=e4g5vJ5B6uM

So 50 factorial (50!) is much larger than Avogadro’s number (6.022x10^23) of questions.  Starship has much more than 50 interdependent parts (by inspection).  So the fact is that Starship is clearly a boondoggle.

Putting the dubious "by inspection" at such a critical point reduces your whole argument to "trust me bro, I asserted it."

Iterative development usually starts with a working prototype that gets scaled up with simple experimental changes.  Uhh, that’s not what is happening.

That's because that work takes place in labs we can't see into. 

If you don't think SpaceX uses iterative scale-up then you don't know Rasky. Dan Rasky. Here's the NASA engineer describing the SpaceX iterative scale-up process used for the Dragon heat shield:

https://www.youtube.com/watch?v=SMLDAgDNOhk

(the whole playlist is worth watching, lots of insights about Musk from an outsider who actually worked there)

Most of the scale-up phase of Starship has already been completed (some of it happening before our eyes in past years), but of course they're still doing small-scale iterative projects such as the one-off prototype tiles we've seen. Clearly these are just a bunch of different prototype designs made by people working at a lab bench or machine shop, and whatever works best will be chosen for the next phase of scale-up.

I actually expect there was at least one round of iteration with ground testing before the tiles got put on Starship, so what we're seeing is the best tile designs that made it through the previous round.

[meekGee]

Doing a thing differently and expecting the same or different result is called RESEARCH.

Doing the same thing and expecting a different result is referred to as INSANITY

No two Starship launches have been the same hardware, and no two Starships have failed for the same reason. So SpaceX is doing good on this front.

When Mr Scott says they keep doing the same thing, he means they keep iterating.

His syntax is perfect, and semantically he's pretty sound.  The logic is lacking, but hey, 2 out of 3 ain't bad.

[AC in NC]

And no, Elon didn't give up on Mars, Peter Thiel is wrong. Elon corrected him here.

Holy Crap!  How can Thiel have butchered the takeaway from what he said Elon said that badly?

[AC in NC]

When Mr Scott says they keep doing the same thing, he means they keep iterating.

When Mr Scott says anything, he's going to delete it within a few weeks because he's a troll

[yoram]

IMO rapid turnaround may be one of the stupid requirement that Elon famously hates. Demanding a rocket ship to perform like an airliner always struck me as being a bit silly. Can the system work economically if a ship turnaround eventually settles in to 15 days? Ten days? Five days?

I don't think the real goal is fast turn around, but cheap turn around, ideally with no to minimal work. If it takes 20 days it likely needs a lot of work hours, so won't be cheap.

An extreme case would be the Shuttle which ended up being extremely non economical because of this.

[meekGee]

When Mr Scott says they keep doing the same thing, he means they keep iterating.

When Mr Scott says anything, he's going to delete it within a few weeks because he's a troll

I didn't know he did that! In a way it's adorable. It shows that he cares... 

[JulesVerneATV]

Starship is just a tool like Grasshopper was a tool, or Starhopper but its possible Starship wont work as fully envisioned then you throw away the tool that doesn't work and move on to the next version. V2 Facing a Curse? Move on to the next Mars Colonyship, 'Das Marsprojekt' Mars Expedition ship.
https://web.archive.org/web/20100531192655/https://ston.jsc.nasa.gov/collections/TRS/_techrep/CR-2001-208928.pdf
Yes some will be feeling disappointment , its possible China will beat people to get Mars samples with some Chinese MSR samples mission  but that was expected, does that really matter? They in China already have done other stuff others didn't at the far side of the Moon. Elon Musk and Space-X is a lot more than just Starship Version 2, there are robots, there is AI, there are satellites, driver-less cars and you need rovers in space,  he has influence just as people had influence before Apollo, there is the underground tunneling the Boring Company.

[spacenut]

Starship version 1 seems to have worked.  It went into semi orbit, and returned from space.  Cameras did show some heat damage on the fins, etc.  Version 2 with Raptor 2's have not worked.  Seems to be Raptor 2's are the problem.  They haven't had a way to test the fin modifications yet.  Maybe Raptor 3's will be better in the 3rd version. 

Each Raptor version has increased thrust.  Is that the problem?  Is plumbing the problem?  Are COPV's the problem?
Would electric pumps work better than COPV's?  A lot still has to work out. 

However, the Booster seems to be working fine and has even landed back at the launch pad.  This is huge, as this is a heavier lifter than SLS.  Upper stages can be modified to be expendable, or 3 stages, etc.  Not there yet as SpaceX is trying to get Starship functional. 

Fast turnaround can be just having 5-10 boosters ready and stored as well as 5-10 Starships.  They could launch every day and still have 5-10 days for refurbishment of those that just landed.  Same with having more launch facilities. 

If SpaceX had a few more drone ships and a few more launch pads, F9 would be launching every day or even twice a day.  It all depends on how fast refurbishment of a booster or a Starship takes to have faster turnarounds. 

[envy887]

If they can't make it work, at some point (when money becomes an issue) they will have to stop trying and scrap the program.  I don't see a middle ground with this outfit.  Here's hoping they make it work.

 - Ed Kyle

That might be a while. The Falcon 9 / Starlink combo are steamrolling the competition, and pulling in a ton of cash.

[OTV Booster]

To me what Starship looks like is just a platform to test rocket engines.  In the movie industry, it’s a Macguffin.  Just a plot device, but never the main part of the story. Skunkworks did this carp.

It’s all about the rocket engines and only the rocket engines.  HLS as a vehicle has not materialized in hardware. 

So I ask myself, why… and all I can surmise is that SX folks wanted to just establish engine performance and manufacturing production capabilities.

It's so funny to me how people still don't understand the concept of iterative development.

SpaceX can tell people till they're blue in the face, but fundamentally if you don't get it then you don't get it. 

No no no.  I’ve done iterative development before it was cool. 

A little descriptive example to help with understanding.

As many of us have left the safe spaces of academia, what you begin to see are the loosely defined aspects of the scientific method.  Unfortunately, what SpaceX is doing is not related to the scientific method.

Doing a thing differently and expecting the same or different result is called RESEARCH.

Doing the same thing and expecting a different result is referred to as INSANITY.

What Starship is doing to the greatest extent practicable is the latter. This is what happens when you’re hardware rich and you have a billionaire that only has one Plank’s time constant to ask a limited amount of questions. 

There is not enough time in the universe let alone on a one hour game show to deal with the quantity of necessary questions that bring a semi-caffeinated/ group think ability to fix something as complex as Starship.

Let’s say there are 50 critical parts that are all dependent on each other.  The number of questions needed to be asked can be estimated as n!

So 50 factorial (50!) is much larger than Avogadro’s number (6.022x10^23) of questions.  Starship has much more than 50 interdependent parts (by inspection).  So the fact is that Starship is clearly a boondoggle. 

Iterative development usually starts with a working prototype that gets scaled up with simple experimental changes.  Uhh, that’s not what is happening.

Mr. Scott, I have to ask a personal question to which you have no obligation to answer.


Do you have a negative opinion of Elon Musk? The accusation has been made in these discussions but I prefer knowledge to opinions and so I ask.


My experience is that many people dislike Musk and use it as a basis for judging all his actions. I share this dislike but still attempt some degree of dispassion, judging his actions independent of my negative opinion.


Another OPINION I have is that despite what I see as a post TwitX descent into wackiness, he is an engineering genius. See Howard Hughs, Henry Ford and Thomas Edison for other examples. Nothing is completely black or white.


So the two part question is, do you dislike Musk and does it shape your dissatisfaction with his StarShip program?

[wannamoonbase]

Funnily, I don't see the biggest risk for SpaceX not from the technical side, but from the political.

Either nationalizing SpaceX or driving out Musk (both openly discussed), would most likely stop any real development of Starship. Regardless of what you think about politics or the political believes of Musk, in a lawless America where the highest court regularly sides with the executive in an unprecedented way, the normal rules do not apply any more. Assuming that Trump will stay in power for the foreseeable future (which at least for me is a given), that bodes not well for SpaceX.

So my assumption is, that the Starship program will not get out of the prototype phase.

ps. I really hope, that I am wrong about that

I don't know if it's probably, but it's for sure a possibility.

The uncertainty and risk hanging over someone is the point of the intimidation.

Like others, I thought for years that Starship was a 'when', but after losing the last 4 ships and the political turmoil it's moved to a 'if' for me.

[sdsds]

With the SpaceX culture failure is always, "Within the set of possible outcomes, even if unlikely." And that's a feature, not a bug. If you're not at least occasionally failing you're not taking enough risk. It really doesn't matter if the cause of the failure is technical or societal.

I wouldn't much care to see these people in a MMA cage fight. I'd much rather assess their skill at no limit Texas Hold'em. My money is on SpaceX.

[TheRadicalModerate]

I'm extremely confident that SpaceX can build an expendable Starship with a reusable SuperHeavy and deliver >100t of payload to LEO.  If SpaceX can really manufacture Raptors at their current scale for $1M apiece and SuperHeavy can get to 20x reuse, it's hard to imagine the cost of a launch being much more than $35M.  If we assume a 70% gross margin (roughly what F9 gets now), that would be ~$120M per flight.  That's a specific cost of $1200/kg, which is about a third of F9's, and half that of an FHE.

That's not a cost revolution, but it's better'n a sharp stick in the eye, and the superheavy lift, even with only partial reusability, is a revolution.

Two biggies that hurt if there's no Starship recovery (not necessarily reuse):

1) You can't land anything on Mars.
2) You can't get Starship crew-certified for EDL, which makes it uncertifiable for launch.


Specific cost is only half the story.  The other half is launch cadence.  Cadence can be limited by either the engine production rate, the hull production rate, or the turnaround rate; it's hard to tell which one is the critical path at this point.  But obviously reuse dramatically affects the needed production rates for a given cadence.

Refueling requires high cadence in addition to low specific cost.

[Robotbeat]

SpaceX will keep trying to make full reuse happen for Starship until they completely die. They fundamentally won't stop with cheap partial reuse. Also, reuse has nothing to do with the problems they've had in the last few flights.

[sdsds]

[...] They fundamentally won't stop with cheap partial reuse.

Right. They already have cheap partial reuse with Falcon and the haven't stopped. It fundamentally doesn't matter whether they're calling this system Starship or Falcon-X; they saw what wasn't working with Falcon and moved to trying something better. They'll do that again if necessary and again it doesn't matter if they call the full re-use version Starship or StarshipFR or ApplePie.

[...] Specific cost is only half the story.  The other half is launch cadence

And the third half is customer demand those massively cheap prices create. I might not be like Isaacman with the resources to self-fund science missions but if under $5k gets my 3U CubeSat to orbit? Sign me up for two or three of those! With that scale of demand and margin SpaceX can fund ApplePie with the proceeds of partially re-usable Starship.

[thespacecow]

Those who think Starship is now an "if" instead of "when", please specify your definition of it "works" so that your prediction can be refuted in the future.

What does it mean by Starship "works":
1. Full reuse - aka relaunch of a recovered ship - achieved?
2. Deployed Starlink operationally?
3. Orbital refueling demonstrated?
4. Starship landed on Moon/Mars?