Abort options for Starship and Starship/SuperHeavy

[Robotbeat]

Right, your arguments have been opinion or begging the question for the vast majority of this thread, you don’t need to keep reminding us of that.

1) I don't think you know what begging the question means.
2) Everything you've posted is an opinion too.  We don't have any data because, if you recall, this vehicle has never made orbit.

1) when you demanded to know what would happen if Starship could never achieve 400 or 2000 consecutive successful launches. To even answer that question would require assuming they couldn’t meet their cost/reuse/reliability goals in the answer.

2) No, it’s also a conditional strategy for achieving the next level of reliability and launch demand size. It’s a possible path to reach very high reliability levels. And I’ve posted actual statistics from the early part of spaceflight, numbers they’ve achieved in the similar F9, etc. not every logical argument about possible things in the future is equally mere opinion. Calling something that has a logical argument behind it “delusional” isn’t “just my opinion, man,” it’s just an uncalled-for insult.

[Robotbeat]

First Starlink and trips to deep space using lots of refueling enabling ultimately 99.9% reliability, then point to point cargo enabling 99.99%, then point to point chartered flights and mass trips to LEO and deep space for thousands of people enabling 99.999%, then hopefully point to point flights for millions of people, and millions of people  and millions of tons to orbit and deep space enabling 99.9999+%.

It's just an opinion, but I think those numbers are completely delusional.  Never, ever going to get there with this architecture.

I don't think you're going to get those numbers no matter the architecture. Well maybe a space elevator might work.

Space elevator is worse. Space elevator is just what people say when they want to wave their hands a lot.

Don't overthink it, I just said space elevator because it was the first thing that came to my mind in terms of what would theoretically be the safest way to get to space. Maybe there are other far out options there that I haven't heard of that don't rely on chemical rocket engines.

Don't know what you mean about waving hands a lot.

Basically exactly what you said about it being some theoretical thing.

Chemical engines don’t have some fundamental physics that guarantees they have to be unsafe. They’re far simpler than jet engines, for instance, and can have all solid surfaces be at lower temperatures than many of the high temperature parts of jet engines. And the stress margins can be just as good as for jet engines. Taking it as axiomatic that chemical rockets have a threshold of reliability beyond which it’s not physically possible to improve them just isn’t true or scientific.

Raptor might never achieve the reliability to enable this, but I’ve seen Xcor describe engine designs that could achieve essentially unlimited lifetime by keeping stresses far below cycling limits.

[Lee Jay]

And the stress margins can be just as good as for jet engines. Taking it as axiomatic that chemical rockets have a threshold of reliability beyond which it’s not physically possible to improve them just isn’t true or scientific.

It is if you include the horrid payload mass fraction of rockets that forces you to save weight any way you can.

[Robotbeat]

And the stress margins can be just as good as for jet engines. Taking it as axiomatic that chemical rockets have a threshold of reliability beyond which it’s not physically possible to improve them just isn’t true or scientific.

It is if you include the horrid payload mass fraction of rockets that forces you to save weight any way you can.

Nope, there are big ways to address the weight of rocket engines while maintaining this stress margin. We've barely scratched the surface of them, and we've had essentially no need to do it until now.

https://www.researchgate.net/publication/343704990_Lightweight_Thrust_Chamber_Assemblies_using_Multi-Alloy_Additive_Manufacturing_and_Composite_Overwrap
https://www.nasa.gov/directorates/spacetech/game_changing_development/projects/RAMPT

Allowable stress of Copper (using wires as an example) is about 180 MPa for one cycle and drops to 120MPa at 5,000 cycles. Not exactly a big deal. Moving the primary stress from being carried by copper to being carried by, say, a composite overwrap is potentially a bigger difference in mass savings than the extra mass it takes to go from 1 cycle lifetime to 5,000 cycles lifetime.
https://www.matec-conferences.org/articles/matecconf/pdf/2018/24/matecconf_fatigue2018_06002.pdf

copper allow used to make 3D printed rocket engines called GRCop-42 has a strength of about 360MPa and a density of about 7.8kg/liter. T1100G carbon fiber has a strength of 7000MPa and a density of about 1.79kg/liter. Even with resin and a quasi-iso layup and strength knock-downs from cycling, composites still would allow a potential weight reduction with a properly designed engine. More than enough to account for higher cycle rocket engines.

[TheRadicalModerate]

I also would assume at least a factor of 3 (half an order of magnitude) improvement in reliability/safety from launches occurring with crew due to higher standard of quality assurance and margin. Not counting high thrust abort systems.

Soyuz launches with crew have a higher reliability than an R7/Soyuz family launch without crew. (This is something we can check on, and I’ll do so later tonight I think.)

I'd expect some improvement.  But you know how you verify such an improvement?

With PRA!

Beyond that, you can only use better QC to improve failure modes that you completely understand, i.e., where the relevant hunk of the PRA model is mature.  For the parts where you simply don't understand the behavior or have provably not thought of all the corner cases, all the QC in the world will do you no good.

If that condition persists, then you need a Band-Aid.

[Robotbeat]

Or you can get an estimate of it from Soyuz statistics.

[TheRadicalModerate]

Or you can get an estimate of it from Soyuz statistics.

Only if a Starship behaves like a Soyuz.  It doesn't.

[Robotbeat]

Or you can get an estimate of it from Soyuz statistics.

Only if a Starship behaves like a Soyuz.  It doesn't.

It behaves better than people concern-trolling on webforums with zero data.

It's impossible for you and me to do PRA on this, whereas Soyuz statistics are public.

[Lee Jay]

It is if you include the horrid payload mass fraction of rockets that forces you to save weight any way you can.

Nope, there are big ways to address the weight of rocket engines while maintaining this stress margin.

Ah, so it's possible to arbitrarily reduce stress with no added mass and no packaging or volume penalties.  Good to know.

Of course you could use those approaches to reduce mass but I guess there must be a law against that.

[InterestedEngineer]

Does it meet 1930s-era civilian air flight pLOC rates?  Absolutely not, until at least 2035 or so.

But this is my whole point:  2035 is too late.

It's too late to put private individuals into cislunar, which requires being able to defray the costs across 12+ people.  Without the private cislunar traffic, NASA won't suffer the embarrassment they need to, and will stand pat on SLS/Orion, continuing to use LSS only as a nuclear weapon to swat a fly.  If NASA stands pat on SLS/Orion, then Artemis will fail and be cancelled.  If Artemis is cancelled, then SpaceX has no platform to subsidize its testing to get ready for Mars--or any other crewed BEO destination, for that matter.

I think if SpaceX could talk NASA into D2-assisted LSS, then they probably could wait until 2035 for a crewed launch/EDL capability.  But I don't think they can do that without embarrassing NASA and their congressional patrons, and that requires private flights, starting somewhere in the 2028 timeframe.

As I said, way up-thread now, an alternative strategy is to subsidize a couple of D2-LSS cislunar flights.  If that gets NASA to move, it can be SpaceX's anchor customer for all kinds of R&D for Mars.  But the much better solution is to get bigger crews launched, sooner.  (This also keeps Maezawa from being really angry and/or litigious.)  I think it's so much better a solution that it's worth doing the escape engineering.

the bar isn’t to get to 1920s pLOC figures before sending anyone to space in Starship. Just need to beat pLOC of Artemis 2, probably around 5% (in *reality*).

40 consecutive launches, re-entries, and landings of starship (and then adding extra margin and quality assurance for crewed flights) would absolutely crush the pLOC levels that NASA will ever get with Orion and SLS. That’s not at all unreasonable to expect by Artemis 3, say.

I'm too lazy and stupid to wade through all the statistical methods, so I just cribbed the Lewis points and the adjusted-Wald score from Ed Kyle's old reliability metrics at the late lamented spacelaunchreport.  (It may be gone, but its spirit carries on in the Wayback Machine...)

40 launches, no failures:
Lewis score: 97.6%
Adjusted-Wald 95% confidence interval = [89.6%, 100%] with a mean of 95.6%

You can get the Adjusted-Wald to a mean of 99% with 189 consecutive successful missions, 95% CI=[97.6%, 100%].  Lewis score would be 99.5%.

If you want the lower bound of the 95% CI to be at least 99%, you need 460 consecutive successes.

Just for grins and giggles, 2 failures in 500 missions would give you a 95% CI=[98.5%, 100.0%], mean 99.2%.

And, since we're also talking about successful landings, from the time the F9 first did a successful landing, its record is (if I tallied it right), 155/164.  95% CI=[89.8%, 97.2%] with a mean of 93.5%.  Lewis point: 94.4%.

Link?

That seems pessimistic, and counting early failures completely disregards a process that has learning capabilities.

How many F9 landings in a row are there right now?  I think it's 130 or so.

[chopsticks]

It behaves better than people concern-trolling on webforums with zero data.

Ah yes, the resort to the concern troll accusation again.

There is zero data that Starship will perform as hoped for either—and how long before it will?—and SpaceX is attempting a number of things that haven't been done before with this vehicle. Space is hard and lots has to go right so people don't die. I'm not saying it won't work, but I just hope that the folks in charge of this have at least some sort of a contingency plan to add an abort system.

That said, I know that there are lots of competent people involved in this. I just don't see how flying people without an abort system for the first few years is going to work out. (Maybe in 10 years or so humans landing propulsively on Earth will be widely accepted)

[Robotbeat]

It is if you include the horrid payload mass fraction of rockets that forces you to save weight any way you can.

Nope, there are big ways to address the weight of rocket engines while maintaining this stress margin.

Ah, so it's possible to arbitrarily reduce stress with no added mass and no packaging or volume penalties.  Good to know.

Of course you could use those approaches to reduce mass but I guess there must be a law against that.

Yes, you can use those approaches to reduce mass, but there are minimum gauge issues that prevent you from reducing much. If you wanted to use the added strength to reduce stress to increase cycle life, there’s a lot more room there.

Additionally, I find it interesting that the engine with the highest thrust to weight ratio ever, Merlin 1D, is also the only one of two highly reusable engines ever made and flown and reused, firing dozens of times. Lighter for a given thrust than the best expendable-only engine.

Even if new materials and manufacturing techniques could be used for making existing engines lighter instead of increasing cycle life, there’s no real strong need for being much lighter as we know current engines are light enough for the required purposes. That cannot be the showstopper for thousands of uses.

[sebk]

Does it meet 1930s-era civilian air flight pLOC rates?  Absolutely not, until at least 2035 or so.

But this is my whole point:  2035 is too late.

It's too late to put private individuals into cislunar, which requires being able to defray the costs across 12+ people.  Without the private cislunar traffic, NASA won't suffer the embarrassment they need to, and will stand pat on SLS/Orion, continuing to use LSS only as a nuclear weapon to swat a fly.  If NASA stands pat on SLS/Orion, then Artemis will fail and be cancelled.  If Artemis is cancelled, then SpaceX has no platform to subsidize its testing to get ready for Mars--or any other crewed BEO destination, for that matter.

I think if SpaceX could talk NASA into D2-assisted LSS, then they probably could wait until 2035 for a crewed launch/EDL capability.  But I don't think they can do that without embarrassing NASA and their congressional patrons, and that requires private flights, starting somewhere in the 2028 timeframe.

As I said, way up-thread now, an alternative strategy is to subsidize a couple of D2-LSS cislunar flights.  If that gets NASA to move, it can be SpaceX's anchor customer for all kinds of R&D for Mars.  But the much better solution is to get bigger crews launched, sooner.  (This also keeps Maezawa from being really angry and/or litigious.)  I think it's so much better a solution that it's worth doing the escape engineering.

the bar isn’t to get to 1920s pLOC figures before sending anyone to space in Starship. Just need to beat pLOC of Artemis 2, probably around 5% (in *reality*).

40 consecutive launches, re-entries, and landings of starship (and then adding extra margin and quality assurance for crewed flights) would absolutely crush the pLOC levels that NASA will ever get with Orion and SLS. That’s not at all unreasonable to expect by Artemis 3, say.

I'm too lazy and stupid to wade through all the statistical methods, so I just cribbed the Lewis points and the adjusted-Wald score from Ed Kyle's old reliability metrics at the late lamented spacelaunchreport.  (It may be gone, but its spirit carries on in the Wayback Machine...)

40 launches, no failures:
Lewis score: 97.6%
Adjusted-Wald 95% confidence interval = [89.6%, 100%] with a mean of 95.6%

You can get the Adjusted-Wald to a mean of 99% with 189 consecutive successful missions, 95% CI=[97.6%, 100%].  Lewis score would be 99.5%.

If you want the lower bound of the 95% CI to be at least 99%, you need 460 consecutive successes.

Just for grins and giggles, 2 failures in 500 missions would give you a 95% CI=[98.5%, 100.0%], mean 99.2%.

And, since we're also talking about successful landings, from the time the F9 first did a successful landing, its record is (if I tallied it right), 155/164.  95% CI=[89.8%, 97.2%] with a mean of 93.5%.  Lewis point: 94.4%.

These are purely naive models assuming no learning whatsoever, i.e. static reliability, which is unrealistic. They are good for runs of serially produced stuff where no learnings are applied. If you want to estimate reliability for consumers of a particular production run of cars, or radios, or ball bearing, or so - fine. But they're invalid for systems where learning is expected to be a core part of the progression.

AFAIR The most naive model which accounted for learning I saw (actually for rocket and spacecraft reliability estimation) was the model when the current fallibility (1-complement of reliability) would be estimated by taking the minimum of the inverse of the doubled length of the unbroken run of successes preciding the last failure and the inverse of the doubled length of the current run. It's based on the assumption that the last failure reason has been fixed, and the next lingering reason is less probable and blatantly assumed it's 2x less probable. Then it also assumed that if the current length is longer that that it's more likely that the improvement have been better and it's most likely in we're in the middle of the next success run, not at the last step of it.

A better model would be to use plain old linear regression of previous failure intervals. Of course you need to have a few failures for the model to be worth something, but in the case of Falcon 9 landings we do have it. Anyone want's to run it?

If one has a system with long history (like R-7) then linear regressions to a 3rd degree polynomial estimate would work better.

[Oersted]

but it will never be safe enough for the general public, which is it's express reason to exist.

Never say never.

Starship is being built by the company which is regularly transporting people to orbit on the Crew Dragon. Nobody is afraid of flying that one. I think that bodes well for the future of Starship.

[chopsticks]

I don't know if this has been mentioned yet in this thread, but during one of the interviews with Jared Isaacman (can't remember with who anymore) Jared was asked about LAS on Starship. He didn't say a lot about it, but he said something to the effect of "Dragon was essentially built for NASA, whereas Starship is a SpaceX idea". I'm totally paraphrasing this, but what I got from it was that Dragon was heavily constrained by NASA's requirements but with Starship, SpaceX is doing their own thing and their engineers are doing what makes sense to them. If someone can find the video of this interview and the timestamp that would be awesome - I don't want to mis-characterize Jared here.

I inferred from what he said that SpaceX was going to figure out a way to not have an abort system as we know it, but perhaps I was reading too much into it. If this is the case, I guess it remains to be seen what NASA will have to say about it, and if they will ever feel comfortable putting their own crew on this thing.

Assuming that SpaceX totally closes the door on abort capability as we know it, what I see happening is private missions like Dear Moon and Polaris being the guinea pigs of the system before NASA okays it for their astronauts. I wonder how many crewed flights this would take.

TBD..

[Lee Jay]

but it will never be safe enough for the general public, which is it's express reason to exist.

Never say never.

Starship is being built by the company which is regularly transporting people to orbit on the Crew Dragon. Nobody is afraid of flying that one. I think that bodes well for the future of Starship.

Dragon has abort modes and an inherently safe method of EDL (doesn't require a liquid propulsion system to work).  And it's still not safe enough for the general public.

[chopsticks]

And it's still not safe enough for the general public.

I wouldn't get too hung up on this part. "General public" if you're referring to an average airline or rail passenger is I believe a looong ways off: whether it's SpaceX flying people or anyone else for that matter.

I personally don't buy a lot into the PR and buzz generated around this. I think regardless of what people want to believe, travelling to to LEO or beyond is going to remain a fairly niche segment of travellers for quite awhile. Yes, the people flying on the dear moon mission are ordinary people in some regard, but as mentioned by Tim Dodd, he had to get a medical evaluation and I'm sure there will be lots of training.

Space will become more accessible, but it becoming as ordinary as boarding an airplane is a fantasy at this point, IMO.

[edzieba]

but it will never be safe enough for the general public, which is it's express reason to exist.

Never say never.

Starship is being built by the company which is regularly transporting people to orbit on the Crew Dragon. Nobody is afraid of flying that one. I think that bodes well for the future of Starship.

Dragon has abort modes and an inherently safe method of EDL (doesn't require a liquid propulsion system to work).  And it's still not safe enough for the general public.

Presence of absence of abort modes does not inherently confer or remove safety. All they do is reduce some specific risk that is not reducible any other way.

We do not care that solid motor abort towers are normally jettisoned just after the first stage burn completes leaving a capsule 'abortless' riding a fully fuelled second (or possibly third) stage. There are abort options using e.g. a capsule's service module on-board motors to escape a stage that gracefully fails (e.g. contained engine failure), but if the stage RUDs then you don't have a backup abort option. We accept that making the stage sufficiently reliable is an acceptable alternative. We do not demand adding abort-abort motors in the event of an abort motor failure, we just accept that abort motors can be made sufficient reliable.

We have demanded that trans-oceanic aircraft must have multiple engines, and have then happily compromised from requiring 4 engines with engine-out to accepting two engines of sufficient reliability without any engine out capability (ETOPS). We do not concern ourselves that aircraft have no abort-modes from wing failure (despite it being a failure mode that has occurred in flight) as long as we have deemed the structure sufficiently reliable and maintained.

No commercial car has an abort system in the event of a steering assembly failure, despite that being a single point of failure. We accept that the assembly can be made sufficiently reliable to not require one.


The common theme is that if a system or component can be deemed to be sufficiently reliable, we are happy to accept that there is no abort mode. In the early days of human spaceflight abort modes were a necessity because there was no avenue to system maturity within the timeframe required to start flying humans within. Starship is not under that deadline: it can wait to fly enough to demonstrate reliability before launching or landing humans, and the only deadline is whether Maezawa is willing to continue to push back the launch date (or willing to pay a little extra for EOR launch/land using Dragon).

[TheRadicalModerate]

Dragon has abort modes and an inherently safe method of EDL (doesn't require a liquid propulsion system to work).  And it's still not safe enough for the general public.

Something to bear in mind:  Commercial human space is operating under a "learning period" regulatory moratorium, enabled by multiple extensions to the legislation that originally established it.  That moratorium may easily be extended yet again, but it's currently set to expire in October, 2023.

At some point, it's going to expire for real, and there will be regulations that all commercial human systems will have to meet.  I would expect those regulations to be fairly forgiving compared to aviation regulations of any sort, but I wouldn't be surprised if they adopted NASA's 1/1000 pLOC for each of launch and EDL, and 1/270 pLOC for a six-month mission.

[Lee Jay]

but it will never be safe enough for the general public, which is it's express reason to exist.

Never say never.

Starship is being built by the company which is regularly transporting people to orbit on the Crew Dragon. Nobody is afraid of flying that one. I think that bodes well for the future of Starship.

Dragon has abort modes and an inherently safe method of EDL (doesn't require a liquid propulsion system to work).  And it's still not safe enough for the general public.

Presence of absence of abort modes does not inherently confer or remove safety. All they do is reduce some specific risk that is not reducible any other way.

We do not care that solid motor abort towers are normally jettisoned just after the first stage burn completes leaving a capsule 'abortless' riding a fully fuelled second (or possibly third) stage. There are abort options using e.g. a capsule's service module on-board motors to escape a stage that gracefully fails (e.g. contained engine failure), but if the stage RUDs then you don't have a backup abort option. We accept that making the stage sufficiently reliable is an acceptable alternative. We do not demand adding abort-abort motors in the event of an abort motor failure, we just accept that abort motors can be made sufficient reliable.

We have demanded that trans-oceanic aircraft must have multiple engines, and have then happily compromised from requiring 4 engines with engine-out to accepting two engines of sufficient reliability without any engine out capability (ETOPS). We do not concern ourselves that aircraft have no abort-modes from wing failure (despite it being a failure mode that has occurred in flight) as long as we have deemed the structure sufficiently reliable and maintained.

No commercial car has an abort system in the event of a steering assembly failure, despite that being a single point of failure. We accept that the assembly can be made sufficiently reliable to not require one.


The common theme is that if a system or component can be deemed to be sufficiently reliable, we are happy to accept that there is no abort mode. In the early days of human spaceflight abort modes were a necessity because there was no avenue to system maturity within the timeframe required to start flying humans within. Starship is not under that deadline: it can wait to fly enough to demonstrate reliability before launching or landing humans, and the only deadline is whether Maezawa is willing to continue to push back the launch date (or willing to pay a little extra for EOR launch/land using Dragon).

I seriously doubt that SS will ever get to be reliable enough for the general public to justify having no abort system. Not without major changes.