Building Rocketships In The Desert
Starship test flight rocket just finished assembly at the @SpaceX Texas launch site. This is an actual picture, not a rendering. pic.twitter.com/k1HkueoXaz
— Elon Musk (@elonmusk) January 11, 2019
This is for suborbital VTOL tests. Orbital version is taller, has thicker skins (won’t wrinkle) & a smoothly curving nose section.
— Elon Musk (@elonmusk) January 11, 2019
Go Elon!!!
At this pace and with some luck he will likely beat the SLS to orbit!
And they aren’t even putting all of their effort into it yet. Musk has said previously that commercial crew is where most of their development resources are going at the moment, after that is completed they will shift development resources to BFR (now Starship).
I wonder. Is it possible Musk has determined that the old guard at NASA will never allow CC to be a success and pushing forward with Starship is the only way to really get crewed spaceflight going?
All those wrinkles remind me of the inflatable Christmas decorations that folks have just taken down off their lawns. Or the skins of those many V-2s on display in various museums.
Looking forward to the flight tests!
Musk tweeted the actual Starship will have thicker, less wrinkly, skin. Makes sense as this is only a “hopper”.
Now that looks like a real rocketship should! None of those flying sewerpipes that pass for rockets these days!
Be still my heart…..
Last night Musk also tweeted that the first orbital Starship may be ready by June ???
https://twitter.com/Astro_C…
Van Horn (Blue Origin) is desert. Boca Chica is not.
The speed at which starship is moving is embarrassing for Orion and SLS given the decade head start NASA has had not to mention the budget differential. If this beats EM-2 to test flying with a crew all the SES managers who have wasted almost two decades of dedicated folks careers working on the boondoggle should all be reassigned to latrine duty at mcmurdo .
I suspect Musk is just as happy that they are not getting in his way.
Shiny prop, but hard to fathom what does this have to do with going to Mars
Duh. Its a test article for a much larger rocket.
If large rockets were an actual bottleneck for getting to Mars, there might be a point in this.
Cost is the bottleneck, a large fully reusable launch vehicle/spacecraft/lander is how you get the cost down to affordable range.
No, launch costs haven’t been the actual bottleneck for any deep space missions for a long time, that’s really misguided.
The spacecraft and probes regularly cost 10x the launch cost, easy. Look at Curiosity, or Juno or any major spacecraft sent out of LEO
To think that changing the launch cost by 2x or even 5x or whatever is going to make Mars missions suddenly affordable is kind of high-school level economics
The reason they cost so much is because they must work perfect without any humans around to fix things. When the instruments are just installed in a Starship, or a pod deployed from one, and when there is a repair shop nearby and someone who knows how to fix them, the costs will drop like a rock.
You are on really shaky grounds trying to claim that manned presence makes things cheaper. Any manned spaceflight project has been order(s) of magnitude more expensive than anything robotic.
It’s a very iffy argument, but I’d be more comfortable if someone had actually studied the numbers. According to a National Academies study, unmanned spacecraft developed to class C/D standards are about 80% successfully while those developed to A/B standards are 90% successfully. Past missions and the general consensus of people in the field is that distinction results in a factor of two to three in cost. That’s for unmanned spacecraft with no changes of repairs. Having people around to fix things could, potentially, permit a very high rate of hardware failures and an order of magnitude or more reduction in hardware costs.
Having those people around would, as you point out, drive costs up by at least an order of magnitude or more. So this becomes a question of which, order of magnitude or more, cost differences is greater. I’m inclined to say robotic missions would be better, but I’m not certain enough to say it’s obvious. Some investigation into the details might be in order, especially since we are talking about a few orders of magnitude. When you ask if 10*X is more or less than 100*Y, you need a really good idea what X and Y are.
Manned presence doesn’t equal manned spaceflight. A satellite launcher/tanker BFS would return to Earth after just a few days at most, they can be maintained on the ground, no manned spaceflight needed.
Even a Mars BFS would only need to flight 6 months, then it can be serviced on the Mars surface, and after 2 years it will return to Earth for maintenance. Again, no manned spaceflight needed, except for manned presence on Mars which is the goal of the Mars mission in the first place.
These conditions are radically different from GEO comsats, which need to last 15 years without maintenance.
Spacecraft like Juno and Curiosity are unique designs which are both highly optimized and designed to have a very low risk of failure. (And Mars 2020 is being developed in the same way, despite claims that it’s just Curiosity with different instruments.) That whole approach and the mindset that goes with it are a result of high launch costs. It’s a matter of only having one opportunity, knowing there will be no second chance, and wanting to get the most out of it.
Lowering launch costs by a factor of two could easily change that. A functionally identical spacecraft, but with higher mass and lower reliability, would cost much less. And if the launch only costs $25 million, that starts to look like a more reasonable choice.
> That whole approach and the mindset that goes with it are a result of high launch costs.
That’s a childish myth propagated by people who haven’t really looked at spacecraft design and tried to understand why things are built the way they are, or why the costs are what they are.
Simple fact check: GEO comsats are effectively serial production spacecraft, with slight variations between sats on same bus. Building them is still insanely expensive, that’s why sat building market revenue is 3-4x of launch markets.
No, that’s the considered opinion of someone who has worked on planetary missions for a quarter of a century, worked on numerous design studies, and seen how low cost missions differ from high profile, class A and B missions. Yes, communication and other serial spacecraft do tend to cost a few times more than the cost of launching them. That’s probably the right balance between launch and production costs. But if the launch costs go down by a factor of two to five (numbers you mention) that balance implies satellite costs would go down by a similar factor. With lower launch costs, packing in the greatest capability possible and assuring very high reliability wouldn’t be worth the cost. And, if you look at other industries, a few hundred million isn’t insanely expensive.
1. I didn’t say “launch cost”. BFR is not just a launch vehicle, the upper stage is a spacecraft/lander by itself, this is why they needed the “Shiny prop”, to test the landing method for the upper stage/spacecraft/lander. So BFR is not just going to reduce launch cost, but spacecraft/lander cost as well.
2. And BFR is not just going to reduce launch cost by 2x or 5x, but more like 50x to 100x by $/kg.
3. Finally, you’re confusing development cost and unit cost and thus comparing apples to oranges. When you launch a deep space probe, the launch cost only include a tiny percentage of the development cost of the launch vehicle since the launch vehicle design is reused many times. But the deep space probe’s cost included both the full development cost and the unit cost.
>And BFR is not just going to reduce launch cost by 2x or 5x, but more like 50x to 100x by $/kg.
Sounds like magic. Also sounds familiar, we have heard this before more than once, see history of Space Shuttle for instance
> But the deep space probe’s cost included both the full development cost and the unit cost.
Deep space probe and satellite buses also get reused.
Commercial satellite are, sometimes, reused. But more often they are a fairly modular design and the “same” bus really means a similar one. For deep space missions, I can’t think of a single example of real reuse since Voyager.
They frequently use the same design for _components_ and a bus _based_on_ a previous one isn’t uncommon. But there is an easy way to see that there is no real reuse of a previous design. Those projects still have a full phase A and B, with full preliminary and critical design reviews (PDR and CDR.) If they were really reusing a past design, they could and should have a very abbreviated Phase A and a delta-PDR which only covered the changes. If it really was a complete reuse of a past design, they might even be able to hop straight to CDR. (Well, presumably the payload would be different, so it would need the full process, but not the spacecraft.)
> I can’t think of a single example of real reuse since Voyager.
Really ? You haven’t looked much then. MER-A and MER-B. Mars Polar Lander -> Mars Phonenix -> InSight
> If it really was a complete reuse of a past design, they might even be able to hop straight to CDR
That’s chasing strawmen, as instrument suites aren’t supposed to be identical in the first place.
I’d forgotten MER-A and -B. Those were actually built at the same time, so technically that isn’t an example of _re_using a design. But I did mention Voyager, and the same thing applies to them.
Since the decision to fly two MER rovers wasn’t in the original plan, it would be interesting to see what the cost estimates were like. Someone must have the numbers on what flying one or three would have cost.
MPL, Phoenix and InSight are not a reuse of the same design. That was my point. They are similar designs, and designs with substantial heritage. But if, for example, the telecom system was _identical_ (as in truly reusing the same design) why did they have a system-level PDR and CDR for it? Or spend any money in phase A and B? The answer is that the designs were _not_ the same, just similar, and that makes design reviews necessary.
But your comment about skipping to CDR makes no sense to me. Why can’t the instrument suites be identical? Mars, for example, is a very diverse place (all real planets are.) If the costs were low enough, sending a few more, completely identical, MER-C, -D and maybe -E, landing in different locations could be justified. You seem to be making the circular argument that low-cost approaches are impossible because missions are “supposed” to be developed in a high-cost manner.
In any case, and even if you change the payload, the process could be massively streamlined. If the payload really were the only thing to change significantly. No Phase A and B work on the systems which did not change. Only do system level PDRs for the payload, and a mission level delta PDR focused exclusively on how the payload changes could impact the unchanged systems. Limit Phase C work on the nominally unchanged systems to any unplanned impact on them which turned up in the mission level delta PDR and limit those system level CDRs to the changes.
NASA has, actually, done that before (although it’s been a long time…) The fact that this was not done with Mars 2020, for example, shows that it isn’t really the same design as MSL. Despite the initial and public claims, they actually made significant changes to most of the major systems.
> If the costs were low enough, sending a few more, completely identical, MER-C, -D and maybe -E
Very rapidly diminishing returns and huge opportunity costs of doing something else with the same resources ( time, money and talent ). Also won’t give order of magnitude savings, although nobody will argue that “per unit” costs will be spread out better.
Look, I’m not arguing high volume of activities will bring per unit costs down, however nobody in the world has budgets to support tons of higher volume of activities.
BTW, I forgot to mention Mars Express and Venus Express. Again as good useful reuse that could be pulled off, but in the end it doesn’t change the cost equations radically
By a very charitable estimate, we’ve seen less than a millionth of the Martian surface from the ground and at a resolution of under 0.3 meters. And only about 3% of it at 0.3 meter resolution from orbit. If you consider adding to that “very rapidly diminishing returns,” I’m not sure what to say. Should we also stop funding field geologists on Earth? (Note I did write _if_ the costs were low enough a MER-C, -D, etc. would be justified. Your argument presupposed they never could be. Your conclusion is just a restatement of your assumptions.)
Mars Express and Venus Express are very definitely the same design. They are closely related, but the payload is completely different, VEx had about two and a half times as much power as MEx, the thermal design very was different, etc. They were’t the same because, well, the were different. Reusing the same design means going directly to fabrication without any new design and development. That is, for example, what the Iridium NEXT constellation did with 75 spacecraft.
And the argument about a sufficient volume (I think you mean number) of missions being unaffordable doesn’t make sense. What _if_, hypothetically, an order of magnitude cost reduction is possible and the budget did not change?
That would mean an order of magnitude more missions would be possible, and that’s the sort of numbers to really take advantage of a standard bus (as in the same, as in no design and development.) It would probably make sense to have the spacecraft and the payload interface standardized but with mission specific instruments. Based on your own comments, can we agree that this would provide a factor of two reduction in cost?
Given low costs and a high flight rate, you could reasonable relax strict requirements on reliability. The missions are no longer unique opportunities and wouldn’t represent a decade or two of hundreds of people’s careers. Based on a National Academies study, the difference between Class A/B versus C/D missions is 90% versus 80% probability of success. So _if_ we could get to that order of magnitude drop in cost and increase in flight rates, we’d end up with almost nine times as many successful missions. Probably more, since if it also involves using a truly standardized bus, experience from past missions would increase reliability. Based on conversations with lots of people in the field, the difference is also a factor of two to three in cost.
Let’s see, that’s a factor of two from using a standardized bus, and a factor of two or three from flying Class C/D missions instead of A/B.
I’d argue that a further factor of two or three in cost could come from less optimization. With high flight rates, the pressure to cram in every bit of capability and get the most out of every gram of payload would be reduced. That (and using a standardized bus) would make the missions a bit less capable. But I’d also argue that this is made up for by the high flight rates. There would be more opportunities, on a shorter time scale, for follow-up missions. And they could have instruments tailored to observe things a previous mission discovered. A capable instrument designed for what you _know_ you want to measure is actually better than an _extremely_ capable instrument designed based on guesses about what you want to measure.
Let’s see… That’s a factor of two, a factor of two or three and another factor of two or three. That sounds like an order of magnitude reduction in cost. That is also a very different way of doing things, compared to current practices. That doesn’t mean it wouldn’t work or would be inferior. It also depends on launch costs also going down by an order of magnitude, and no one of the above ideas would work on its own.
The grasshopper is a shiny prop, but hard to fathom what does this have to do with reusable 1st stages.
https://uploads.disquscdn.c…
“The grasshopper is a shiny prop, but hard to fathom what does this have to do with reusable 1st stages.”
Nothing, zero, zip, because it’s a reusable 2nd stage test article.
Wondering why some of you are so dour about this. Has anyone ever built a vehicle – this large – this fast? Has anyone ever tested a rocket stage that will land on another world? THAT is what he is doing.
When thinking about Starship’s various capabilities, one of which is a huge Tanker, one has to think “ACES who? “
Exactly! And it could get even wilder,
Musk also stated that they’re building “sections” of the full-size orbital Starship Mk-1 in LA, which ISTM leaves open the possibility of those sections being shipped to Boca Chica for another “water tower” buildout. That, or the big hanger the tower crane is being stored in becomes a factory.
IF either happens, international geek-gasm in 5, 4, 3, 2…
Bingo! Folks are so stuck in the old way of doing things its impossible to see the magnitude of the paradigm change that is occurring. It would take NASA longer to just do the viewgraphs then it took Elon Musk to build the Starhopper.
It is part of a SYSTEM not a stand alone article.
Unless their plans have changed (quite possible), the Starship would also be a stand-alone vehicle for suborbital transportation. Of course, I have my doubts on that idea, but that’s a different matter.
Well put, V.
BTW: That model was also used in “Flight to Mars”. 🙂
https://uploads.disquscdn.c…
I would like to know whether Starship will go to a pressure supported structure. Despite the hassle of maintaining pressure, it proved a reliable and lightweight strategy for the original stainless steel launch vehicle, the Atlas, and the use of stainless allowed the tanks to be integral with the skin. it would also eliminate the wrinkles.
From Musk’s tweets. It is non balloon propellant tanks with autonomous pressurization from broil-off propellants for propellant settling. He also stated that winkles will be smoothed out with thicker skin for the tanks.
I’ve taken issue many times with the appearance of ISS and with other vehicles, mostly because form following function does not imply ugliness.
Here we have a lovely demonstrative case that reminds me of the first 747.
Musk said Starship and SH will be self-supporting.
As a lifelong reader of Sci-Fi who lovingly remembers all the golden age magazine covers, this design brings tears to my eyes! Bob Heinlein would have loved it. Asimov and Clarke would have too. I really don’t care what he does next, after seeing that rocket ship Elon can do no wrong in my book. John Campbell you passed too soon, but from wherever you are now, I know your smiling down on this. Elon Musk, our modern version of Delos D. Harriman.
I am just making an observation here; we have seen photos of a steel structure that resembles a reusable rocket. But I have not seen any photos of tanks, bulkheads, plumbing, wiring or avionics going into the structure. Just a shell. And if those bells at the bottom of the structure (or vehicle) are real engines, would they not have end caps on them to protect from the harsh salt environment down on the Gulf? Or even be covered and purged?
Seems like there is a lot work ahead before this ever flies (reminiscent of the first 787 Rollout).
The engines are the Raptors used in testing, they are being used as sizing and fitting pieces. The actual flight engines were shipped from Hawthorne recently and will be slotted in once the rest of it is built.
As for “plumbing”, have a look at the schematic layout of any modern rocket, all the plumbing is in and around the engines. The rest is just the tanks, the bulk of the rocket body. There’s nothing we would see.
As for the “bulkheads”, there’s only two, one in the bottom section, which we didn’t see being built until it was well along (no-one knew what it was at first, so no-one cared), the intertank dome was seen being built but it hasn’t been installed yet. The upper dome is the upper section itself. Wiring/avionics, if installed yet, would be internal, and not visible in the images available.
(Remember, they started the build on a concrete surround that prevented people from seeing what was going on inside. Tonnes of material was craned into the bottom section that we didn’t see in detail. We have no idea what’s inside, only that it was responsible for most of the worker activity.)
As for “tanks”… uh, it’s a rocket vehicle. It basically is just tanks with engines strapped on. Tanks are what we’ve been watching being built.