I've been reflecting on the Starship program the last week and one thing has become obvious to me. SpaceX is enjoying the freedom to try and fail in a way they couldn't with Falcon 9.
— Everyday Astronaut (@Erdayastronaut) September 5, 2025
Doing anything "experimental" on the Falcon 9 was risky because it was SpaceX's only source of…
14 thoughts on “Starship Thoughts”
Comments are closed.
I’d also say SpaceX is testing and strengthening Starship in a way NASA never tried with any of their vehicles. Some may think China will beat us to the moon, but China will know nothing compared to SpaceX in regard to heatshield capabilities to a level needed to exploit the moon or any other body in Space.
I think you’re right. I think China has a lot more to figure out than most people assume. And there’s the “invincible opponent” syndrome working here.
It won’t be as easy for China as everyone thinks.
I remember all the Ranger missions to the Moon, before the Surveyors landed. The more other countries find out how hard getting to the Moon successfully is, the more impressive the US successes become.
Even recently many moon landings failed at the last minute.
I saw a YouTube video, that got “ratioed” in the comments, that SpaceX hasn’t yet solved the mass-fraction problem of a fully reusable orbital rocket, namely that Starship can carry a payload other than its thermal protection system.
The claim is that these tests we are seeing are an effort to see where Starship can shed enough weight to carry a useful payload, remembering that even with expendable systems the payload is a low single-digit percent of the launch mass, and at the same time not burn up or break up on reentry.
So the problem is not whether Starship can be launched and recovered and reused, but if whether this testing process, which seeks to see what mass reduction they can get away with, will result in a useful amount of orbited payload in fully reusable mode?
Anyone know anything about this? Does SpaceX have a design that can orbit a specified payload, and the tests are to wring out the bugs in this design? Or do they not yet have a design, but they hope to converge on one through the testing and refinement of the vehicle?
These kinds of anti-SpaceX things about on the InterDweebz. It was true of Starship v1. V2 has a payload of about 25 metric tons above the mass fraction and is the testbed. V3 is a redesign around the Raptor 3 engine, and the proof of the pudding will be in the eating. v4 is like the current “finalized” Falcon 9 Block 5.
The OP video is wrong, I think. I’d call the complete redesign of Falcon 9 from 3×3 engines to 8 + 1 about as experimental as it gets.
SpaceX took a wise approach to Falcon 9 development. They incrementally improved the rocket through several versions as they improved the Merlin engines. They introduced super chilled propellants to further improve payload capacity. After testing basic recovery technologies with Grasshopper, they tested it during several commercial launches. It took quite a few tries before they were successful at landing a booster. Based on lessons learned from studying those boosters, they evolved the Falcon 9 into its ultimate V5 version. They also implemented payload fairing recovery to further improve affordability and accelerate launch rates. Even then, they’ve continued to refine trajectories and implemented reduced size upper stage nozzles when applicable to continue improving operations. They also switched from aluminum to titanium grid fins and improved landing gear legs to reduce refurbishment costs. All of this while increasing launch rates year after year.
In the comments, Dodd brings up a good point. It doesn’t matter if China goes to the Moon before we return. He points out there will be quite the contrast between a tiny capsule and a skyscraper and what that means for what each country can actually do on the Moon.
It would be nice to beat China but it is more important to focus on fulfilling our strategy.
China can use its “tiny capsule” to start a major lunar exploration program including a small moonbase, while waiting for its “skyscraper” to be ready/ See the planned “Apollo Applications Program” that would have been done while waiting for the “Space Transportation System.” The trick, he notes, is to actually *build* it. Dismissing the tiny capsule shows he’s not as smart as he thinks he is.
Note AAP was restricted to Skylab and STS to Shuttle. Two monumental fails in one! Such a deal!
I don’t think he is dismissing it so much as noting that Starship will be beneficial when it comes time to do the building. I took his comment as that it is more important what you do when you get there vs when you get there but as you point out, you still need to get there.
I don’t think Tim is “not as smart” at all. Or, if he is, it’s in assuming that the PRC has anything resembling a good shot at getting a crew to the lunar surface before we return.
First, to do that, the PRC still has to be a going concern in 2028 or 2029 or 2030 or whenever it is capable of first attempting a crewed Moon landing. This is hardly a cinch bet given its financial condition and other problems – mostly self-inflicted.
Second, there is no prospect of a “major” lunar exploration program by the PRC for a minimum of several years to a decade following any notional initial landing.
The initial PRC lunar crewed mission architecture uses two large, expendable LM-10 rockets for each visit, one to launch the lander and another to launch the Earth-to-lunar-orbit-then-Earth-return crew capsule. Each of these rockets is a triple-core vehicle using 5-meter cores, 21 booster and first stage engines, two 2nd stage engines and three third stage engines. Long March 10 will not only resemble Falcon Heavy, it will have almost as many total engines – 26 vs. 28.
This is a lot of large rocket parts that need making. The only other 5-meter diameter rocket currently produced in the PRC is the Long March 5. The LM-5 uses 12 total engines, two on each of four strap-on boosters, two on the core stage and two more on the second stage. The boosters are liquid-propellant units with 3.8-meter hulls – about the same diameter as SpaceX Falcons.
The PRC has never built and launched more than three LM-5s in a single 12-month period. To launch a single crewed Moon mission will require the construction of more than twice the tonnage of 5-meter hull structure needed to launch three LM-5s.
It will require more than twice as many engines as well. Unless there is major new factory capacity about to come on-line capable of 5-meter hull and large engine fabrication, the current PRC capacity to fly Moon missions looks roughly comparable to our own limits for SLS-Orion-based missions – namely, one every other year if everything goes right and less than that if it does not.
Thus, if the PRC wants to fly a crewed Moon mission as soon as 2028, it already needs to be working on the two rockets needed for a preliminary uncrewed test mission. I don’t think it’s to that point of capability yet as there has been some talk of increasing the size of the Tiangong space station and that will take at least a year’s production of LM-5s to do.
Of course the PRC could always copy the Artemis 4 page from NASA and launch a crew on pretty nearly virgin hardware, namely the second-ever LM10 to fly – assuming the first LM-10, with the lander as payload, was a success. Perhaps the PRC would take such a risk if it looked as though Artemis 3 could actually fly in 2027 or 2028. There is the Apollo 8 precedent, after all, impelled by sightings of the N-1 on its pad.
But after a first crewed lunar mission, it could take two or more years to mount a second. The first mission is only supposed to spend six hours on the surface – not even as long as Neil and Buzz did back in the day. And the lander design incorporates a crasher stage to handle most of the descent. The limited on-surface endurance, the lack of much downmass capacity beyond the crew and their suits, and the crasher stage all present big problems anent setting up any sort of long-term presence.
The crasher stage, in particular, presents a lot of problems for return visits to the same location. A second mission to the site of the first landing could easily result in taking out the PRC equivalent of the Apollo 11 landing site.
The only way the PRC is going to be able to land significant mass – other than crew – on the Moon is to graduate from the marginal pair of LM-10s to a reusable LM-9. Lord knows when, or even if, such a beast will ever fly. The coming decade is not going to be a good one for the PRC. More existential priorities may well intrude in the interim.
To be pendantic about the history, wasn’t the concern about the “heavy Zond” launched from a Proton rocket (almost but not completely unlike our Titan III is capability)?
I read the “zond” means “probe” as in a space probe, and the Zond name applied to unmanned, robotic missions to the inner planets. Because of their penchant for secrecy, but for state security and also for saving face if a mission fails, they gave the Zond name to unmanned tests of a crewed vehicle for the loop-around-the-Moon-without-entering-orbit mission. We called that vehicle a “heavy Zond” because it was different from the much lower mass genuine inner-planet space probe vehicles.
The Soviets tried mightily to “scoop” the US on a non-landing lunar encounter mission because their N1 moon rocket was way behind schedule and probably didn’t have enough payload for a practical lunar orbit rendezvous mission. Towards this end, they flew the unmanned test of the loop-around-the-Moon mission with a Proton rocket launching a “heavy Zond”, and each time it “burnt up, real good!” on reentry because they hadn’t quite solved the guidance requirements for an Apollo-style skip reentry to manage thermal loads for a reentry with the higher velocity than a low-Earth orbit mission.
Did the Soviets finally get an unmanned heavy Zond mission to work, but by that time Apollo 8 happened. The Soviets didn’t see much point in continuing with the heavy Zond program after that because what’s the point of being second to reach the South Pole or climb Everest or whatever the goal was. I know this “‘tude” on the part of the Soviets from hearing from an actual Soviet giving a talk at the State Historical Socity of Wisconsin on the University of Wisconsin-Madison campus.
What an actual Soviet was doing on campus during the twilight of the Soviet Union was that he had connections to an actual Egyptian in our Biomedical Engineering department. He was Valentina Tereshkova’s boss, and she was supposed to be the speaker which was how the event was advertised. I was so hoping to get her signature next to her mission in my copy of Oberg’s Red Star in Orbit, but she didn’t show because she was busy in their version of Congress trying to keep the Soviet Union together.
I was then going to give my copy of Oberg to her boss Konstantin Froulov, if I have his name spelled somewhat right, but he glanced at actual Soviet KGB minders sitting in the front row and diplomatically refused this gift, stupid me not realizing this was a faux pas on my part to offer it with KGB dudes sitting there and not saying a word. He did mail me copy of “Korolev”, written in Russian which I can barely read, but is a prized book in my collection.
Perhaps “heavy Zond” was more of an impetus to Apollo 8 than the N-1. My knowledge of Russian manned and “manned adjacent” lunar and lunar-ish space activities during the Moon Race is mainly what was public knowledge at the time and what incidental information I have encountered since. I’ve never made any concerted effort to delve deeply into the then-non-public history of that era of Soviet space activity.
My point in bringing the subject up at all is the matter of institutional incentives and worst-case-foremost mindsets in such a milieu – on both sides. Apollo 8 was risky to the point of borderline insanity. We only essayed it because we were fearful the Soviets were much further along with their lunar program than they actually were.
This works in both directions. A lot of what the Soviets did during that period was because of their superior knowledge of the state of our program vs. our knowledge of theirs.
I do not want to see the US taking any Apollo-8-level chances just because too many people in the government have an unwarranted case of the yips about where the PRC program actually stands and what it can do and when.
As with the Soviets back in the day, the present-day PRC knows a lot more about the state of the US program than we do about the state of theirs because we conduct ours in public and they, by and large, do not.
As I noted previously anent production capacity, the PRC lunar program has very real limits on what it can do and how soon and how fast it can do that.
Thus, it is hardly far-fetched to imagine the leadership of both the PRC nation and the PRC lunar program getting a case of the yips not too far down the road comparable to what the US government and NASA experienced back in the day and which resulted in the dice roll that was Apollo 8. That could easily result in premature efforts on their parts that go tragically awry.
But the US program is already plenty Apollo-8-ish as things stand. Artemis 2, as currently planned, is far riskier than is prudent owing to question marks about its heat shield and life support system. We “can’t” take time to test it more because the SLS-Orion stack’s glacial production cadence has NASA and the rest of the US gov’t. convinced that taking time to do so would lose us any chance to Beat the Chinese[tm].
But if – because God looks out for drunks, fools and the United States of America – we are successful in both launching and getting the Artemis 2 crew back alive in the first half of next year, and Starship progress is also significant by that time, then the psychological ball would once more be in the PRC’s court. If it looks as though we might actually pull off Artemis 3 in 2027 – or even in 2028 or 2029 – the PRC, assuming it is still around, could well feel impelled to dice with fate also – most probably by attempting an initial manned landing mission without a preliminary full-dress unmanned rehearsal. This could be impelled by the same sort of production capacity limitations that make it “impossible” to do reasonable testing of the SLS-Orion stack before sending people Moonward on it.
As Mark Twain noted, “history doesn’t repeat, but it often rhymes.”