FAA stays grounded in reality as SpaceX preps for takeoff SpaceX supremo Elon Musk expects the next Starship to be ready to launch in three to four weeks following the monster rocket's latest failure. The timeline could charitably be called "optimistic" – something doubtless preying on the minds of NASA managers worrying about the Artemis Moon landing mission and the Starship …
Probably... (speculation follows)
They perhaps shaved structural margins a bit too much on the booster. The structure vs weight tradeoff is where a rocket engineer earns his paycheck, especially in the first stage.
Or perhaps the maneuvering was just too sporty and they need to tone it down. Perhaps you can't just flip a SuperHeavy booster as hard as a Falcon 9.
Maybe they turned up an issue in the flight termination system or guidance in the second stage. That could be hardware or software. Maybe the FTS overdid it instead of underdoing it like last flight.
They certainly fixed the pad issue, the issue with not all the first stage engines firing, the flight termination system, and the staging, so I'm sure they'll have some sort of solutions to what broke on this flight.
And then we'll see what breaks next! Wheee!
My money is on the heat shielding. You could see tiles coming off. Of course the Space Shuttle orbiter had the same problem and they lost a bunch of tiles too. Difference there was that the orbiter was manned, so it was a slight bit more critical.
The next one will probably make it to orbit, but we'll have to see how many separate pieces re-enter over Hawaii.
You can sure tell they're big fans of Kerbal Space Program.
On the booster flip. I suspect the fuel sloshed, which took the fuel out of the lines, then the engines went off. Another option is the blast on the top of the booster from the second stage opened the top of the can .
The heat shields coming off. Would be interesting to see how far that survived re-entry. Stainless is a very different material to aluminium. But you are right, tiles were clearly lost. The question is why the second stage didn't make it to orbit. I've not seen any real explanation of that.
This post has been deleted by its author
>Another option is the blast on the top of the booster from the second stage opened the top of the can .
The heat from the hot staging will have also increased the pressure in the fuel tanks. If that was higher then they expected it could have contributed to the engine failures. The bottom of the second stage will also have experienced higher temperatures during the first moments of hot staging. It will be interesting to see if the stick with the hot staging procedure with the next flight as it was a rather late change to the design.
On a similar note, the booster engines shut down in stages, probably to reduce the levels of deceleration "pushing" the fuel upwards, potentially either cause pressure issues or hydraulic shock, and so was a mitigation. But I wonder if the thrust from the Starship engines doing a hot staging "pushed" back on the booster, causing an increase in deceleration rate and resulting in fuel pushing "upwards" and/or hydraulic shock" to levels more than expected? Whatever caused the problem, it does seem to have been an issue with fuel feeds to the engines as they seemed to go out in a cascade very quickly after re-light.
This post has been deleted by its author
An analysis I saw of the telemetry showed the booster decelerating markedly when the second stage engines fired for the separation. This would have created negative G's in the booster's propellant tanks disrupting the smooth flow to the engines. The booster had reduced thrust considerably by then down to just three engines. Perhaps that was too much.
its not ideal to lose the tiles as it impacts the reusability of Starship, but its not loss of vehicle type problematic at this stage, unless entire chunks start dropping off, stainless steel is very different to aluminium, its got a much higher melting point so is expected to survive some tile loss. It was notable even the day before launch they were doing some tile repairs, and there looked to still be some anomolous placed tiles, maybe on weld lines, maybe defects, maybe even SpaceX were experiementing to see what might happen if they lost tiles, or which fixing method works best, its ultimately a test vehicle.
The vehicle's shape also helps mitigate re-entry heating. The bow shock sits closer to more tightly curved surfaces (e.g. the leading edge of the shuttle's wings). Because Starship is very big and curves very gently, the bow shock should sit a lot further from the surface of the vehicle, which should help reduce the heat transferred to the vehicle. Tiles missing from these areas might not matter so much. Any tiles knocked off from around the control surfaces might be a much more serious issue.
There is some fascinating footage taken by the YouTube site Astronomy Live from the Florida Keys https://www.youtube.com/watch?v=CTcSMh4VYow showing the Starship tumbling out of control before it exploded (I assume by the fight termination system). There are also two fairly small and one large release of gas clearly visible in the live launch footage and a corresponding loss of oxygen pressure. I noticed these releases as I watched the launch live and thought that these were probably not a good sign. It is also interesting to note, although it may have nothing to do with the failure, that the launch was held up for a short time at the 40 sec. hold for a problem with the second stage oxygen pressure. I'm sure we'll learn more as the telemetry is examined. It is good to see that the launch pad suffered only minor damage, mostly being attributed to the "sprinkler system". As I noted in a previous post there was a massive reinforcement effort on the launch pad https://www.youtube.com/watch?v=09DDpHdIYgU that is a long watch (25+ minutes) but very interesting. This was a major contributor to the launch pad staying in one piece.
There is another video I believe on the same YouTube channel that was an early analysis of the explosion, It shows that the explosion left the crew/cargo portion of Starship intact, the rest way destroyed. The video of the tumbling Starship clearly shows that it was still intact.
Common Sense Skeptic has a good video breaking down the launch here https://www.youtube.com/watch?v=ka5id7ZQKL4.
He also uses footage from Astronomy Live that shows the top portion of Star Ship tumbling towards Florida after it exploded. That could be an issue for SpaceX as it's a major chunk of spacecraft that is falling uncontrolled towards a population centre.
"That could be an issue for SpaceX as it's a major chunk of spacecraft that is falling uncontrolled towards a population centre."
That chunk was well away from land, but that was the case this time and why they have to thread the trajectory carefully so the vehicle and debris paths aren't intersecting land.
"There is some fascinating footage taken by the YouTube site Astronomy Live from the Florida Keys https://www.youtube.com/watch?v=CTcSMh4VYow showing the Starship tumbling out of control before it exploded "
That video shows the upper half of Starship tumbling post explosion. If that was the result of the FTS, they still have loads of work to do on that system if such a large section remains. Obviously, SpaceX had no clue about what was happening with Starship at that point going by video of the command center and everybody at the time looking like everything was normal and they had just had an expected loss of signal. Since a person in Florida was able to get imagery, a question must be raised asking why SpaceX couldn't have deployed any sort of tracking especially since this was a test flight and as much data as they could get would be critical in this sort of case. Puerto Rico is a US possession so having tracking there wouldn't have required any sort of complicated issues over shipping the telemetry equipment. Of course, Elon isn't such a hero there. NASA might have even be roped into using the TDRS sats since SpaceX is supposed to be developing a lunar lander based on this design.
SpaceX's own feeds, as feeble as they were once again, give some indications that several off-nominal things happened with Starship. The booster had an explosion in the engine bay just a second before exploding. Was that FTS, or was it a tank rupture at the common bulkhead? It could have been FTS and SpaceX belatedly has said it did activate, but a frame by frame look could be interpreted the other way quite easily.
If you watch the booster data at separation, you will see that there was some deceleration (likely caused by the blast from the second stage engines). This would have causes the fuel to move away from the bottom of the tanks, which could have lead to the engines being deprived of fuel for the boost-back burn.
SpaceX had ullage cameras on the Falcon to see what happens to the propellents during flight, and I wouldn't be surprised if they have the same for Starship. If they do, it would be interesting to see what happened at staging.
They were aiming for 50% thrust on the three booster engines at stage sep ... If it was a gas phase ingestion issue, it could be as simple as increasing stage 1 to 60% and reducing the second stage sep power by a few percent to balance the remaining fuel load.
On the other hand, it seemed to me that the first stage issue actually began at the boost back turn when the first ring was reignited. Hypothetically it could be that (as someone said) things were a bit too sporty or it could be that an engine/fuel line blew out and progressively took engines on one side with it before the FTS cut in (my feeling). Be interesting if it was a major oxygen feed failure and is the same problem that later took out stage 2 ...
Engine development was a generation ahead of launch for IFT1 and IFT2. Some of the changes were for ease of manufacture but the reduced part count improves reliability. Previous engine changes may fix problems that will be identified during the mishap investigation.
This was SpaceX's first full attempt at hot staging. Now they have real flight data to compare with I expect their simulations will improve. It might be possible to fix hot staging and the boost back burn with a software update. If not, it did not take long for welders in Boca Chica to install new baffles in the tanks to get the belly flop landings to work.
For IFT2 the heat shield tiles over the circular welds were glued on. They are attached more securely for the current occupants of the rocket garden. The attachment is now tested with a suction cup device.
I am sure SpaceX can have another Starship ready in four weeks that will explode later in the flight. There is no shortage of Starships and such explosions would be worth it for the data collected for IFT4. In real life, the FAA are not staffed, funded or resourced to keep up. I would be surprised if they are ready in two months and not surprised if it takes three. SpaceX will not sit on their hands during the wait so there will be more improvements for IFT3 and IFT4.
The tone of the article provoked responses, some of which contained information I had not found elsewhere. It is a part of how journalism has functioned for decades.
Although to be fair, electric bikes are far more likely to catch fire than electric cars, for example there's been 17 deaths in NYC from e-bike fires in 2023, in London there's been 137 fires and 3 deaths in 2023 (well, up to mid September).
"While SpaceX fanatics are trying to paint the second Starship Super Heavy launch as a success..."
While habitual Musk / SpaceX detractors are determined to paint it as a failure. Seems to be a lot of that about. The truth is somewhere closer to the middle ground, as usual.
This latest test - key word being test - of a highly experimental vehicle showed substantial improvement over the previous test, and those improvements were made in a very short time. That is impressive, and certainly indicates a degree of success, just as no improvement, or even deterioration, would have been a failure.
That said, there is clearly room for further improvement. Indeed, the whole point of the test, it's predecessor and successors, is to identify things that don't go to plan and facilitate remedial action. From that perspective, the test was also successful - two RUDed stages will provide lots of opportunities to identify things that didn't go to plan. Even so, the test did not achieve everything that was planned (though I doubt anyone expected that it would) and those parts of the plan therefore failed by definition.
Naturally, the FAA will want to know what happened, and how that outcome can be avoided in future, though not as badly as SpaceX will want to know, I expect.
Had this been a real mission, resulting in the total loss of not one but both stages of a tried and tested, flight proven vehicle, then the gleeful proclamations of failure would be amply justified. But it wasn't. It was a test, an experiment, some aspects of which had never been done before. I'm sure that most of the failurites understand the nature and purpose of experiments perfectly well, but are unable to resist the temptation to poke at SpaceX / Musk.
Perhaps, instead of claiming success or failure, they could agree that progress has been made, and will likely continue to be made, towards improvement in human spaceflight capability. That can't be a bad thing.
It was a success - but it wasn't as much of a success as it potentially could have been.
They've demonstrated hot staging - deceleration is generally bad for fuel staying in the pumps, but that's a relatively trivial fix, just don't throttle those three engines down as far and you don't end up decelerating... the booster is important to reuse, but also it's never the primary mission.
We still don't know what actually went wrong on the ship, but the booster demonstrated the IFT system worked as well (which is a significant improvement on last time).
There are a few bits of hardware very close to ready to test, and stage zero looks in remarkably good condition.
"They've demonstrated hot staging"
That's not saying much. Yes, they hot-staged, but they don't know if that injured the booster in a way that led to its failure.
"We still don't know what actually went wrong on the ship, but the booster demonstrated the IFT system worked as well"
Maybe, maybe not. After seeing CSS's breakdown and frame by frame view of the booster just before it exploded, a tank rupture and subsequent explosion is also a viable explanation. It's already been shown that an explosion at the aft end of Starship will cause a buckling at the common bulkhead of the tanks which could also mean the Booster is also susceptible to that failure mode.
"That's not saying much"
Have you managed it yet, on a rocket this powerful?
Demonstrating hot staging is a significant acheivement.
We do know that the booster survived for quite a while, and SpaceX have much more data than we do, they're not relying on random youtube streamer's equipment (impressive though much of it was)
"Flight two of the Starship Super Heavy improved on its predecessor, but only if you imagine that your self-driving electric car did not crash into a tree and catch fire. Instead, it made it a bit farther down the road, and then caught fire."
Now I'm NO Musk fanboy; the guy is an egregious wankpuffin with an ego the size of a small planet. But I'd have to call that a rather negative tone.
The first launch was a complete clusterfsck.
The latest attempt was an 'almost worked'. The booster apparently performed pretty much perfectly, but the attempt to boostback and recover it resulted in loss of control and was terminated by a manual RSO destruct command, AFAIK. The upper stage apparently performed pretty much perfectly, but was terminated by a computer-initiated auto-destruct command for no particularly obvious reason.
When you think how many Falcons they blew up in their early days, they actually seem to be on a pretty good trajectory with this latest toy.
How many Falcons do you think were blown up? The first 3 Falcon 1 launches were failures, then they got it right and promptly retired the rocket.
In 275 launches, the Falcon 9 has had one complete launch failure, one partial launch failure (one of 2 satellites didn't make the correct orbit). One other early version blew up on the pad. Every other launch has been a success.
If you mean how many Falcons blew up on landing attempts, then yes. There were a few on tests after the primary mission was a success. Then they got it right and land every time now. The last attempted landing that failed was in February 2021. Who else can land an orbital class rocket?
"When you think how many Falcons they blew up in their early days, they actually seem to be on a pretty good trajectory with this latest toy."
And, of course, not forgetting the number of early Starship "hops" that ended in RUDs too, but to be fair, this is effectively two separate rockets, each with the own foibles and problems doing things that have either not been done before or not been done on this huge size of vehicle and then combining them and their "unknowns" into one big launch stack :-)
I wasn't surprised the first launch ended in explody bits, but was surprised at the amount of damage to the launch pad and tower. I wasn't *too* surprised at the second lauch booster explosion and was really getting hopeful for Starhip as it just seemed to keep going up and was sadly dissapointed but hugely surprised that it too didn't fully succeed.
I've said it before and I'll say it again. SpaceX seem to operated a little like Kickstarter projects. There's a goad you want and need to reach, and then there's "stretch goals" that would be nice to reach. On the whole, SpaceX do generally seem to meet their main goals and sometimes reach some of the stretch goals. If nothing else, it's a lot more existing than spending 10 years, going massively over budget and massively of target date for a "first time success" :-) And as someone else suggested, a first time success could mean it's over engineered and over weight as well as over budget!
As with their landing attempts, they kept trying until they got it right. Then they improved the process to the point that it is now more surprising when it doesn't land successfully. The last 162 attempted landings have been a success with the last failure in February 2021.
Since they already have the next few Starship stacks built and ready to fly, they will get this right soon as well. With the SpaceX methodology, design the rocket, fix all the known issues then launch it to find any unknown issues. Fix them and try again until all unknowns issues are known and fixed.
It is far cheaper and quicker to test things to destruction and fix what goes wrong than to try and anticipate everything that could possibly go wrong before your first launch.
"And as someone else suggested, a first time success could mean it's over engineered and over weight as well as over budget!"
It could mean that or it could mean good engineers. It's said that Henry Ford had his people looking at what broke on his cars to determine which parts never broke to see what was too good. It's not bad to do a bit of over-engineering, getting a system going and then seeing what you can dial back to save mass and cost. If it does the job, it might not even be a necessary thing to change anything. The plan is to get these rockets back so there will be ample opportunity to evaluate them provided they can get to the point where they come back and can be examined in one piece. There will come a point where Elon will run out of willing investors to keep pumping money into the company and they have a large number of successful launches to get to just to be able to fulfill the lunar lander contract. It also can't be ignored that they still have to perfect ship to ship cryogenic fuel transfers and before that, come up with proven connections they've been paid to develop. The launch pad can't "just" survive, it needs to be robust enough to launch 5 ships or more per day to support the mission architecture they've dreamed up. After the 1st grain silo exploded, Elon commented that the OML loses 5cm/s of steel from the rocket blast. Several 4-6 second exposures per day would require the top parts of the structure and possibly the bidet to be exceptionally thick and quickly replaceable.
I agree, the tone of that statement was definitely off. It's a little disingenuous to compare a highly experimental test article, being flown as a test, to a commercially available, publically used vehicle. I know self-driving cars, electric or otherwise, aren't in common public use, but that is how the statement presents that scenario.
It seems to me SpaceX's way of working (break things, then fix them) is fundamentally incompatible with the FAA's.
The FAA either needs to become more lenient (since there wasn't a chance of anyone getting hurt this time) or SpaceX will be unable to fulfill its commitments to put American astronauts on the Moon by 2025.
The first test flight was an unmitigated disaster so it warranted a thorough investigation. But the latest launch was well clear of any risk to human life and property and the FTS did its job. I believe the FAA should wrap up this investigation within 8 weeks and grant SpaceX a new launch license within that time.
I wouldnt have said flight 1 was an unmitigated disaster, it wasnt the best outcome for sure :) but the worst case scenario is the fully fuelled Starship & booster exploding on the pad, destroying the launch tower and taking the tank farm with it. Closely followed by they get off the pad, but the rocket pitches off towards South Padre island with a flakey FTS.
Anything better than those two setups is considered a win,and flight 1 didnt hit those issues, and flight 2 certainly didnt, even if the overall outcomes dont look like it to the outside SpaceX observer, or El Reg, there are successes there that feed into the next iteration of design and launch.
Musk has previously complained the FAA arent particularly well setup for this kind of thing, and no doubt with Thanksgiving and Christmas holidays ahead, theyre a federal agency after all, the delays in paperwork will no doubt delay the next launch attempt to the new year at the earliest, so I agree 8 weeks at the earliest is the realistic turnaround.
whilst SpaceX already have another booster just waiting for engine installation, and 4 or 5 completed Ships to pick from. so the fact Musk is suggesting theyll be ready again within 4 weeks, rather than suggesting he is being his overly optimistic self, is actually hinting at the fixes dont require major reworking of their design, which is a good sign.
That was in commercial service before the booster could land for re-use. I can see the same happening with Starship/Superheavy, where it would be commercial viable to use it as a launch vehicle even if the booster is not recovered.
As with Falcon, booster recovery (starting with simply soft-landing in the ocean) will be attempted every launch, with the data that's collected being used to refine the engineering until landings are as "boring" as for Falcon.
I'm surprised they aren't building a "catch only" tower half a mile or so away from the launch site... I know there is seriously limited space available since it's a nature reserve...
I'd have thought that the oil rigs would have made a good starting point for this kind of platform
Flight two of the Starship Super Heavy improved on its predecessor, but only if you imagine that your self-driving electric car did not crash into a tree and catch fire. Instead, it made it a bit farther down the road, and then caught fire. Who knows, maybe it'll get all the way to the shops before catching fire on your third go?
Everyone said I was daft to build a castle on a swamp, but I built in all the same, just to show them. It sank into the swamp. So I built a second one. That sank into the swamp. So I built a third. That burned down, fell over, then sank into the swamp. But the fourth one stayed up.
In the 50's the Atlas rocket had 5 officially known failures before it reached orbit with the Comsat satellite. And this is not counting all the non-disclosed failures during its secret development as the ICBM it originally was intended to be. And I bring up the Atlas first because it was the first major rocket development in the US, and second because like starship, it had a pressurized steel shell. Starship is of course orders of magnitude more complicated, but so is our capacity to build rockets.
All of these flights are incredible successes that show a VERY RAPID development process, with efficient real-life testing, and not overly expensive lab testing - the kind the existing pork-barrel funded space industry loves.
Many of the upcoming flights will continue to fail before Starship and Boster return safely and precisely to earth. And I am sure everybody at the FAA understands that perfectly. The ones that don't want to understand are those colluded with the legacy space industry, that have every interest to retain their unproductive jobs, and delay Starship reaching maturity.
Interesting to know how Bill Nelson has changed sides, now that he realizes the fight with China is on :)
"All of these flights are incredible successes that show a VERY RAPID development process, with efficient real-life testing, and not overly expensive lab testing"?
The problem with the empirical development model is that it generally makes the assumption that because something hasn't broken in a small number of tests, it's not going to in future. In reality there will be a mix of simulations, lab tests alongside the "light the blue touch paper" science, but it's still worrying that they've had these failures.
Are you asserting that the stuff that's broken is the only stuff that could break? I don't believe that, and it begs the question what else on Starship has not been adequately modelled, but because by chance it's failure mode hasn't yet been initiated will assumed to be safe?
Rockets are complex, so are large jet airliners and their engines - outside the aircraft industry, most people have a very poor understanding of just how complex a modern airliner is, and even if the operating environment is more forgiving, the complexity is easily on a par. Yet we manage to design, build and get to service airlines and brand new engines without this half-baked "whoops, another RUD, never mind!".
It's entirely possible that whatever failed on the launch test was modeled or simulated and found to be ok. How else do you improve your modelling and simulations if not from real world testing of those predictions?
As soon as they get a starship / booster back in one piece I'm sure it will be examined in great detail to find parts that "almost broke". Then it will probably be launched again to see what breaks next time. Of course simulations and modelling are important but there's no substitute for doing it for real...
Also, ref planes: Boeing would like a word...
"It's entirely possible that whatever failed on the launch test was modeled or simulated and found to be ok."
Yup. Once you start designing on the ragged edge of oblivion, testing is the only way to find what breaks vs what gets payload to orbit
Sure, you can design so it's guaranteed NOT to break but it might then be too heavy to do the job
"yet we manage to design, build and get to service airlines and brand new engines without this half-baked "whoops, another RUD, never mind!".
Part of that is each new aircraft isn't trying to completely re-design every aspect every time in the same way that Elon is trying to do with Starship. The next Airbus or Boeing model will build off the previous model it replaces. As analysis tools improve, each new aircraft will have gone through more virtual testing that helps identify problems before any hardware is sent off to the machine shop to be made. Yes, things can be over-analyzed, but not spending the time required can lead to yet another $200mn or more of parts and labor being spread out on the floor of the Gulf of Mexico and the need for investors to step up and contribute more funding.
The last aerospace company I worked at had management suffering from the Anti-NASA disease. They fervently believed that if NASA did something a certain way, they wanted it done differently. It didn't matter what it was. I know that some of NASA's policies are overly cautious or based on problems that have been overcome by advances in technology. It's important to understand exactly why they mandate something be done a particular way since many times it comes from mishaps and finding ways to never have those happen again. The beauty is that many times asking will get an answer. I have a pretty good collection of Standards manuals from aerospace companies and they document the same sorts of ways of doing things that NASA has done. I tried to get the company I was with to make a start on those sorts of policies and the engineering manager was all for it, but the suits wanted nothing to do with it.
I can see the negatives (the rocket blew up!), however, we can also see that SpaceX hit its targets for this launch and missed its stretch goals. I'll just point everyone to this article over on Ars which is written by someone who has followed SpaceX since the outset and knows their approach to development:
"Sorry, doubters: Starship actually had a remarkably successful flight"
https://arstechnica.com/space/2023/11/heres-why-this-weekends-starship-launch-was-actually-a-huge-success/
Eric wrote the literal book about the history of SpaceX so actually knows what he's talking about.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
