95 comments

[ 2.1 ms ] story [ 462 ms ] thread
I can't imagine the FAA will be thrilled with this...
(comment deleted)
Is the relationship between the FAA and space companies really as hostile as people allude to?
I can't speak on the relationship between FAA & Spacex, but I can't imagine it's great considering the public statements that have been made (ie. Twitter).

But, FAA takes safety incredibly seriously. Range safety is top priority, and a malfunction of flight termination gear would be a showstopper. If there is indeed a serious design problem with their FTS it would ground the vehicle for a lot longer than the 1-2 months they're claiming for next test flight.

> it would ground the vehicle for a lot longer

This is what Elon is quoted to be saying. [1] "The longest item on that [requirements to next launch] is probably requalification of the flight termination system"

> the 1-2 months they're claiming for next test flight

Do you have a source for that claim?

1: https://tlpnetwork.com/news/2023/04/starship-ift-postlaunch-...

Elon sated 2 months the day of the launch and again in the interview discussed in this thread.

https://www.youtube.com/watch?v=mmIqSPux3FY

Thank you for your response! (And you have such an excellent user name!)

I don't want to sound like a stickler but I'm not hearing him claim 1-2 months for the next test flight. That is the specific thing I'm asking about. I even downloaded the subtitles and was reading through it.

Are you maybe confused about this part? "I think from a technique from a rocket standpoint and Pad standpoint we are probably ready to launch in uh six to eight weeks"

Or maybe this one "and hopefully be ready to fly again in a couple months uh once again". That's not a claim that in 1-2 months they will fly again. That is a hope that in a few months they will be able. A "couple months" can be very well more than 2, and a hope is not a promise.

Thanks!

Those were the parts I was talking about and I agree it is far from a firm claim or commitment. Even "plans" are subject to uncertainty. If I recall correctly, There was also a bit where he talked about expectations/aspirations for 4-5 launches this year.

PS, where does krisoft come from? It sounds familiar?
Ah. Thank you for asking. It is a very mundane thing. Literally my name (Krisztian) + soft for software. :) I choose it when I was very young as a "brand" for the programs I was writing.
(comment deleted)
from the article "In one of many spins on the day's failures.."

I stopped reading there, the author has a clear bias against SpaceX. Expectations and what "success" meant for the test were set well in advanced from multiple sources.

Defining expected failure states is good, because it allows you to distinguish expected failures from unexpected ones. What it doesn’t allow you to do is re-cast failures as successes, which is the spin that the author identifies.
Starship not separating, launch stand disintegrating, 5-6 raptors not working, HPUs exploding, all of those were failures. That doesn't mean the test as a whole was a failure, but you cannot say that everything worked fine.
it was from the very beginning meant to be a destructive test. the fact that it didn't blow up on the pad marks it as a success!

Every millisecond of lift it gave was telemetry that will refine the end product.

This. Was. Never. Meant. To. Succeed. As. An. End. Product.

the launch, short flight and destruction was a wonderful success. Mistakes were definitely made... but thats the fastest way to learn! (As I often tell my kids)

There's a wide spectrum of possible outcomes for a destructive test.

Blowing up on the launchpad during fueling would've been a destructive test. So would splashing down in the ocean off Hawaii. That doesn't make them the same; one went well, the other did not.

i am willing to admit being biased. On the other hand, every requirement of the test was met. The failures lie in the expectations of the viewers IMO.
Honestly I think your bias is showing here. There were many failures. The self destruct should have worked, and they did not expect to basically blow up the launch pad.

I get that no one expected the rocket to work perfectly, and so the expectation is that it was going to go up and explode. That's reasonable, and I've watched other SpaceX flights with similar expectations (and occasional explosions) while considering those a success.

But there were multiple failures here, even inside of those parameters. Their safety systems, which were supposed to work, failed. They also failed to safely launch the rocket, in the sense that they scattered debris from the launch pad over hundreds of acres. Both of those can reasonably be considered massive issues that should not have happened.

They could have implemented an early warning system for wildlife.
It's 2023, not 1953! Launching giant rockets with no plan for what happens 5 seconds after launch is not "meeting expectations".
technically, i think they wanted 10 seconds of telemetry data. Everything after that was bonus.
There's no way in heckfire that the FAA green-lighted the launch of an 11 million pound rocket to get 10 seconds of flight data.

This article summarizes some of the flight objectives:

https://spacenews.com/faa-issues-license-for-first-starship-...

There was a planned stage separation, 9+ minute total engine burn, and peak altitude of 235km.

So Musk was arguing that the reason it didn't explode as it should have is that he just built it too darn strong? I'd have thought they would have tested the bomb on some big hunk of steel just like the actual tank to make sure that it worked right. They test all the other parts of a rocket, surely they do the termination system too.
His point had nothing to do with the termination system.

The rocket tumbled for four complete revolutions. Most rockets will be destroyed by inertial and aerodynamic forces well before the first revolution (they are designed to withstand forces along the major axis, not significant side forces).

It was helpful from an aerodynamic force POV that the rocket was well over 100,000 feet before it started to tumble. Still, the inertial forces when the full stack started rotating must have been huge.

That's not exactly a direct response though to the question unless you're suggestion is that SpaceX's destruct mechanism consisted of expecting aerodynamic forces to destroy the rocket during a tumble.
I interpreted Musk's comment to reflect the fact that rockets generally will be destroyed if the tanks lose pressurization and/or the rocket tumbles. The (load bearing) tank walls usually cannot support major side loads. The Starship stack needs to be a lot stronger than "usual" rockets to withstand the rotational forces and aerodynamic forces involved in landing them.

Ref: https://headedforspace.com/why-and-how-rocket-fuel-tanks-are...

OTOH, most rockets are transported horizontally. (insert picture of Falcon9 going down a road on a flatbed here). This means that they are designed to withstand 1g side loads. SuperHeavy is always roughly vertical. Starship of course has to withstand insane loading horizontally during it's crazy descent and flip stage, but SuperHeavy stays vertical.
The tanks are held in supporting cradles when horizontal so that their weight is distributed properly. Sometimes the tanks must be pressurized when transported horizontally or they would be damaged.
The charges worked fine. They blew holes in the tanks as planned. The issue is that this didn't cause immediate structural failure as intended. But structural failure of the rocket while flying through upper atmosphere at supersonic speed is not something that you can test on the ground.
^ THIS ^

This is the whole of Rocket Science. You can't test for some conditions, you can only test IN them by actually launching a rocket.

I think of it like diving into a pool. There is no way to practice diving into a pool before actually diving in. You learn to dive by diving.

I read that the destruction mechanism is explosives that rips open the fuel tank, internal to the rocket, but apparently then requires atmospheric pressure to collapse the rocket shell, and at the altitude it was initially at the pressure was insufficient to do that.

If this is really the mechanism then I suppose tough to test (at least expensive to have large enough low-pressure test chamber to blow stuff up in), but it does seem odd they wouldn't just have explosive charges to destroy the rocket shell as well.

The "rocket shell" is the "fuel tank". They blew a big hole in the side of thing, which usually is sufficient to destroy any rocket in seconds. Turns out this one it took a while for the damage to cause complete failure.
Interesting! Do we know how soon after it started tumbling they fired the self-destruct ?
Here's a video of the flight, where you can see the charges activate and venting start at about 15:12, the linked time: https://youtu.be/NCYSVmSPM7E?t=663

You can decide how long that is before and after, though I'll note that it may be complicated to determine what's not nominal: Starship has an unconventional staging maneuver that's basically a spin.

I'll also note that "they" didn't send anything; the FTS is autonomous, and will trigger on violation of its conditions, which is probably (1) departure of the instantaneous impact point (where the rocket will hit if all thrust stops now) from a safe ground corridor or (2) indications of disintegration of the vehicle. "The vehicle is doing something weird" is not a condition for flight termination.

Good thing one of the day's many failures didn't take the rocket on a course over Port Isabel or Matamoros. Someone really needs to explain why the flight was permitted to continue after the hydraulics were lost.
It wasn't. There was a 40 second delay between FTS signal and explosion.
Thrust vector control was lost 85 seconds in. Then they continued onwards for several minutes more, and even then range safety couldn't terminate the flight for 40 seconds. That's not a good sign, also because anyone capable will either have been fired or left on their own accord a long time ago.
It was within the flight corridor during that time.
Let's skirt as closely to the edge of the safety envelope as we can. That'll work!
TVC was lost at 85 seconds, FTS was initiated before 90 seconds.
IIRC, RocketLab launches from Wallops were delayed by about a year due to delays in FTS certification. Doesn't bode well for seeing Starship launch again any time soon.
RocketLab's largest hurdle was certifying their generalized AFTS software for any rocket and any spaceport, this won't be an issue for SpaceX.
It’s really frustrating to me how much public info is getting “dumped” in Twitter spaces, which are for the most part impossible to get back. It’s hiding useful information that would let people make more informed decisions, even if just on something like their own opinion of how something happened… it’s as if you put press releases up on a time limited Snapchat… it’s dumb. And I can immediately think why Elon is doing it, it’s harder for people to go back and dredge up what he says and report on it if the primary source is gone. It’s his choice (unless legally obligated by other circumstances such as reporting requirements) but it’s deeply frustrating to try and follow Space industry news when you have one of the most newsworthy recent events disappear into a smokescreen of hate posts online and primary sources get turned into disappearing mirages in twitter spaces.
This is an an understandable frustration, but I don't think it is a underpinned by a reasonable expectation. In the post internet world, people have come to expect information to be available immediately with minimal effort; to be able to find the airspeed velocity of an unladen swallow with a click of a button[1].

I think the most challenging part is the expectation of immediacy and that someone will pre-package the information. If you are fine waiting a year or two, there will likely be books on the topic with extensive information, maybe some FOIAed original reports.

[1] Im guilty of this too. I googled the quote to get it right and was frustrated that the complete text wasn't the first response.

Oh I don’t mind it not being pre packaged… if anything the death of blogs and the rise of “a video on YouTube will be the only form ever used to present information” has forced me to get used to watching or listening to entire chunks of stuff to get the information out of a primary source…

My biggest complaint was that SpaceX are posting primary source material into a twitter space that other people have to archive. For instance the Reddit post earlier linked to a YouTube video posted by the account. “Elon Musk History” which feels pretty ephemeral given it has only been online since a bit over a week ago. It bugs me that official primary source information of relevance to the public, to the space industry, and to space nerds in general, has to be captured and archived away by passionate ransoms on the internet… people should not need to be doing OSINT style scrape and archive tactics on SpaceX public statements by the CEO regarding the first integrated test flight a rocket that will be part of the Artemis moon landing run by NASA … the tinfoil hat urge to label this as Elon forcing people to engage with Twitter features to make them more relevant … but it’s more likely he wanted to be able to do this briefing from his phone regardless of where he was at the time and decided to just do it in twitter “because that’s good enough” (the unsaid subconscious decision making being that obviously it’s good enough because he bought the whole company)

If this was better managed, it would be archived and published by Elon or SpaceX to some other channel, but it’s not, it’s left on the floor of the internet as trash headed for the delete pile unless someone jumps in to make a copy for historical record.

Again, I agree with the desire for more documentation, but what I disagree with is the idea of doing it more thoroughly or not at all. That simply results in less information. You can look at what other companies do as a comparison and you don't get nearly as much information. SpaceX seems to me to be far more disclosing. For example, you could find raw data from on board Starship sensors informally released to the public the day of the launch. I haven't seen anything like that from the SLS. I would rather have that information out there for willing individuals to sift through than not at all.
Ahh, here’s where I’ve definitely not been clear enough then.

I’m not trying to imply they should do nothing if they won’t try and pass the bar I’m setting… and managing this information officially themselves…

I’m lamenting the fact they aren’t trying to live up to the bar that used to be considered normal for official communication from a company, that the official communication will be available once they are done speaking… and hilariously enough I just realised what this is, we went from the modern YouTube era where companies wanted to be in more control of the content they posted and would this run the PR event then published the “official take” to a corporate controlled channel and hold copyright and all the power it gives them… back to the pre-YouTube era of the press conference where reporters brought their own voice recorders and shotgun mics and all hustled for optimal position, recorded their own copy of what the people speaking said, and then all either wrote it up themselves or passed it back to the news desk to have a writer do the article writing… and it feels very stupid to step back like this when all the bullshit can be replaced by a shell script the company runs.

> It’s really frustrating to me how much public info is getting “dumped” in Twitter spaces

But the alternative is not that you get a nicely packaged information dump in some public channel.

The alternative is that the company says "There are multiple anomalies we are investigating alongside our governmental partners." And then x months later they say "The regulatory agencies gave the all clear for a new test launch." You as the public is not expected, nor usually privy to these information. Maybe you can FOIA it out of the FAA later, if you are lucky.

Anyone know where I can find a recording of the "Twitter audio chat on Saturday"? As-is it's not clear the article's body supports the claim made in the headline.

> Musk claimed it was because "the vehicle’s structural margins appear to be better than we expected."

What do high "structural margins" have to do with inability of the self-destruct system to respond to commands?

This is pure speculation (since again, I haven't been able to find the primary source), but it sounds to me like the self-destruct system did respond immediately, but the command to self-destruct wasn't issued because the vehicle took 40 seconds longer than expected to reach a state where a self-destruct was necessary. If that's true then this isn't a critical safety issue like the headline is implying.

Thanks! Looks like the portion referenced in the article headline is at 3:16: https://youtu.be/mmIqSPux3FY?t=196
From Elon's description there it sounds like he believes the self-destruct system did immediately activate, but there wasn't enough detonation cord for the resulting explosion to immediately destroy the spacecraft at that altitude. That's where the "40 seconds" number comes from.

He also suggests that re-qualifying the abort (self-destruct) system is the item that will take the most amount of time before they'll be ready for the next launch.

> Musk: "The outcome was roughly in what I expected, and maybe slightly exceeding my expectations, but roughly what I expected, which is that we would get clear of the pad."

I'm still surprised by such a low bar.

It's in line with their previous style of iteration and small incremental changes on Starship, where they often have the next test article already lined up. I guess we're just not used to seeing it with a complete, large rocket stack.
Exactly the sort of questions I had before I wrote my post https://news.ycombinator.com/item?id=35772716

I wanted to get the primary source but can’t unless I get lucky and someone saved it… and I can find the saved copy somehow… and I even know it happened to go look for it (which is a major issue when these things happen late at night in my Timezone and I may be asleep from before it’s even posted as an alert let alone when it’s done and disappears)

If anyone has a copy, yeah I’d love the link as well.

(comment deleted)
Trying to make a case for SpaceX, my best excuse is that the tumble somehow prevented radio reception of the remote destruct command.

Based on having watched Scott Manley's video though, I am surprised an onboard auto-destruct did not kick off immediately when it went through the first full rotation.

I mean, that would actually be a terrible excuse though, no? The spinning should be expected enough that you go with a radio system/setup that can handle it.
The summary from the reddit thread makes it sound like the "structural margins" statement was unrelated and about the somersaulting, not about the delayed self-destruct. The article may have taken it out of context.
The FAA probably used NASA modelling software to certify the SpaceX Flight Termination System. They probably used the same software to certify the FTS on SLS, Vulcan and other large rockets that haven't had an implicit or explicit FTS test. That might call into question the safety of those rockets...
Common Sense Sceptic did a whole analysis of the launch pointing to this too:

   https://www.youtube.com/watch?v=ErDuVomNd9M&t=2s 
Note: CSS is very anti Elon Musk and SpaceX. However, they are usually very thorough and include a lot factual information in their material. So it can be interesting to see, just to hear an opposite (less positive) view of Elon Musk's statements and endeavors.
Von Braun said that they learned more from a launch failure than they did from 10 launch successes.

This is an example of something that they never would have learned from a successful launch.

Von Braun knew that you can't have a giant rocket deployed like a blowtorch onto concrete structures. He knew that without ever trying it.
To be fair, Von Braun and parts of his team had a huge amount of experience from many years of testing experimental smaller rockets at Peenemünde (the museum is great) and other places.

You can find footage of V-2 test launches where the entire test article comes falling/crashing down into the countryside.

I would also assume that approvals for risky launches or going hard on gear was a bit easier to get during a war time weapons research operation, so they presumably had their cowboy years.

> You can find footage of V-2 test launches where the entire test article comes falling/crashing down into the countryside.

Absolutely. Here are some of the test footage:

https://www.youtube.com/watch?v=7YFU4KaJSSc

(not really fancying the music, but couldn't find one without.)

The concrete launchpad that SpaceX built has already withstood forces greater than any rocket in Von Brauns lifetime and every other rocket that has ever flown.

The successful half-thrust Starship test applied more force than the Saturn V or SLS in development do at full thrust.

People quick to criticize the spaceX pad conveniently ignore this fact, as well as the fact that SpaceX knew it was not reusable and was already planning to upgrade it.

If Von Braun actually thought what you claim, he has already been proven wrong.

> The concrete launchpad that SpaceX built has already withstood forces greater than any rocket in Von Brauns lifetime and every other rocket that has ever flown.

"withstood"

I'm not sure what you are getting at. The half-thrust test had very little damage and was more force than every other rocket on the planet.
Von Braun knew that if you didn't ask questions, your government might just give you Jewish forced labor for your rocket plant, which you'd gladly accept.

Von Braun also knew that if you built missiles for the power-hungry US military, they'd conveniently forget the fact that you were a Nazi.

That's much less convincing for the FTS in particular, which needs to be reliable even (or especially) in early testing. Since you expect everything else to go wrong, you need to be able to stop it from going wrong on top of people.
News also came out that the rocket was leaning shortly after launch. Not a great situation with a faulty flight termination system.
Clickbait and lying.

It did immediately respond - by detonating the self-destruct explosives. It's just that that didn't immediately destroy the whole rocket.

(comment deleted)
A couple things.

The first is that Starship has a AFTS (Autonomous Flight Termination System), which means that it's commanding itself to destruct if it predicts that it will be going out of bounds, rather than the ground sending the command to it. I don't think there is a way to command the rocket remotely after launch.

The other thing is that it did break up when it was around 30 KM above ground.

I'm baffled that they didn't use a linear charge for the FTS. Was it that much cheaper to just put the two little hole-poking charges?
I suspect it isn't cost but the desire to have as little explosive as possible for worker safety.
sounds like they did. Elon mentioned a "longer cord" not adding a cord.
It "responded" immediately to the "command" (an onboard automatic trigger) and the explosive charges ruptured the tanks instantly. However, the explosive charges which ruptured the tanks didn't cause the vehicle to disintegrate immediately, as it was stronger than anticipated and at a high altitude with lower wind force. (No one has flown a rocket of this construction and size before.) The flight was, however, pretty well terminated at that point (the FT in FTS).

They will be adding more explosives for the future to ensure structural failure.

That isn't what Elon stated in the interview (but you may still be correct).

[Edit: You are correct]

"The longest lead we dont have is probably requalification of the FTS. Um, Because we did initiate the FTS, but it was not enough to.. it took way too long to rupture the tanks. We need.. basically we need more detonation cord to unzip the tanks at altitude and ensure the rocket explodes immediately."

It sounds like the FTS worked as intended, but the holes created in the fuel tank were undersized, allowing the tank to maintain enough pressure to remain structurally sound. You can see 4 plumes of gas venting from the 4 locations of the FTS boxes while it tumbles. I would expect them to switch to long shaped charges, possibly radial in orientation.

It sounds to me that the FTS was triggered automatically by the onboard computer, but that Musk concedes that it possibly should have triggered (or been triggered) sooner:

Musk: Time for AFTS to kick in "was pretty long," about "40 seconds-ish."

On the other hand, rockets have blown up on the pad due to over-aggressive AFTS, so I would imagine the bounds were pretty loose for an initial test flight.

I wonder if it also just replied with a U+1F4A9 PILE OF POO.
"Looking into it".
I can't find the documents/video but you can see that the self destruct explosives do indeed go off.

All that they do in starship is punch holes into the fuel tanks. Then, as soon as air mixes with the fuel in the correct ratio, boom.

However, there was too much pressure in the fuel system and it was jetting out so strongly that it appeared O2 wasn't mixing with the fuel.

After about 10 seconds you then see it explode.

Aegis would have taken care of it easily with an SM-3. I wonder if some Arleigh Burkes or other Aegis-enabled ships are on standby every time there's a launch. Or perhaps the Air Force helps with a Patriot battery.