Comment by samwillis | Hacker News Reader

Comment by samwillis | Hacker News Reader

Parent Full thread

The working theory from people on Twitter who generally "know their stuff" is that the FTS blew holes in the tanks, but where normally this would result if the tanks immediately structurally failing they didn't. This was probably a combination of the stainless steel contraction, pressurised tanks, and under specified FTS. It took a good 40 seconds from FTS to structural failure.

There have also been questions as to why it appears the FTS was only point charges rather than a cord down the side of the tanks.

There is a good stabilised video of this portion of the flight here (https://twitter.com/Erdayastronaut/status/165250639080235827...), you can see the plumes of propellant from the sides of the tanks where the FTS has made holes.

Elon also just did a Twitter Spaces where he answers a load of questions, summary here: https://twitter.com/thesheetztweetz/status/16524519714109358...

Key takeaways:

> From a "pad standpoint, we are probably ready to launch in 6 to 8 weeks.'

> "The longest item on that is probably requalification of the flight termination system ... it took way too long to rupture the tanks."

> Probably an 80% probability of reaching orbit with Starship this year, and "I think close to 100% change of reaching orbit within 12 months."

> "We're going to putting down a lot of steel" under the launch tower before the next Starship flight.

> The vehicle's structural margins appear to be better than we expected

> At T+27 seconds, SpaceX lost communications due to "some kind of energy event." And "some kind of explosion happened to knock out the heat shields of engines 17, 18, 19, or 20."

> Rocket kept going through T+62 seconds" with the engines continuing to run. Lost thrust vector control at T+85 seconds

They will never launch within 8 weeks. That time frame is insane and not enough to repair the damage at the launch site.

Also I don't expect a FAA to give an ok until SpaceX has cleaned up the surrounding area.

Doesn't the specific "pad standpoint," qualifier at the beginning of the statement mean to make clear that the pad will be ready in 6-8 weeks, not the rocket?

Otherwise, why prefix with it?

Yes, this is "Elon time", so x2. Four months?

If the piles under the launch mount aren't damaged (which would surprise me!), it's "just" filling in the hole, pour some concrete and cover with the new water cooled steel plate. Then replace any other stage zero kit that got damaged (the tall tanks particularly)

I think he's probably right the longest thing to solve though is flight qualification with the FAA.

x2? x10 has historically been more accurate.

I also love how Elonauts have decided that FAA qualification is a mere trivial formality as opposed to a necessary check on a founder deciding to go ahead with a premature launch because he wants to talk about it on 4/20 for example.

The fact that the FAA check will be the longest is no defense. It’s a reflection of the fact that the FAA gave him an easy way out and all the critics of the FAA’s decision (including flagged stories on HN) turned out to be absolutely right.

He predicted it would launch 4/20 and it launched 4/20. So current factor is 1x.

Sigh. Yes, perhaps you are accurate that it’s not always 2x but your 10x things is pretty childish.

So what if it’s not 2x but more? Why are people like you so resentful of someone building something cool you literally can’t wait to point out any flaws. It’s sad tbh.

Because unrealistic representation of timelines can sound like a non-technical manager eager to claim credit for arduous research and mechanical work they don't seem to understand enough to give accurate forecasts for, just as much it might sound like an excited geek who can't wait for a cool thing.

Maybe it sounds like aggressive and optimistic goals. Maybe it sounds like someone who doesn't understand that nine women can't give birth to a child in one month, and cracks the whip like they could if they tried hard enough.

Have you seen the footage of the debris super heavy launched into the ocean/surrounding wilderness? I'm fine with "someone building something cool", I'm sick and tired of those same individuals pushing the negative externalities on the rest of society through lack of investment/planning.

If I was your neighbor and build a potato gun, but the potatoes kept ending up in your yard, would you be happy for me or ticked off?

Some dust dropped on the wilderness and ocean is a nonissue, it's a weak sneeze compared to what the mining and construction industries do every day without fanfare. The issue is the dust dropped on the people in that area.

What else can we whine about? Did the rocket maybe hit an insect on its way up?

Hardly any one lives there, and the people who get compensated. Wildlife is not going to have major, and I say major issues from this.

You sound like the environmentalists you don’t want to build solar parks in the Sahara because it gets in the way of a few goats.

Take a step back. Perspective.

https://freakonomics.com/podcast/failure-is-your-friend-2/

Space shuttle challenger disaster discussed at 17:35.

Nobody is resenting anything cool, they are pointing out that when it comes to packing 4kt worth of explosives into a vehicle and flinging into the air you better take your time to get it right.

This is a private company which is not directly accountable to the government as NASA would be. There won't be a congressional committee holding them to account or removing the people in charge if they fuck up. Sure they might get a good scolding from congress if they make mistakes, but they'll get to pay their way through the courts to shirk accountability.

The fact that the FTS failed on top of all the other problems is cause enough for alarm and indicative that they are not being thorough enough. It was sloppy work, and when Elon opens his mouth about unrealistic timelines he is openly stating that he will continue to apply pressure on the team to rush things and cut corners for results. There is nothing cool about running an aerospace program like that.

Falcon 9 Block 5 is literally the most reliable rocket of humanity as of today (the longest streak of successful launches ever), and on the top of it, partially reusable and fairly cheap to launch. It took several explosions (that you seem to consider "sloppy") to get there, but the result exceeds anything that competition came up with.

There is an adage that people who love sausages shouldn't visit the meat plants where these are made. This is harder to do with the development process that SpaceX uses, given that it is all over the news and you can't easily escape the sight. But please do not confuse the development process with the end product.

As of now, Starship is in the not-even-beta phase. Its current problems have no relation to how the mature rocket will perform.

Note that, unlike other vendors, SpaceX didn't dare to launch human crew with their Falcon 9 until they were very sure that the rocket performs well and has a very good launch record. Compare this attitude to many other human-rated vehicles that actually killed astronauts. Nothing sloppy or rushed there.

Your point is a copy paste of every negative failed spacex launch reaction.

Yet somehow they’re still the most successful rocket creating entity ever, and are close to creating a launch system that will reduce launch costs by 100x.

So what exactly is not working here, and why are YOU so upset about it?

Cybertruck, Robotaxi, FSD ?

We are still waiting for all three many years after Musk promised they were coming soon.

Roadster, Dojo (this one I remember actually believing he sounded so sure it was a done deal years ago), Semi. 10x is generous given his history , can't believe how many people still find him credible

Yet he’s still beating blue origin, rocket labs, every major car company trying to create electric cars, and a charging network.

Promising 10 awesome things and delivering 2 still gives you more awesome things than you started with.

At best you are a grumpy old man who isn’t happy with anything, at worst a resentful anti success advocate.

What have you created in the last 10 years?

I like seeing people be successful. With Elon there is always too much celebrating of announcements vs celebrating successes. He's done well selling EVs and Falcon 9 is great. Mostly he lies though so that's what I assume with any forward looking statements

It’s not lying. It’s being wrong in the timeline for deeply complex technical innovations. This is like me calling my engineers liars every time a task takes longer than expected (which is nearly every time). This whole “predict the exact day humanity will solve its most ambitious technical challenges like self driving cars or you’re a liar” attitude is incredibly toxic

Being wrong once on a date 5 years away isn't necessarily lying. Being many years wrong on frequent "by the end of the year" predictions in a way that happens to help you sell products is lying.

How many times has a simple coding task taken longer that you estimated? I guess you’re a liar too.

X10 would be a generous assessment of when FSD will fully SD.

The rule of thumb is that it takes 4 weeks for a concrete driveway to fully cure.

I admit I've never built a launchpad for the world's biggest rocket, but I'd wave my hands to say it might take a bit longer for that concrete, and it won't get poured immediately.

How do you water cool a steel plate under a rocket without the coolant instantly vaporizing and blowing up the plate?

> instantly vaporizing and blowing up the plate

It sounds like you are thinking about the water coolant being in or behind the steel plate.

But if you pump water on top of the steel plate there it can instantly vaporize without blowing up anything. If you have sufficient flow then that can keep the steel plate from melting.

In the Spaces thing, he described it as being like an upside-down shower head, and to expect lots of steam billowing out during launch instead of dust, so I don't think it's a closed system.

NASA Kennedy launchpad deluged test https://youtu.be/LNkmwrTjKuo

That flow is even bigger than it looks, seriously impressive.

> The IOP/SS peak flow rate is 1.1 million gallons per minute, high enough to empty roughly two Olympic-size swimming pools in one minute.

https://www.nasa.gov/content/launch-pad-39b

4.16 mega litres a minute.

Must be some fair old pumps on that setup!

They’ll need pumps bigger than the rocket’s fuel pumps.

Nah, just some water towers. You can fill the water towers slowly using small pumps, then drain them quick using gravity.

> They will never launch within 8 weeks.

And the quotes don’t say that they will. In fact they imply that they won’t.

Also, it's not enough to simply repair the damage.

They need do proper failure analysis and engineer a solution so this damage won't happen next time, 6-8 weeks seems way too short. Elon is suggesting just slapping down a shitload of steal, but I really doubt that will meet standards.

I don't understand why Elon insists on re-inventing the wheel when we know how to deal with rockets this size. The Saturn V was similar in size and power, the launch pad at Cape Canaveral is still standing, just build a copy.

Super Heavy has more than double the thrust of Saturn V, how is that similar?

He's got many reasons. Some of them are probably not good reasons. If the project fails, it's a good bet some of the reasons are in conflict. Also, some reasons drive bad decision making as the project proceeds. This is just a big, vivid example of how many projects die:

1. Not giving up project goals or specs when they start to look dubious

2. Sticking with cheap (in time, money, or both) solutions when more-reliable solutions might be called for

3. Letting sunk costs drive bad decisions about the above and other project related decisions

In this case there are unpleasant downstream effects of making safer decisions: Starship might not be the Moon lander Artemis 3 should rely on. All changes made in Boca Chica will have to be retrofitted to Kennedy Space Center.

Lastly, Musk has made pronouncements about Starlink's viability resting on Starship. That could be hyperbole, but Starlink needs 30-100X more subscribers to become a first tier telecom provider, and it can't support that many without the planned constellation of tens of thousands of full-size second-generation satellites.

I mean challenging the “well known” engineering orthodoxies like rockets landing being almost impossible and not worth doing seem to have worked out ok for him?

I love it when people take one small issue and point out how stupid he’s being, yet ignore how by challenging every point of engineering on rockets he’s almost at the point of making a mars transportation system possible.

Rockets landing was never considered impossible, NASA even built and tested several such vehicles long ago. What was determined is that is is not economically viable which even SpaceX has started to realize on some launches where they ditch the landing in order to be able to carry more payload.

The launch pad at Cape Canaveral may not be exactly what Starship needs but it would be a good start. Like flame trenches and placing the fuel storage tanks far enough away from the launch pad.

> What was determined is is not economically viable which even SpaceX has started to realize on some launches where they ditch the landing in order to be able to carry more payload.

SpaceX hasn't "started to realize" that, they always knew that some F9 launches would be fully expendable and they've been occasionally expending F9 boosters from the start. Not because there's something wrong with the economics of booster reuse generally, but because expending a booster is sometimes (not usually) worth it to get more mass into orbit (roughly double to GTO with Falcon Heavy if they expend the core stage.)

That extra performance was always known to be required for some payloads, but isn't required for most payloads.

What was stopping any other launch system from sending fuel tanks to LEO and beyond?

Probably not a whole lot technically. By now we could have had several fuel tanks in Mars orbit and a few on the ground, rather than all those orbiters and ground vehicles.

Trade-offs.

What would be the point? Who'd pay for it? The real money is in launching satellites into orbit around earth and you don't need orbital propellant depots for that. Serious Mars colony customers don't exist.

I know they're planning it for Starship HLS to the Moon, maybe it will finally be done then (I'm not holding my breath, Starship HLS's role in Artemis still seems farcical to me.) But until then, it's never really been an idea worth developing.

Doesn't look like my tone came across correctly.

I agree with you.

They had done static fires at half thrust, equaling the thrust Saturn V produces without issue.

Why spend hundreds of millions or in the case of Saturn V billions on a governmentally overspecced launchpad when a launch mount and a concrete pad would do just fine?

Manned space exploration is hard to "overspec." It is risky af under the best circumstances. This illustrates a conflict in Starship design goals: Big enough to colonize Mars but cheap enough to lunch Starlink satellites. Might not be reconcilable.

Rather, it is incredibly easy to overspec. Everything is important and then you get SLS like costs. Compare to Falcon 9 which took a similar path to Starship and today works beautifully for human launches.

Take the SLS crawler looking to cost $1 - $1.5B. What a waste of resources.

https://spacenews.com/nasa-audit-reveals-massive-overruns-in...

Apollo cost 2.5% of USA's entire GDP, every year, for ten years. Space exploration is insanely expensive. It is unclear if the requirements of space exploration can be met while making a rocket intended to be at least 5X cheaper per kilogram to orbit.

Trying to meet wildly divergent, even contradictory goals, is a good way to build a doomed project. Now add schedule pressure to remain the sole vendor for a lunar lander and you get bad project management decisions.

Spacex have already succeeded with Falcon 9? Starship is the next magnitude in cost reduction following the same formula.

I am not doubting that Falcon 9 is a great success. It took 6 years of attempts to reliably land the Falcon 9 first stage, at some cost to capacity. The second stage is not recoverable and it appears unlikely it ever will be.

"...following the same formula" of rapid iteration is what they are doing. It worked well for learning how to fabricate a giant rocket out of stainless steel. That formula is bumping up against the cost of iteration. They're trying an Agile-ish approach to making the biggest ever rocket when Agile might be insufficient to, for example, make car software at VW Group.

Starship is all about finding ways to whittle down margins from the status quo

They obviously took a calculated risk. If they never took any risks they wouldn't be where they are today. It's easy to criticize in hindsight, but you win some you lose some.

In Boca Chica, underground water is too high AFAIK. Any trench system would risk being flooded.

Who said they’d launch in 8 weeks? He only said the pad would be repaired in 8 weeks

>"We're going to putting down a lot of steel" under the launch tower before the next Starship flight.

I understand that Starship needs the ability to operate in places where launch infrastructure can't exist. But that's a problem for another day. Put in a flame trench + water suppression system and get the thing to orbit.

The main problem isn't Starship with its six measly engines on a lower-gravity celestial body (e.g. faster lift-off).

The main problem is the Super Heavy booster, tearing the ground apart with 33 engines at full thrust while struggling against 1 g downwards. This is a very Earth-specific phenomenon, because of our relatively high gravity. You can easily do a Single Stage to Orbit launch with just Starship on Moon, Mars, Europa etc. It is very likely that Super Heavy boosters will never leave the Earth.

A flame trench would probably be a lot more inconvenient to them. Right now they can freely move around most of the launch pad. This is probably important for their rapid iteration, which is what is going to get them to orbit as quickly as they can. If a big chunk of the launch pad is consumed by a trench, that area becomes inaccessible by their platforms, cranes, etc.

Plenty of types of cranes can span a trench.

Not really. As far as I understand, the booster is only for launching from Earth. From elsewhere (moon, Mars) Starship can launch alone.

Then maybe it would also require that much less thrust?

Isn't the water table too high at that location? If that's true then it was never a realistic possibility to build a trench.

That's not an insurmountable problem, it just costs more money to deal with. People have been building structures below the water table for hundreds of years using cofferdams and caissons. Very realistic, just a bit pricey.

I thought the issue with throwing up a bunch of dust is that the planet this rocket was starting from is inhabited.

The dust is a problem for the community in that immediate area on the order of miles, but on a planetary level it's nothing compared to the dust the mining, construction, etc industries produce. Per the ATF: "In 2015, the latest date for which statistics are available, the U.S. used slightly less than 4.5 billion pounds of explosives." Of course measures are often taken to reduce rock dust falling on people, but away from populated areas it's common to blast rock in the open air, in vastly greater quantities than the debris lofted by this rocket.

Not quite. It is speculated that the rocket engines were damaged by chunks of the pad hitting them. That is a danger to the launch itself even if the dust wouldn’t bother anyone else.

The upper stage, with 6 engines, is the part expected to land on the Moon and Mars and take off. Still, that's going to be challenging: Lunar regolith wants to clog and grind everything it gets into.

I'm going to violate one of my Prime Directives about posting on Musk threads, but there's just no way they'll repair that pad in two months.

Even if Musk was a Stalin-esque Maximum Leader, it couldn't happen. That's not even considering the other regulatory hurdles you might encounter when your planned building expels apartment-sized chunks of concrete at 83 m/s.

Of course, regulatory is . . not the primary problem, even if you're of the persuasion that things like safety and the environment don't matter. Some chunks rebounded and almost certainly hit that rocket. Even without "lotto ball concrete debris", Dougherty et al figured out in 1984 that high impulse continuous shock would reflect, resonate, and shake a rocket to bits. It's research related to the "water deluge" system; arguably, the Soviets came to the same conclusion with Energia, based on the N1 work from the decade prior. Minimum safety distance, that's to protect infrastructure that takes a long ass time to put back, and the V2 rocketmen started paying pretty close attention to that even for their suborbital shots.

I guess I'm not too sure what got learned here, aside from the fact that Raptor is a pretty great engine, still burning at all after that abuse.

Also, Saturn V's S-1C had 13 launches and 13 orbits . . fifty years ago. Why is that? Well, there's lots of reasons, and one of them is the cost of labor. Those big F-1 engines were essentially hand-fitted pieces of welding art, but, counterpoint, computing and other tech would MORE than make up for a surfeit of cheap master welders[1]. Leadership from capital is another. I'm not going to say anything negative about anybody, but let's take a look at one of the maximum leaders from the era: Thornton Wilson. Born on a Missouri dirt farm, ISU BS Aero Engineering, CIT MS in same, NAS award winner, as undergrad he worked on B-47 and B-52, and then kept on working aerospace UNTIL HE DIED. In 1969, the buck stopped with this guy. This is who capital put in charge of the big booster back in the day. How does capital do leadership now?

[1] You would think, right? Maybe these computer things aren't actually all that useful, but I don't believe that. More likely, computing has a whole bunch of secondary effects on organizational health that are really, really, really bad, which partly offset the amazing gains.

Reply on news.ycombinator.com