SpaceX Halts Rocket Launch 10 Seconds Before Planned Liftoff | Hacker News Reader

SpaceX Halts Rocket Launch 10 Seconds Before Planned Liftoff

bloomberg.com

129 pointssillypuddy64 comments

There's an old quote that seems applicable here:

"It's better to be on the ground wishing you were in the air, than to be in the air wishing you were on the ground."

Since I'm embarrassed by how many votes this is getting without any citation, I'll just point out that I probably first learned about it from something Tory Bruno (CEO of ULA, the company that launches the Atlas and Delta rockets) said, seen here:

https://twitter.com/spaceflightnow/status/776854635055284224

This morning, it was posted by Doug Wheelock:

https://twitter.com/Astro_Wheels/status/832972785152573441

It appears to be said by many people before them, in different forms - I can't find any information about who first coined the phrase.

This is also a common statement at skydiving dropzones.

There's a related truism in aviation - "There are three things you don't need - runway behind you, airspace above you and fuel on the ground."

"A good rule for rocket experimenters to follow is this: always assume that it will explode"

-- Astronautics, issue 38, Oct 1937

Wise.

This is far from the closest to launch SpaceX has called an abort. Many may not remember 3 years ago when they called the abort at T-1 after main engine ignition: https://www.youtube.com/watch?v=5v0k_0PTLnI

To be honest, though, neither case is really abnormal. There's a whole reason it's "T-minus" and "T-plus". T-zero is the point of no return (on the STS, this was when the solid boosters were fired). An abort at any point before then is still considered normal. Aborts after T-zero are anomalies.

(FWIW, even the STS has aborted take-off post-main-engine ignition: https://www.youtube.com/watch?v=8bkYP3pU76I . The main difference between the SpaceX Falcon and the STS is that the Falcon is capable of re-igniting its engines almost immediately, making a same-day re-attempt of a post-engine-ignition abort possible. The STS main engines require a manual reset, making abort post-engine-ignition more costly.)

It wasn't an abort, but the Mercury-Redstone 1 had a "T-plus" stop time [0]. It attempted to launch and the engine shut down immediately after lift off. It didn't explode, the escape rocket jettisoned, the capsule deployed its drogue, and the the rocket settled back on the pad untethered [1]. It came to be referred to as the "four-inch flight".

After considering some risky ways to deal with the fully-fueled rocket, flight control ultimately decided to wait until the LOX boiled off. Chris Kraft declared, "That is the first rule of flight control. If you don't know what to do, don't do anything."

[0] https://en.wikipedia.org/wiki/Mercury-Redstone_1 [1] https://www.youtube.com/watch?v=7O4V7JfeTSU

And if you like stories like that, "Failure is not an option" by Gene Kranz is a must-read/listen. It seems every single flight of the early American Space Program was a nail-biter.

When you're responsible for it, or even a part of it, every moment is a nail-biter. You know exactly what can go wrong and how bad it could be.

I had found it odd, knowing that SpaceX ignites engines at T-2 (and releases clamps at T-0), watching the Arianespace launch earlier in the week. They ignited the main engine at roughly T-1, then revved it up until T+7, and then fired the solid boosters and released at T+8 or so. https://youtu.be/pgj0kSXEtdQ?t=890

The STS fired the solids at T-0, thus being the "point of no return", why is it that Arianespace waits until +8 to fire solids? Do they just consider igniting the liquid main stage as "no return" ?

Also, SpaceX had a T-1 abort after main engine ignition later too. (the one you showed doesnt even have landing legs!). The later one was SES-9 in Feb 2016 - https://www.youtube.com/watch?v=LuS5mehCAMI

A boat was down-range and the cryo propellants had warmed up and started getting bubbles in it... the engines fire at T-2, run diagnostics, determine if they are running properly, THEN release clamps for lift off, but in this case, the bubbles caused lower than expected thrust, so it turned itself off.

The Arianespace count is just a difference of opinion. The US space industry use T-0 as the moment of departure from the launch pad, whereas the European program uses T-0 as the moment of main engine ignition.

Ariane 5 has aborted after T-0 but prior to igniting the boosters, video here: http://www.space.com/11266-ariane-5-launch-aborted-engine-ig...

An abort after T-00:02 was iffy for STS. If an abort was called after T-00:02, the RS-25s kept running until T-00:00 just in case the SRBs lit anyway.

Makes you wonder why anyone ever thought it was a good idea to put solid boosters on a man-rated rocket...and yet NASA is repeating that "mistake" with the SLS.

It's a way to use NASA budget to prop up military contractor ATK.

Not a problem this time because the capsule has a launch escape system

A launch escape system which uses parachutes, which will not react well to descending through a cloud of burning solid rocket fuel...

Didn't Congress order NASA to do that, this time around?

I can't quite imagine what failure modes were they thinking about. Those failure modes must have had a nonnegligible chance of lighting off _both_ SRBs, but a very small one of lighting off a single SRB.

IIRC, the SRBs generated enough thrust to break the hold-down bolts even in the event the frangible nuts don't fire. I'd guess the thinking is that, if both fire you might as well have the main engines running than not (since, as I noted before, the SSMEs couldn't be re-ignited in flight). True, if only one SRB fires, you're pretty much toast no matter what, but NASA loves their contingencies of contingencies of contingencies. For example, I've read that there were serious doubts that a Return-to-Launch-Site (RTLS) abort would actually, you know, be survivable, and yet it remained one of the official abort modes throughout the shuttle program.

In fact, there was at least one case (and maybe two, on consecutive launches), in which one of the serial SRB holddown bolts did not fire. The SRB, as designed, tore right through it.

Regardig RTLS, commander Young refused to test it on STS-1. Though it was engineered and had code and procedres written for it, RTLS was never really considered a viable option.

I assume the very mechanism that ensured both lit also introduced the latency in the abort mechanism they were concerned about.

> FWIW, even the STS has aborted take-off post-main-engine ignition

This happened five times, in fact, all computer-initiated: https://en.wikipedia.org/wiki/Space_Shuttle_abort_modes#Redu...

Musk tweeted last night that there was a potential leak issue that might cause a scrub during the launch. They added an automatic check for the problem, and it appears that's what caused the scrub.

Since this is a launch to the ISS, the launch window was tiny- miss it by a second, you've missed it completely.

They'll fix the issue and probably try again tomorrow.

Edit: my mistake, Musk has tweeted that there was another probably-unrelated-maybe-issue, and they called it off just to be safe. Still, provided the minor issue doesn't turn out to be something more major, tune in tomorrow at 9:30 am EST.

The issues were unrelated. The leak from yesterday was in the system that spins up the turbopump prior to engine ignition (using compressed helium).

The scrub today was related to some wonky readings on a sensor that measures the position of a hydraulic actuator that controls the direction of the second stage engine (the 'TVC' or 'thrust vector control' actuator).

How tight is the launch window for rendezvousing with something in orbit? A second? A few milliseconds?

It depends on the capabilities of the launcher. For Falcon 9 it's several minutes, but since that's shorter than the amount of time it takes to 'recycle' from an abort (even if the issue can be fixed in zero time), it's effectively an instantaneous window.

Depends on the orbit and the fuel available anything from seconds to hours. A fully loaded Dragon will have an "instantaneous" window, but with spare fuel for corrections during launch it can be an hour or more.

This seems informative:

https://www.reddit.com/r/spacex/comments/3vcoxh/why_spacex_h...

Not super tight, but they don't have a lot of leeway either.

And this is why we write unit tests!

Ars now says that Musk is claiming to have personally scrubbed the otherwise "green" launch due to irregularities he spotted in a control system for the upper stage. Not sure we are getting any full story yet. Not sure he has full a story yet to give.

https://arstechnica.com/science/2017/02/behind-the-scenes-of...

Does he literally have a finger on an abort button? I find that very hard to believe.

NASA provided a live video of the launch prep. Maybe an hour before the launch, the commentator discussed two issues, and said that for both of those issues, they would make a call on it at T-1 minute.

Later on, the video showed SpaceX launch control center, with a group of people huddled together talking - I'm assuming they were discussing the issues.

One of those issues was resolved quite a while before the launch, but not a lot was said about the second issue - the second-stage engine actuator.

The launch was scrubbed at T-00:00:15. In the SpaceX video [1], you hear someone saying "Hold hold hold", followed by "Launch abort is started".

If I were to speculate, I'd suggest that the second-stage anomaly went away, but no-one could explain why. I think the launch was scrubbed, not because someone said "This will go wrong", but because no-one could say "This will go right". Even if Musk did personally scrub the launch, it was for an issue that had been an ongoing discussion for maybe an hour, and that they had already determined would be a (literal) last-minute decision.

[1] https://www.youtube.com/watch?time_continue=1834&v=V5bG37hzw...

Thanks for the link.

I find the people that work on this stuff fascinating!!

To have that much precision (with 15 seconds left, left just stop).

I can right-click and deploy some changes to a production server or run a database script that has a massive impact on the data.

But to imagine that many movings parts (ie those workers specifically involved in the launch) and to have those parts all working together. A database change has a massive impact to "our data". When a launch goes wrong - really puts into perspective what "massive" really means...

Amazing!

think...

#> shutdown -h now

> Does he literally have a finger on an abort button? I find that very hard to believe.

Why? Anyone can call a hold until the T-10 second mark (after that the computer is the only thing that can call an abort). All they have to do is call "Hold, Hold, Hold" on the countdown net, and the person who actually has their finger on the button will press it.

So "anyone" actually means those on that line ... not the engineers at home watching on TV and certainly not the general public

Yes... Is that terribly surprising...?

There was a large grouping of people talking in the control centre visible on stream about T-1m. I imagine it wasn't "I suddenly decided to call it off on a whim at T-10s" more like "I noticed an anomaly, talked to the team, and decided to call it off, the final decision was made by me at T-10s"

Looks like yes.

https://twitter.com/elonmusk/status/832995083519614976

From the replies to that: "During a launch, anybody can call Elon and abort the launch." [0]. It seems like they want to go in the complete opposite direction of when NASA launched the Challenger despite an engineer telling them it would explode [1].

[0] https://twitter.com/zlsadesign/status/832995766436179969

[1] http://www.npr.org/sections/thetwo-way/2016/01/28/464744781/...

Sounds like the Andon cord[1] approach. It does seem to be superior in many contexts to a strict command and control hierarchy.

[1] https://en.wikipedia.org/wiki/Andon_(manufacturing)

Just want to clarify that by "anybody can call Elon" he means "anybody at SpaceX." [0].

[0] https://twitter.com/zlsadesign/status/832995820328804352

"Anyone can call" ... i hope the howard stern crowed dont get wind. With public events you have to filter incomming information. I assume that "anyone" actually means spacex people on the team. He isnt watching twitter.

Think of someone shorting stocks. Even a one-day delay can impact all sorts of companies. A good prank call could make a real difference.

Huh? I've worked on lot's of construction jobs where "anyone" can stop work if they see an unsafe situation. Obviously "anyone" doesn't refer to every human on the planet, it refers to relevant personnel. How is it even possible to be confused by this?

I was also wondering the same as there is another comment above about the exact same thing. Some people on HN really take things too literally.

There's no way he can't just command an abort from wherever he is.

No. With something as important as a space launch, the idea of a 10-second abort based on a phonecall is crazy. They need to be sure it isnt disinformation. That late in the game i would expect all communications to be only from verified sources, sources on a list with dedicated pathways and roles. Millions, billions are on the line.

I have taken issue with musk tweets that have been less than accurate before. Time will tell.

It doesn't cost nearly millions or billions to scrub a launch... It definitely costs that much to have a failed launch.

He was on site with engineers. This is such a bizarre thing to be doubtful of.

I just can't get over how terrible an idea a PayPal founder starting a rocket company sounded. "Good ideas often sound bad," no joke. I was a pretty harsh critic of SpaceX in its early days (I was an aerospace engineer and have followed private spaceflight since its inception) and once they got over some failures and staffing issues, they've really blown me away.

Alt-space community laughs about some aspects of "traditional aerospace" mentality for about quarter of century if not more. It's mostly about their opinion that rockets somehow are rocket science, pardon the pun.

Given how glacial rocketry progress was by early 2000 it's no wonder people from outside of the industry were pretty annoyed, not to say more. Elon Musk isn't exactly alt-space guy - that would closer describe John Carmack or others - but the effect which he produced on world rocket industry is nevertheless pretty healthy.

It's probably hard to believe that SpaceX is just a correction to a industry which got too swamped with status quo. Somebody - may be SpaceX itself - can actually demonstrate not only reasonable results, but significant innovation.

Aeronautics is just as stagnant. Whenever news of new small aircraft startup hits HN though, most comments are negative.

>"Given how glacial rocketry progress was by early 2000 it's no wonder people from outside of the industry were pretty annoyed, not to say more."

I think this is on the money. I think people might get more interested commercially if getting out of LEO per attempt was more reliable vs the current focus on bringing the cost's down with typical systems we have been using in rocketry (which seem to not have increased the reliability). It would be really cool if we heard about advancements in alternative exit strategies[0][1], but it seems we're still far off.

I'm personally the most interested in electromagnetic rail assist for the first stage, with some sort of scram jet for the second stage, best most practical stuff will remain classified military research (looking at the US Naval programs) and the public ONR[2] reports still bring up issues as far as durability of rails over repeated use which probably is going to require some major break through in HEP & matsci.

With the Falcon 9 current specs, the first stage takes about 162 burn time, and assuming its traveling roughly at 1.66 km/s[3], its first stage travels to around around 270 km (ISS is at about 400km in LEO) with a max paylod of about 2.28 * 10^4 kg, I don't wanna be anywhere near the launch of the kinetic EM slug that needs to have at at least ~60,340 MJ of PE between the 1st-2nd stage transition (setting aside any given directional vectors, and if present or near present scramjet array could support such payload). The navy doesn't even expect a 32MJ weapon until mid 2020[5], so who knows, maybe if I can be alive in 60-80 years, we will have built huge CSP arrays in the middle of the desert to be able to launch things into LEO with more reliability.

But another hundred years of redox seems more likely.

[0] https://www.nasa.gov/topics/technology/features/horizontalla...

[1] https://en.wikipedia.org/wiki/Non-rocket_spacelaunch

[2] https://en.wikipedia.org/wiki/Railgun#Outstanding_issues_in_...

[3] https://www.quora.com/How-fast-does-the-first-stage-of-Falco...

[4] https://en.m.wikipedia.org/wiki/Falcon_9

[5] https://news.usni.org/2015/04/14/navsea-details-at-sea-2016-...

Betting on failure of anything ambitious as SpaceX is fairly safe. Most such projects fail. It's not even clear to me in retrospect that there were signs that SpaceX was more likely to succeed. I really sat-up and took notice on one of their early launches where they had an abort and then re-prepped and launched in under 24h (IIRC?). That's something that few others could achieve, and they were doing it cheaper than everyone else.

This reminds me of this video: https://www.youtube.com/watch?v=8P8UKBAOfGo

Pretty harsh.

The right call was made, yet they still need to start showing that they can regularly launch without blowing up on the pad, cargo and all. In other words, this was the right choice in a bad situation.

It's really hard to say they're in a bad situation, given the company's age, the newness of the technology they're flying, and the relative low cost: https://en.wikipedia.org/wiki/List_of_Falcon_9_and_Falcon_He... compared to https://en.wikipedia.org/wiki/List_of_Atlas_launches_(2010%E... or https://en.wikipedia.org/wiki/List_of_Ariane_launches_(2010%...

Six/Seven launches a year puts them quite comparable to Ariane, and only slightly behind the nearly twice as expensive Atlas flights. Nobody is going to fault SpaceX for taking a breather for a day to help ensure their hardware reaches orbit rather than launching through an anomaly and raising the risk of a failure.

SpaceX need to show they can launch frequently, period. Higher launch frequency with lower reliability is a market that aligns with SpaceX's commercial customers.

If I have a hugely valuable payload, why would I entrust it to a company that has a reputation of launching when they shouldn't?

Scrubbing a launch is a much smaller deal than blowing up and losing the payload.

Yeah, I mean they're both undesirable outcomes given a competitive market, but one is "sub-optimal", the other is a catastrophe.

I highly doubt that, considering how costly the rocket and the payload are.

It would probably drive up insurance costs significantly too.

The webcast showed T - 0:00:13

That's what the timer reading froze at, but is that the same as the time when an abort sequence was initiated?

This is the first time that I saw a livestream where an abort happened, so I'm curious. Is the timer freeze standard?

  is that the same as the time when an abort sequence was initiated?

You can watch it on SpaceX' video - the instruction 'hold hold hold" is given at T-00:00:15, the clock freezes a couple seconds later, and shortly after that, you hear "launch abort is started".

https://www.youtube.com/watch?time_continue=1834&v=V5bG37hzw...

  Is the timer freeze standard?

It's mission dependent. Some missions allow for the clock to be paused, then resumed [1]. Some missions have prearranged holds [2] built into the mission plan.

For a mission to the ISS, the launch window is too short to hold then resume, so a hold on an ISS launch is effectively a launch abort.

[1] https://en.wikipedia.org/wiki/Space_shuttle_launch_countdown

[2] https://en.wikipedia.org/wiki/Built-in_hold

Reply on news.ycombinator.com