Hacker News new | past | comments | ask | show | jobs | submit login
SpaceX nails triple booster landing after satellite delivery (bbc.com)
513 points by support_ribbons on Apr 12, 2019 | hide | past | favorite | 120 comments

This really looks good for launch costs long-term.

The Heavy costs $90 million per reusable launch with a 64,000 kg payload, or $1400 per kilogram.[1]

Current versions of the Falcon 9 and Heavy can fly ten times with virtually no refurbishment between flights.[2] The only part they throw away is the $7.5 million upper stage.[3] Their expendable cost is $150M and they haven't really started reusing yet, so if they actually reuse each rocket ten times, they have a lot of room to lower prices; ten launches would be $($150M - $7.5M) + ($7.5M * 10) = $21.75M per launch, or $334/kg for the rocket itself. Launch cost won't be quite that low because they also have labor, fuel, and so on, but it looks like they can get well under $1000/kg just with the Falcon Heavy.

The larger and fully-reusable BFR should do even better.

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

[2] https://www.space.com/spacex-falcon-heavy-triple-rocket-land...

[3] https://www.nextbigfuture.com/2019/04/spacex-recovered-6-mil...

Note that a 64,000 kg payload LEO launch doesn't cost $90 million because it would have to fly expendable.

For the moment they also throw away the fairing, which I assume isn't included in your $7.5 million as it's been reported to cost $6 million by itself (at least for regular Falcon 9 -- not sure if the Heavy fairing is different). They've been working on fairing recovery, but have yet to succeed.

EDIT: scratch that, apparently the recovered the fairings as well per a comment below. Neat!

the fully reusable falcon heavy can IIRC only lift about 8000kg to LEO. expendable is volume-limited but if you want to lift a concrete block, you can :)

Reusable is 8000kg to GTO, not LEO. LEO performance in fully reusable configurations is 30+ tons.

So assuming 32 tonnes it's $668/kg for the hardware (assuming full payloads).

Probably also have to build in enough slack for occasional failed landings, but we don't know yet what the failure rate will be after the kinks are worked out.

You'd be hard pressed to fit 32 tons inside Falcon Heavy's fairing. It's far more likely to be volume constrained than mass constrained when flying to LEO.

That could bring satellite costs down a lot though - if launches are cheap and not weight limited a cheap steel frame satellite with heavy solar panels becomes feasible - all the old rules about building to optimise weight and not price go out the window.

I wouldn’t be surprised if spacex also enter the satellite building business after mass producing starlink.

And that makes solar power satellites more feasible, especially with the BFR.

Musk has expressed skepticism, famously asking "what's the conversion rate?" The answer is that microwave transmission is 40% efficient with today's technology. This isn't bad considering that average daily sunlight in geosynch is five times higher than on the ground,[1] and you don't need much storage or backup.

The early SPS designs were hugely expensive monolithic beasts, but current designs use lots of identical parts (of several varieties) that self-assemble in orbit, so you get economies of scale with factory production.

A year ago I read the book The Case for Space Solar Power [2], which broke down the costs in detail. With pre-SpaceX launch costs they got 15 cents/kWh, including manufacturing and ground stations. I plugged in the BFR cost Musk was claiming (I think I used $50/lb) and got about 4 cents/kWh.

[1] https://en.wikipedia.org/wiki/Solar_irradiance#Irradiance_on...

[2] https://www.amazon.com/Case-Space-Solar-Power-ebook/dp/B00HN...

Hmm, what's the cost to put solar panels in heliocentric orbit? Dyson spheres anyone?

Light weight satellite construction will still be a very high priority because making the structure of the satellite lighter extends its operational lifespan.

The reason is attitude control. Reaction wheels can be used to orient the satellite, but eventually reaction wheels get saturated. At that point you need to use thrusters to cancel out the reaction wheel saturation. Those thrusters require propellant, e.g. something to push against. The greater the ratio between propellant mass and structural mass the satellite has, the longer it can remain operational before the propellant runs out and your reaction wheels finally saturate, leaving your satellite tumbling useless without attitude control.

That's a good point, though can add more propellant too if you're not mass constrained.

It will be interesting to see if Falcon Heavy, and more likely starship, have a significant impact on what sort of satellites are built, and the way they are built. If they do it will probably take years before we see any changes, however spacex is ideally positioned to cash in on a transition to more mass-produced satellites if that does happen.

What if SpaceX could combine cheap satellites with some form of reusable propellant-less or super high ISP propulsion? I'm thinking of solar sail or solar-ion drive tugs.

Not only that, but apparently they recovered both fairings: https://www.google.com/url?sa=t&source=web&rct=j&url=https:/...

So 5 out of 5...

It's an interesting issue with the fairings, and I'm interested to see what (or if) they're going to solve the issue with the saltwater contaminated fairings.

Being a SpaceX launcher, as the article indicates, means they can accept higher risks, but in turn has a greater financial risk to the company. It's times like this where I'm glad SpaceX is a private company.

According to a reddit thread they've saltwater hardened the fairings and now they can reuse them:


The first reuse mission seems to be an internal one though.

That's just speculation, I wouldn't cite it as an authoritative source.

The part that says they're able to reuse these ones is a tweet directly from Elon Musk; he also confirms they were recovered from in the water, not caught in the air.

That information was already covered by the news article linked in the root comment above.

The additional suggestion by a commenter in the Reddit thread that "they've saltwater hardened the fairings" is speculation.

>The part that says they're able to reuse these ones is a tweet directly from Elon Musk;

And we know that Elon Musk only ever tweets true information.

> in turn has a greater financial risk to the company

A financial risk you can count as R&D cost.

SpaceX has also positioned itself for some launches that would be acceptable to loose, like the deployment phase of starlink: loosing a few mass-produced satellites has to be much cheaper than loosing a big one-off satellite that takes months to rebuild.

Same link but this one doesn't report your click to google: https://www.teslarati.com/spacex-falcon-heavy-fairing-recove...

I see multiple requests to ampproject on the official page, so Google is getting data regardless.

Not that I'm avoiding google, but this link connects to

google.com googlesyndication.com googletagmanager.com gstatic.com

When rendering the page, so this just proves that avoiding sending data to google is very tricky indeed!

It does work for those using a content blocker or tracker-blacklisted DNS service.

I guess salt water isn't that much of an issue after all?

Yea, I wish we'd get a more detailed explanation of the accounting on this. For all I knew it was totally plausible that fairings would be destroyed by sea water, and it was also plausible there'd be no damage. The part that confuses me is why they would spend so much effort trying to catch them with Mr. Steven if they are in fact serviceable after a swim. Maybe the idea is that salt-water refurbishment-and-reuse saves a bit of money compared to expendable, but mid-air-catch would have saved more but turned out to be too difficult?

I think their customers would be more comfortable with (re)using a Caught fairing, so that was Plan A. Also not all fairings are the same. Most have acoustic blankets and other components that need to maintain strict ratings. Remember, many payloads are more expensive than the rocket and the fairing requirements are usually margined-up by the satellite manufacturer to avoid blame for failure.

The fairings are also quite big, so I imagine they get destroyed by waves very quickly if the sea isn't completely calm.

There was a sibling comment which mentioned some kind of anti-salt-water treatment, I guess that was Plan B and crossed the finish line while Plan A was still struggling?

Developing the recovery process has given them the chance to examine multiple fairings after they landed in the ocean. I'm sure they took the opportunity to see what refurbishing them would take, and perhaps that's what informed this decision.

I think that is only an issue with hard to clean equipment, like engines.

SpaceX thought it was enough of a problem for the fairings that they built a boat [1] to catch them before they hit the water. Either they've redesigned the fairings or they've re-evaluated their post-recovery process to make it possible to reuse fairings recovered from the water.

[1]: https://www.space.com/41614-spacex-mr-steven-catcher-boat-up...

So now they have recovered the fairings? SpaceX just keeps getting farther and farther ahead of everyone else.

There's a great media thread on reddit where people post their own photos and videos: https://www.reddit.com/r/spacex/comments/bbhz9a/rspacex_arab...

Here's a great video of the entire flight and landing of the side boosters in one shot, made with custom tracking software: https://www.youtube.com/watch?v=cEZZkEXAD6Q

The boostback burns in that video are amazing. Lots of dynamics in what looks like a shock-induced interface of the rocket exhaust and hypersonic atmosphere.

For those interested, booster separation is just after the 3:00 mark. The whole separation, flip, and boost-back dance is incredible to watch.

I never realized there were launch observers in the water - what a great idea!

Nice video capturing the sonic booms from the booster landings here: https://twitter.com/jefffwilliams/status/1116486329284595717

Yep, shared that with the office and family. That's amazing. It really does feel like the future.

I’m always surprised at how loud the single engine landing burns are.

Big fan of making technical spacecraft manoeuvres sound like cool skateboard tricks.

I'm more impressed by the booster landings though :-)

Seeing a launch/landing in-person is on my bucket list.

I saw the live stream yesterday and when the two side boosters landed side-by-side synchronously it was almost as if it was out of a movie. We are living in the future!

On this launch, it looks like they offset the timings to reduce the chance of tipping over.

If you get time, go look for the first FH booster landings. The timing is almost perfect and the long exposure images looked beautiful.

They are supposed to be offset for radar purposes. The fact the first time was so close was a mistake, fortunately it worked.

>On this launch, it looks like they offset the timings to reduce the chance of tipping over.

On reddit a lot of people enphasised that it was to avoid interference. Which interference, no one knows.

Incorrect. Gift that you give to me, no one knows,

It's certainly very cool. I saw the night launch/landing of Falcon 9 out of Vandenberg a few months back. Certainly worth the overnight trip.

Most impressive part for me was that you see the booster make it's landing burn and touchdown BEFORE you hear the landing at the observation site.


Every single time I watch a successful SpaceX launch I get chills and also get so excited. Congrats to everyone who has worked on this.

I was at Kennedy Space Center's Apollo/Saturn V viewing site for this launch and it was amazing. If you ever have the opportunity to be there for a Falcon Heavy launch, do yourself a favor and go. It's quite an experience and totally worth the money.

Yeah, if you have a week of free time. I planned to go this weekend but chickened out at the last second-turned out in my favor since they launched yesterday, four days after the original date.

Yes, there may be delays or they may postpone the launch. I went on Wednesday and decided to stay one more day for yesterday's attempt and I am so glad I did. It was worth it.

I've seen a falcon launch (not the falcon heavy). It is indeed amazing

What was amazing about it? Tell us the story.

Aside from the whole historical factors of the location itself (for instance: the humans sent to the moon were launched from that same launch pad), watching the launch from a video or live stream does not tell the entire story.

To hear the roar of those 27 engines and feel it vibrating on your chest and the bleachers is an emotional experience. Then you can see the boosters coming back, and their entry burn and you can follow them down to their landing zones. A few seconds later you hear the sonic booms.

While all this is happening, you also have the energy of the happy crowd cheering and celebrating what humans are capable of doing.

It gives you hope and an exciting future to look forward.

It's definitely the sound that most can't be reproduced at home. It really sounds like the air itself is being ripped apart. That's why I want to go back and see another launch. Also, the flame is SO BRIGHT, video doesn't do it justice. It's miles and miles away and it's almost like looking at a welding flare.

Here's the video cued up to the booster landings: https://youtu.be/TXMGu2d8c8g?t=1643 with the main core landing a few minutes later.

If it's not obvious, the main core landing was much harder this time around, because of how fast it was going due to the two boosters imparting a lot of extra speed on it before they separated. If you think about it, it'll have somewhere between 1x and 3x the velocity of a normal Falcon 9 core

It doesn't necessarily make it harder. There's a burn to slow the booster down for re-entry, and having the side boosters do a lot of the work means more fuel / delta-v can be saved to slow the core booster.

The FH is apparently capable of 8,000kg in a reusable GTO launch ( https://en.wikipedia.org/wiki/Falcon_Heavy#Capabilities ), and this satellite was only 6,000kg. So it looks like there would be plenty of fuel available to slow the core down well below the survivable maximum.

It's the same steps but it is harder, the center booster was going faster than any other core that has been recovered. The recovery itself was ~300 miles further out than any other recovery.

Swinging a bat at a 10mph ball is the same technology as swinging at a 90mph one. I don't think anyone would say that they were equally as easy.

My point is that it's less a question of technical difficulty / tolerance, but more fuel / delta-v / mass cost.

In the case of a bat swing, to bat the ball faster, your muscles need to be stronger, and the ball and bat need to withstand greater pressures. In the case of a vehicle travelling through space / thin atmosphere, it only needs more fuel to burn to slow it down. Analogously, you don't parallel park a car from 70mph, having travelled somewhere on a highway.

Here's a graph a booster's speed over time, showing the reentry burn: https://i.stack.imgur.com/xFYIh.png ( from https://space.stackexchange.com/questions/20246/what-is-this... )

Can we finally stop having sci-fi films in which the whole outcome of the movie is solely determined on some heroic manual flying of a spaceship while simultaneously using manually aimed weapons like it's World War II?

> Can we finally stop having sci-fi films in which the whole outcome of the movie is solely determined on some heroic manual flying of a spaceship while simultaneously using manually aimed weapons like it's World War II?

Probably no, we can't stop having some of that. We can have her alternative, and, in fact, a lot of space combat in movies and TV is not resolved that way (both because it involves highly automated maneuver and targeting, and because, especially for combats that determine the outcome of the story, it is resolved by a social or technological advantage, often one provided by the quest making up the body of the plot, not by maneuvering or targeting (except to the extent that technical or social advantage impacts those abilities.)

I like that about the Expanse novels, automatic systems control a lot of the lower level work in battles. Of course humans still come up with the crazy ideas and sort of execute them on a higher level.

The Apollo 13 astronauts would like to share a couple of anecdotes with you.

Apollo 13 (the movie) had plenty of careful checklist readbacks and minimal explosions or shaky cameras. Most scifi is more like Michael Bay in space. See Gravity for a more egregious example of a movie that tries to play itself off as "hard scifi" but ends up just being an over the top action movie.

Yeah, The Martian is like, the hardest sci-fi book I've ever read, but I found the film adaptation... lacking

Seems like it'd be tricky to do from a story telling perspective.

Totally removes all agency from protagonists. Every fight becomes a deus ex machina - "welp, our targetting subroutines continue to be superior to their maneuvering systems so they died. Hooray us?"

Or I guess more probably just a numbers game: "welp, our effective weapons range is larger than theirs, so that's that" with an MCRN exemption clause.

Isn't this how real-world warfare is trending already? A vehicle full of soldiers may die before they even know there was a threat flying overhead. But it's not very heroic-looking to just keep the laser designator on target until the bomb lands.

I suppose that's why so many contemporary thrillers or even far-future science fiction movies devolve into hand to hand combat for the climactic showdown. It looks more interesting. (Though personally it makes me groan and want to throw things at the screen.)

You can't really write a compelling story where your logistics make the hero rather than... heroism.

I wonder what the makers of Zero Dark Thirty did to keep it compelling. I haven't seen the movie myself. The real-life mission resulted in only one light injury among the Americans and involved zero fist fights. Osama bin Laden himself was killed before he could fire a single shot or throw a punch.

> Every fight becomes a deus ex machina

Sounds like a book by Iain M. Banks.

My favorite sci-fi anticipates increased autonomy of technology and incorporates that smoothly into the narrative.

Also, can we start seeing this kind of re-usable multi-stage rocket used for achieving orbit in fiction?

Maybe look closely at the landing on the Hudson river. A good sci-fi plotline for getting out of hairy situations is heroic manual flying that is only made possible due to the automated systems.

If you're talking about Star Wars, definitely not, because it's intentionally unrealistic.

Other sci-fi, maybe.

I find this, and SpaceX, so inspiring. It's easy to feel that technology just moves itself forward. After all there are billions of people on Earth with people working on a countless array of different problems. And each day things seem to move forward, almost inexorably so. And so it's easy to feel that the value of the individual is really relatively low.

Yet imagine a world where Elon did not exist and thus SpaceX did not. Much of what he's done is stuff that we could have been working on decades ago. For some time we were. NASA as early as the 70s had already laid out plans for a Mars expedition including a tremendous space ship that would be assembled and fueled in orbit to take 5 man crew on a 600 day manned expedition to Mars, including landing of rovers similar to the moon.

Those plans got canned by Nixon, and space never really recovered. Not only did we not "inexorably advance" in space, we regressed. Today we're struggling to do a manned flyby of the Moon - when we went to having barely put a man in orbit in 1962, to putting a man on the moon in 1969. The point of this is that technology does not advance by itself, let alone inexorably so. I think it's extremely likely that had SpaceX not come to exist, it's entirely possible that we would still be effectively where we were at near the turn of the century.

Progress of our species, in spite of there being billions of us, is still dependent upon the individual. And SpaceX's plans have very much followed the old quote of Gandhi, "First they ignore you, then they laugh at you, then they fight you, then you win." But of course we've not yet won. As remarkable as this achievement is, it's but the starting line for where we need to be. And that line will not move forward unless we move it forward. But "we" does not mean waiting for somebody else to do so. As SpaceX and Musk have demonstrated, it's ultimately up to the individual to get up and move that line forward -- for the betterment of all.

if you want to talk inspiring, consider Gwynne Shotwell. She's tasked with turning Musk's crazy ideas and timelines into reality. I can't imagine a tougher job and she's killing it.

Gwynne Shotwell is pretty amazing. But don't forget that Musk was smart enough to hire her, and she had enough confidence in him to take the job. A very large part of being a successful tech revolutionary is hiring the right people to work for you.

I always read about Musk like he does the whole thing on his own. Never read about all the engineers that work their asses off to achieve these results

The way I see ot, ot goes both ways. He has taken a lot of shit for Tesla,SpaceX so obviously when something good happens then too he gets a lot. The dynamic is a bit similar to actors/actresses in movies. If it flops, nobody is blaming the makeup artist,dialogue writer. Btw, i do find some things related to Elon's life overglorified(those parts do not involve SpaceX,Tesla) BTW, read this thread about Tom Mueller. https://mobile.twitter.com/mekkaokereke/status/1081619342377...

Really inspiring. This thread got picked up by many news sites. I just googled "spaceX engineer lumberjack" to find this.

He constantly emphasizes that it's the entire team bringing SpaceX to what it is. However, there is an important difference between working for a company and leading one. Boeing/Lockheed, Arianespace, NASA, ESA, Roscosmos, etc, etc not also only have brilliant scientists and engineers working for them, but also have billions of dollars and have been around for many decades.

The big difference in outcomes comes from the leadership. These other companies are headed by people of different abilities, different ideologies, different technical abilities, and so on. This is why they've all been, more or less, surpassed by a company started on a shoestring budget in 2002.

In other words, I imagine if you had let Musk have complete control over ULA (united launch alliance - an arguably anticompetitive merger between Boeing and Lockheed) he'd not only achieved as much as he has already, but likely far more given their vast resources - both economic and human. Similarly, if you put Marilyn Hewson [1] (president, chair, CEO of Lockheed) in charge of SpaceX's people in 2002, SpaceX almost certainly would have catastrophically failed.

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

Large existing discussion at https://news.ycombinator.com/item?id=19639965

For some reason the discussion in this thread is far better than the earlier one you linked.

I wonder how they do the booster landings?

Is it precise modelling and standard control algorithms, or something more exotic like neural networks?

It's done using convex optimization


How is "booster landing" mapped to a convex optimization problem?

Lars Blackmore, who works for SpaceX, has a page about the techniques that are used here:


See also: https://www.aa.washington.edu/facultyfinder/behcet-acikmese

Can anyone give a ELI5 (for developers) of how this technique works?

The control parameter space for the rocket is non-convex because, for one thing, it can't be throttled down close to zero thrust. You can think of the parameter space as having a hole in it. An optimal control algorithm might want to make very delicate low-thrust corrections, but that's not possible.

This technique is based on "the idea of relaxing the nonconvex control constraints to a convex set in such a way that the optimal solution to the relaxed problem is guaranteed to be the optimal solution to the original problem."

By "relaxing", what they mean is that the new model (the convex set) actually contains some parameter values that aren't achievable, but OTOH it is geometrically susceptible to being analyzed by an efficient optimization technique. So it's a clever replacement of a hard problem with an easier one. The complex bit, which I don't understand, is how they show that this replacement will always give the same value as if they had solved the (real) hard problem, i.e. that the solution will never actually use the "illegal" parameter values.

Maybe someone else can give more insight.

I read a paper on this—it's convex because any optimal trajectory will have the engines at full throttle whenever they're on, so they just fill up the parameter space, confident that illegal values will never be required.

can you link to the paper please?

I don't know which paper they mean, but here's a relevant paragraph from one of the papers available from Blackmore's page:

"In this paper, we unify the convex optimization approaches of [1], [2], [17] and extend them to handle thrust pointing constraints. While convexifying the problem with nonconvex thrust pointing constraints, we develop a geometrical insight into the problem that establishes a connection with “normal systems” [18]. A normal linear system is defined in the context of optimal control theory where the system is said to be normal with respect a set of feasible controls if it maximizes the Hamiltonian at a unique point of the set of feasible controls. In the case when the set of feasible controls is convex, a system being normal implies that the Hamiltonian is maximized at an extreme point of the set [18]. Our convexification result has a similar geometric interpretation since it establishes lossless convexification by ensuring that the Hamiltonian is maximized at the extreme points of a projection of the relaxed set of feasible controls. This set is then shown to be contained in the original nonconvex set of feasible controls, thereby estab- lishing that we can obtain optimal solutions of the original nonconvex problem via solving its convex relaxation."

I wonder what emergency procedure is in place in case the landing is off-course. Does it blow itself up?

As far as I understand, the Flight Termination System is safed before the landing burn. It's more dangerous to have it potentially still active after landing. So at this point it can't blow itself up anymore.

Until the last moment it also doesn't target the barge itself, but slightly to the side. Only very shortly before landing does it correct the target to the actual landing position. So in most cases it would simply hit the water instead of anything else when there is a major issue.

If you watch the livestream for yesterday's falcon heavy launch on youtube the second camera angle is a rendering showing trajectory, and you can actually see the center core adjusting closer and closer to the droneship with each step as it gets more confident that it's able to land safely.

The booster's trajectory targets offshore or away from droneship in case landing burn fails. The landing burn brings the booster onto the landing pad. See CRS-16 where the booster landed in the water when the booster lost roll control when the grid fins failed.

They've had a few "on purposes misses" over the years. The Falcon Heavy demo launch had the center core purposely miss the droneship as it was coming in way too fast since 2 of the engines didn't light.

IIRC it purposely missed the droneship by only a few hundred meters, but it was going something like 300 miles per hour!

CRS-16 landed in the water offshore when it lost control.



Almost certainly not - there's all sorts of precedent from rolled-away cars to high wind incidents that just because something has ended up on your property does not make it yours unless the owner has abandoned or discarded it.

No, but if one crashes in the ocean, sinks, and SpaceX does not attempt to salvage it, it legally becomes a derelict and the property of whoever who goes lift it up.

Probably not, but maybe you could fine them for littering. http://mentalfloss.com/article/70708/nasas-unpaid-400-litter...

Most likely not, but considering my house would likely be on fire and the hearing damage, I feel a booster would be decent compensation.

you can, as long as SpaceX recovery truck has not reached your door

I'll swap the house numbers with next door. They'll never find it.

I know this is increasingly off-topic for hn, but the mental image of being woken up by loud booster noises, then seeing a huge booster sitting in your now-ruined garden is hilarious to me.

We certainly do live in the future.

In the anime "Rocket Girls" they landed in the school's swimming pool.

Can someone explain why reusable mode can't lift as much?

Because it has to reserve fuel for landing.

Am I going mad I only see two booster landing in the video?

The two side boosters were landed close together on dry land. The core third booster was landed down-range on a sea platform.

On the live stream, the video feed cut out for the landing of the third (middle) booster.

This almost always happens. It's really hard to keep a stable satellite connection, while a 150ft. booster is blasting the deck with the exhaust from 3 firing rocket engines.

It doesn't fire all three engines down to the deck. The landing burn starts with one engine (in order to have more control authority when the two outboard engines start up), then two more engines get lit, then they shut down shortly before the actual landing (reducing the TWR makes the landing a bit easier).

The way the text appears with a downward-drawing red line really steals my attention. There are rockets on screen and I kept having my eyes avert to the fragmented text in the bottom left corner. I had to watch the video twice to keep my eyes on the rockets. I'm thinking of the UI/UX implications, subtle and effective.

It's just a newsflash, the rockets themselves are a backdrop.

For the real thing, watch the SpaceX video directly from YouTube: https://youtu.be/TXMGu2d8c8g?t=1181

Write them an angry letter.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact