[Edit: First the video was public, then it was made private, now it seems to be "unlisted" but viewable. So no guarantee that it'll still be there in another minute...]
Why is this a good thing? Because it proves a point: the rocket automatically adjusted its trajectory, and continued on its course as if nothing had happened.
The Falcon 9 is the first American rocket since the Saturn 5 with an engine-out capability. Its multiple symmetrically-arrayed engines allow for it to compensate for a sudden loss of thrust from one side: the other side reduces thrust as well to stay balanced, and everything else just burns a little bit longer. That was the theory, anyways. Today they put it into practice.
For any other rocket that you've seen launched in the last 35 years, that would have ended the mission catastrophically -- but the Falcon shrugged it off like nothing had happened. They always said they could do that, and now they've done it. Congratulations, SpaceX! Creating a fault-tolerant rocket is much better than creating a faultless one (since that's impossible).
[Edit]: In the absence of the video, the engine anomaly is being widely reported on Twitter: https://twitter.com/spaceteam/status/255128401927610368
Then again, it is unfair of me to make relatively uninformed declaratives about tonight's launch. The engine blowout could have been cause by an incredibly unlucky set of circumstances, and the engineers at SpaceX could have built a vehicle that survived a catastrophic failure that would have doomed any other craft to date.
Given what I've seen from SpaceX, I am positive that they will get to the bottom of the incident and have it corrected before their next launch.
No doubt they should do an investigation to see if the cause of failure can be determined and corrected. It's possible that the fault may no longer be an issue; this was flying the soon-to-be-retired Merlin C engines, and the design or manufacturing flaw may not have carried over to the Merlin D engines. But what they shouldn't do is spend billions trying to make the rocket perfect, which has been the traditional approach. A few million + redundancy is much better.
This has been done many times in history, in Saturn V and Space Shuttle for example.
But in this case, since the engines are so small, it seems the explosion tactic paid off very well.
Hopefully they can test Merlin-1D adequately. If you aim for reusability, your reliability must go up considerably. If you plan to fly something for 100 times, you probably can not tolerate 98% per-flight reliability. A lot depends on assumptions of course.
And you're right -- for reusability, they do need to get their per-engine reliability above 98%. That would produce a mission reliability of 97.1% (assuming that 2 engines out causes a loss of vehicle, which I understand is only the case during the earlier phases of flight). At 99% engine reliability, you get 99.3% per-flight reliability -- which strikes me as the point where reusability starts to become very sensible. I have considerable confidence that SpaceX will be able to do this.
Bringing a RAID approach to engine configurations seems to have paid off quite handsomely.
SpaceX is calling it an "unplanned shutdown", but it's unclear at what point the shutdown was commanded. It's possible that some off-nominal parameter triggered an engine shutdown, at which point the sudden vacuum inside the engine bell would have caused the nozzle to collapse. (Note that this happened at Max-Q, the moment of maximum aerodynamic pressure, so the outside force on the nozzle would have been significant). This would have had some impact on the aerodynamics of the vehicle (with the flight dynamics software compensating perfectly), but it wouldn't really have been a threat to the other engines, and is not too dissimilar from some unplanned shutdowns which the Saturn V experienced. It wouldn't really count as an "explosion". It might just be a (relatively benign) software bug which triggered the shutdown.
On the other hand, the engine might have disassembled itself first, followed by the shutdown order to keep it from dumping uncombusted propellant. This would be an actual explosion of the engine -- contained by the kevlar shielding to keep it from damaging the surrounding engines. If this is the case, then it's a huge validation of SpaceX's RAID-like design philosophy, since I don't believe that any launch vehicle has ever survived such an explosion, much less completed its mission to perfection.
I'm responsible for webcasting Copenhagen Suborbitals launches, and we're having a lot of great discussions about what works and what doesn't in a live webcast of a launch. We obviously want our webcasts to be as interesting as possible, and so I'd like to ask you a question:
What do you think could be done better in this Space X webcast? Is there something you're missing? Something you think would be cool? Something that's bothering you? Are the speakers good? Why?
I'm sure we can learn a lot by asking potential viewers what they think, and implement it for next summers launches.
And good luck to SpaceX!
For example, they mentioned they don't want the rocket to fly through clouds because of water vapor. Why is that? How does it interfere with the rocket? What can it cause? How confidently can we measure water vapor in atmosphere?
Simply, much more technical information and more educational.
And Dragon is in orbit! Congratulations to the SpaceX team!
I'm sure there are some steps before, between, and after those two. :)
It would also make it more exciting to watch, since it makes the sense of progression and success more tangible.
The speakers reading from a prompter is a bit too obvious at times but this is a minor annoyance. It's awesome that they are actual engineers/managers involved with SpaceX's day-to-day operations.
(It might make sense to setup an off-site poll for this, I'm not sure Hacker News' commenting system will scale well for this kind of feedback.)
I realize many engineers are not good at speaking to general audiences since they might use jargon or are really slow, stutter etc.. Probably they can improve a lot if you pay attention to it and consider it important and devote the necessary resources like time and money.
I saw the news the other day that you guys had written a video mixer for Linux, with integrated Chyrons, transitions, and all that. That's great work, but I do have a question about the motivation behind it.
As a former television technical director, I was curious what made you choose to write a custom software video mixer. The alternative, of course, would be picking up a cheap digital mixer, like a Grass Valley, and encoding its program output instead. Was the impetus financial in nature? Technical, since making a video mixer train work is a little challenging unless you have experience (reference signals, CCU if you have fancy-enough cameras, and so on)? Are you mixing audio in your software suite as well, or do you have an external mixer? (The rationale for each is the same, IMO.)
I'm just curious. Not saying the work isn't valued, just steaming ahead on writing a custom software mixer is an interesting decision. Production houses usually have old digital mixers in storage and might be willing to donate to your effort, too.
Launches are done in the Baltic Sea, 30-40 km. east of the Danish island Bornholm. In order to stream live we managed to get a wi-fi connection going from Bornholm to the mission control ship 40 kilometers away, but we couldn't reliably get enough bandwidth to stream all our camera positions to Bornholm. Furthermore internet connection on Bornholm is not very good. The result being that we needed to do the videomixing at sea. Only problem was that the broadcast studio, and thus the people that need to do the vdeomixing, were placed in Copenhagen. The result being this videomixer where we can send two streams from the launch site at sea: the livefeed and a mosaiq of available camera positions. The studio in Copenhagen can then remotely mix the livefeed based on the mosaiq using a webbased interface, or a commandline.
Right on, thanks for answering.
Very cool stuff.
In example, think how meteorologists on TV use color on a map to show the how much rain has fallen.
dark green -> green -> dark yellow -> yellow -> orange -> red -> blinking purple
Just use a gradient with colors that won't interfere with whatever map you're using. Maybe just "yellow -> orange -> red" as it progresses to its target. Then once it hits it, switch to white, and continue to chart the movement as long as you broadcast.
Congratulations to the whole team at SpaceX!
Go for quality information, even if it's a bit rough around the edges from a BS marketing standpoint.
Get a REAL rocket scientist spreading the word.
gives me the shivers every time
A timeline of what is supposed to happen - Since we had no sound we were left wondering what was happening, and when the next event should be. It would be nice to have something like:
T-30 - XXX
T-5 - YYY
T-0 - Launch
T+2 - ZZZ
NASA contracted to launch the Curiosity rover on the Atlas V rocket, for example, so how is this different? Well, the Atlas V wasn't developed commercially, it was developed as part of the EELV program. This was a USAF program which sought to develop launch vehicles to carry military payloads into orbit in the future. Boeing and Lockheed Martin developed those launch vehicles (which became the Delta IV and Atlas V) but government had a lot of say on the design of the vehicles and ultimately paid for most of the development cost of the vehicles. This is different from SpaceX's example where they developed their own launcher from scratch to their own design, specs, and manufacturing and management processes, etc.
Also, if you look at other NASA launches in recent history the contrast is even stronger. NASA drove and funded development of the Delta II, the Shuttle, the Saturn V, etc, etc.
Additionally, operations for the spacecraft component, the Dragon, is being handled by SpaceX themselves and not by NASA, which is a new thing, especially in the realm of manned spaceflight.
NASA isn't just buying launches, nor are they just buying Dragon spacecraft and launches, they are buying cargo delivered to ISS.
Beyond that, the interesting bit I think is the difference in the contracts that SpaceX has, and the contracts that have traditionally been given to space contractors. Traditionally overruns would be covered by government funding, but in the case of SpaceX they eat their own overruns. The result of this seemingly minor change in funding is that they are far less "quasi-government" than space contractors traditionally have been.
Also, and this is big, frankly SpaceX is operated in a much more nimble and entrepreneurial fashion. The primes, while having lots of very smart and well-meaning people working for them, have BDC-itis. Musk wants to dramatically improve the world and the species, and put a colony on Mars. The leadership of the primes seem to just want to make money. Musk/SpaceX is more exciting and compelling for a lot of folks, plus they're more likely to lead to spin-off missions down the road.
Early rocket programs were military programs, rockets make excellent weapons, they can reach around the world and touch someone in less than 90 minutes. The military paid for, and managed, programs for putting payloads into orbit. Of course the military doesn't have a rocket factory, so it hires contractors to build rockets to its specifications, and to research questions the military wants answers too, Etc. NASA worked with the same contractors as part of the civilian space program to give scientists access to space.
In the military programs, function was important, cost was not. That is because the military has "cost plus" contracting, they say "we need a rocket that does X" and the contractor builds it and charges the cost of building it plus the agreed upon 'profit' and the military pays it. There is no incentive to be economical and no incentive to change.
SpaceX on the other hand invested its own money in building its rockets (that's the 'private' part) and they are selling services on these rockets at a price that makes them a profit on their investment (that is the 'commercial' part). They have been re-using research done by NASA and the defense contractors in their work so they have an advantage over earlier efforts in terms of costs. That said, by approaching it as a business rather than a government contract they have removed a huge amount of cost out of the process.
Because of the process, SpaceX isn't necessarily constrained by the restrictions a nation-state might put on them (although in reality if they tried to sell launch services to say Iran the US would prevent that) They also have the right to re-use their technology for their own enterprise so if Bigelowe Aerospace needs someone to launch their space hotels, SpaceX can offer that service without either Bigelow or SpaceX needing 'permission' from the military.
The amazing bit here is that nation-state space programs were possible because nobody expected to 'make' money, they just took tax payer money and burned it at a high rate :-). In this case SpaceX is looking to recoup their development costs and then some. Not being under the whims of a nation-state budgeting authority (like the US Congress) allows them to make progress quickly.
Previous systems had heavy input for specs and design from NASA. With this, NASA basically says, "X pounds for Y dollars to ISS" and SpaceX is figuring out all the details.
Think of it like the difference between the cost of developing a new product such as an iPhone before scaling it out (cost plus) and the cost of the delivery of the final product (a final complete iPhone from the city store).
Every time I watch one of these I feel that I'm watching history being made. And I wished I could stop holding my breath through the last 20 seconds or so. I also feel like I'm 6 years old again.
It's about turnaround time and precision. Parachutes are messy. They use pyrotechnics to deploy, they have to be packed very carefully and precisely, etc. They also don't lend themselves to precision landings well. They're fine for landing in an enormous landing zone in the ocean or in some unpopulated, featureless desert, but otherwise they're not so good for targeted landings.
Now, compare this to propulsive landings. You "waste" a lot more fuel and cut into the payload of your vehicle but on the other hand you get to streamline the turnaround process to just a matter of refueling and inspection, which you would do regardless of the landing mode. More so, by landing the vehicle in a designated spot you dramatically lower the cost and time necessary to put the vehicle back into service. You don't have to drive out to the desert and retrieve the vehicle (which could take a day at least) you just pick the thing up with a crane and you move it a few hundred meters to the hangar / processing facility.
First, reliability. When you engineer a vehicle as large and as complex as an orbital launcher which only operates once in its entire lifetime you typically need to over-engineer a lot of key parts of the thing in order to ensure a high level of overall reliability. This runs counter to the sorts of optimizations necessary to bring production costs down by orders of magnitude. More so, when a full up test inevitably results in the destruction of the vehicle (because it is expendable) and tests cost as much as a launch (tens of millions of dollars) it makes it very, very difficult to evolve the design of the vehicle extensively.
That leads to a catch-22, you have to run a lot of launches to make the vehicle design significantly cheaper to manufacture. Buuuut, now you've vastly increased the development cost of the vehicle so you've erased all of the cost gains you've made.
But, if you design for reusability then you can actually increase reliability because the cost of the vehicle is amortized over multiple flights, so you can have a more expensive vehicle.
Also, while a lot of aerospace components (like fuel tanks, electronics, etc.) can be mass produced with the right design, this is a lot more difficult with rocket engines which have a very high number of precision machined parts made out of special alloys. This makes the engines the long-pole in the costs of a rocket, and it's very difficult to reduce those costs. Right now SpaceX is already the world's largest manufacturer of high-power rocket engines, so if there was a way to cut costs on them they would already be on top of them.
Now, back to reusability. Another thing that you get from reusability is that testing can be much cheaper, since you can make incremental changes to a vehicle design and then re-test it.
With reusability you have the best of all 3 worlds. You have reliability, you have low per-flight costs, and you have an enhanced ability to prove out potential vehicle redesigns economically.
Now, if the size of the launch market were much larger and reusability were more difficult then the equation could change, because it might be easier to recoup expensive cost-saving efforts over a shorter period of time.
By the time the boosters were recovered disassembled, had every inch checked for damage (visible and xray), and rebuilt, the cost was more than if they'd just left off the parachutes and built a new pair for every launch.
In theory if you landed the first stage of a liquid fueled rocket back near the launch site you could just gas it up and launch again, since the stresses are so much lower. I doubt that's what will happen, though - they'll give it a pretty thorough (and expensive) going over the way NASA did with the SSME.
There are two ways this could make sense from a cost perspective. One is after a confidence-building period they just do quick inspections and live with a slightly higher failure rate on subsequent launches. This would mean the second and subsequent missions are unmanned only, and it would have to bring the cost down so much that satellite makers could afford the risk.
The second is they just do relatively quick and cheap inspections and the increased failure rate is so small it's balanced out by the decrease in failures due to manufacturing error. This is really the holy grail of the rocket business, since it means you start to operate on the same economics the airlines do, where fuel is the cost driver. In rocket launches fuel typically comes to about 2% of the cost.
Putting a kg into LEO takes about the same amount of energy as flying it from LA to Sydney on a 747. The reason sending that kg to Sydney is so much cheaper is they turn the plane around and fly somewhere else instead of throwing it away.
While a controlled, retrorocket touchdown is obviously s huge engineering challenge and greatly complexifies the vehicle and increases its launch weight, the possible cost savings apparently make it a desirable goal.
Also, I believe SpaceX hopes to perfect the technology as part of their strategy to be make their tickets to Mars "round trip" :)
I'd mute it except that I want to keep the thing open in another browser tab so I can switch to it when the show starts.
Lesson learned: don't use repetitive audio crap while waiting for a web cast to start. Just have silence, then start talking when it's actually starting.