Slashdot Mirror
Elon Musk Explains Why SpaceX Prefers Clusters of Small Engines (arstechnica.com)
An anonymous reader quotes a report from Ars Technica: The company's development of the Falcon 9 rocket, with nine engines, had given Musk confidence that SpaceX could scale up to 27 engines in flight, and he believed this was a better overall solution for the thrust needed to escape Earth's gravity. To explain why, the former computer scientist used a computer metaphor. "It's sort of like the way modern computer systems are set up," Musk said. "With Google or Amazon they have large numbers of small computers, such that if one of the computers goes down it doesn't really affect your use of Google or Amazon. That's different from the old model of the mainframe approach, when you have one big mainframe and if it goes down, the whole system goes down."
For computers, Musk said, using large numbers of small computers ends up being a more efficient, smarter, and faster approach than using a few larger, more powerful computers. So it was with rocket engines. "It's better to use a large number of small engines," Musk said. With the Falcon Heavy rocket, he added, up to half a dozen engines could fail and the rocket would still make it to orbit. The flight of the Falcon Heavy likely bodes well for SpaceX's next rocket, the much larger Big Falcon Rocket (or BFR), now being designed at the company's Hawthorne, California-based headquarters. This booster will use 31 engines, four more than the Falcon Heavy. But it will also use larger, more powerful engines. The proposed Raptor engine has 380,000 pounds of thrust at sea level, compared to 190,000 pounds of thrust for the Merlin 1-D engine.
For computers, Musk said, using large numbers of small computers ends up being a more efficient, smarter, and faster approach than using a few larger, more powerful computers. So it was with rocket engines. "It's better to use a large number of small engines," Musk said. With the Falcon Heavy rocket, he added, up to half a dozen engines could fail and the rocket would still make it to orbit. The flight of the Falcon Heavy likely bodes well for SpaceX's next rocket, the much larger Big Falcon Rocket (or BFR), now being designed at the company's Hawthorne, California-based headquarters. This booster will use 31 engines, four more than the Falcon Heavy. But it will also use larger, more powerful engines. The proposed Raptor engine has 380,000 pounds of thrust at sea level, compared to 190,000 pounds of thrust for the Merlin 1-D engine.
Redundancy is always good
Seems like a good idea to me but I'm no rocket scientist.
I thought the F in BFR stood for something else than Falcon...?
It's how you use it.
up to half a dozen engines could fail and the rocket would still make it to orbit
Not if the fail catastrophically. If one blows up you've had it. This is probably more likely than a computer failing and burning down your data centre, so a factor worth considering,
Obviously SpaceX has calculated this but Id like to see a graph of the probability of flight failure of a rocket with 5 big engines and a rocket of 31 small engines. The more engines the higher the chance one will not work but also the higher the redundancy. The fewer engines the less chance one will not work but also the greater the chance one going out dooms the flight.
I came to the datacenter drunk with a fake ID, don't you want to be just like me?
Do you have to shut down computers opposite to the one that fails to maintain "computing balance"? Can a computer that fails blow up and take a few adjacent computers with him? Can a failing computer cause a cascading effect by sending bogus signals through the network that makes other computers fail? Does a failing computer fundamentally alter your mission profile to the point that you have to change the computations for ALL other computers?
Maybe we should stay with car analogies. They aren't any better, but at least we're used to them being garbage.
We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.
I read the headline and said to myself, "Because there's not a single point of failure..." *pats self on back* (gotta do that sometimes...)
"That's different from the old model of the mainframe approach, when you have one big mainframe and if it goes down, the whole system goes down." Except that mainframe doesn't and AWS does.
I think you raise the risk of failures by adding more engines.
It's basically for the same reasons as tesla batteries are made from thousands of the small lithium cells that are already mature.
The next rocket will be the 'Big Falcon Rocket'.... sure... 'Falcon'... ghehehehe. The BFR 3000 ;-)
It would be interesting to know why this engine redundancy wasn't leveraged to save the center core of the Falcon Heavy when it attempted to land on the drone ship. They claim two of three engines failed to fire. If so, why wasn't the system programmed to automatically try to fire two alternate engines in that failure mode? Unless the failures where of a more catastrophic nature of course...
If nothing else, this shows Elon knows nothing about mainframe computers.
The real reason they are taking this approach is larger engines are ruinously expensive and remain fundamentally outside the reach of private companies and the domain of nation states. The materials science research required to replace those 31 engines with say 3 or 5 engines would run up a bill pushing $100 billion. The Chinese and the Russians will remain in the game because at the end of the day that's exactly what they will do while SpaceX sits back and watches. With smaller engines you pay a substantial weigh penalty for the buttressing and gimballing that goes with each engine. All else equal a design with few engine is more efficient. And the redundancy argument is frankly nonsensical because the probability of failure goes up with the number of parts and smaller engines = more parts = a greater chance of your QC missing defects = a greater chance of BOOM. This is why the US always went with fewer but larger engines and the cash strapped Soviets went with smaller but more numerous engines. It was wholly a question of available funds.
That's a terrible analogy, because mainframes pretty much never went down.
Years and years of uptime, components which could be swapped out of a running system ... who needs redundancy when you have reliability?
I kid of course, the mainframe wasn't perfect. But I've encountered systems which have been continuously running for staggering amounts of time.
As a UNIX guy, long uptimes were the norm. You fucking Windows pansies with your monthly reboots or rebooting just to make whatever stupid problem you're having go away simply don't grasp that machines being up for months if not years was a pretty standard thing for a very long time. If you had to reboot more than once a year you had a pretty broken system.
Now due to low quality software, people don't expect uptime to be more than a few days -- heck, I know IT departments which expect their users to boot daily, which is stupid. Who has time to waste on that shit? Like I have 20 minutes a day to reboot and another 25 or so to re-open my applications.
What were we talking about? Oh yeah, rockets .. rockets are awesome.
Now get of my damned lawn.
This preference for engine clusters is like using a stuff-ton of laptop cells instead of a much smaller number of automotive cells in the Tesla battery pack?
Say what you want about Musk, SpaceX, etc
I love the fact that they include "pricing" on the SpaceX website, like your just buying a refrigerator...
Falcon Heavy is only 90 million to GTO btw :)
1) more engines is more parts, which is generally a recipe for more failures
2) more engines probably means more weight which means less total lifting capacity per unit of thrust. Even something as simple as a tank, a single larger volume tank be less material than say two tanks each of half size volume.
--
To use a computer analogy; suppose you have two disks.
raid 0 or raid linear-> maximum chance of catastrophic failure
Put one disk in a drawer and use the other -> safer than raid 0
raid 1 -> redundancy least risk of catastrophic failure.
--
So for what he is saying to be true. There must be minimal risk an engine failure affects other modules. If it detonates for example, does it damage the adjacent modules? The flights controls have to be capable of reliably altering plan to deal with a failure module; burn other thrusters longer, change attack angle, whatever. Finally you have to always leave enough extra capacity to tolerate some number of failures.
Repeal the 17th Amendment TODAY! Also Please Read http://www.gnu.org/philosophy/right-to-read.html
Even the Russians used multiple smaller engines in their space program. And if I'm not mistaken the Saturn V used five of them.
I always heard it was the same basic acronym as BFG the gun from Doom.
Calvin:Do you believe in the devil? Hobbes:I'm not sure man needs the help.
No! There are two extreme cases. One, where the probability of total system failure is the sum of the probabilities of failure of components (bad), the other where it is the product (good) 1% chance of failure + 1% chance of failure = 2% chance of total failure
You can't sum chance of failure like that. That's not how the math of it works. (think about it - if you take that to it's logical conclusion with 200 failure modes each at 1% chance of failure you can end up with a >100% chance of failure which isn't possible) First you have to determine whether the failure modes are genuinely independent or not. But even if you have two completely independent failure modes with a 1% chance of Failure A and a 1% chance of Failure B, the total chance of Failure is NOT F(A)+F(B) = 2% because there is a probability of both failure occurring simultaneously so the real probability will be less than 2%.
1% chance of failure * 1% chance of failure = %0.01 chance of total failure
It doesn't work like that either unless those failure modes are such that both have to occur for a failure to occur.
That's different from the old model of the mainframe approach, when you have one big mainframe and if it goes down, the whole system goes down
In the case of the mainframe the redundancy is build in. You don't have to use 100's of mainframes because they almost never go down. I've been working on mainframes for the last 20 years, and I can count on one hand the number of times the mainframe was down in a production environment in that period.
Try it! Library of Babel
I don't think this article is particularly newsworthy for anyone who's familiar with the subject or even stopped for a few seconds to think about it.
But here are a few points why multiple smaller engines are better in this case.
* Mass production makes things cheaper and sometimes better.
If you look at the cost of an engine, the raw materials are a pretty small percentage of the total cost, It's more about the manpower and tolerances required.
If you have a guy performing a certain task maybe once every two months, he'll be slower and less proficient at doing it than say every few days.
And overall, economies of scale make more engines cheaper to manufacture.
* Redundancy, as mentioned already in the thread.
SpaceX has already lost one of the engines on one of the earlier flights and continued to complete the mission.
They have walls between them that prevent the explosion of one from damaging the next.
And it's only going to improve for Falcon Heavy and BFR, they'd be able to lost multiple engines and compensate for the imbalance.
* Telemetry collection.
You get to build up a history of past performance a lot faster with ten engines than you do with one or two.
After 54 flights, SpaceX has gathered operational telemetry on 486 first stage engine firings and 54 second stage ones (Not including all the test firings).
* Throttling, maneuverability, unique thrust characteristics.
In the early stages of landing R&D SpaceX had a rocket called the Grasshopper, which was modified Falcon 9, that was able to fly up, hover, and then land.
Most larger engines would not be able to throttle low enough to maintain a hover.
This one is just a guess, but I imagine it's much easier and faster to gimbal a smaller engine, and you don't have to put the smaller, cheaper gimbaling hardware on all the engines.
It seems that the exhaust from the multiple engines behaves somewhat similarly to an Aerospike engine, which gives the configuration some extra efficiency.
That's all I could think of, but I'm sure there are more reasons.
1) more engines is more parts, which is generally a recipe for more failures
Only if you hold everything else in the system constant which is clearly not the case in most real world systems. To use a car example, modern cars have a LOT more parts in them than cars from 40 years ago but they also are demonstrably more reliable. Same with jet engines. Modern ones are more complex and with (usually) more parts but they also are more reliable. The relationship between number of parts and reliability is not a simple linear one. Many of those added parts actually contribute to the reliability of the overall system.
The most extreme example of this sort of thing ever attempted was OTRAG.
John Carmack had some interesting things to say about that at his now-defunct Armadillo Aerospace website, some of which have been preserved at Wikipedia here.
Ceci n'est pas une signature.
If nothing else, this shows Elon knows nothing about mainframe computers.
I think that comment says more about you than it does about Elon. Do you seriously think Elon doesn't get that it's an imperfect analogy used to make a rhetorical point?
No, ULA uses Russian RD-180 and RS-68 on their rockets. SpaceX's Merlin is a newly designed engine that is similar to some that NASA have used in the past.
I'm a good cook. I'm a fantastic eater. - Steven Brust
free musik mp3 songs http://www.metrolagu.info/
Wasn't this the design approach of Armadillo?
The classic IBM mainframe had super reliability, which included redundancy.
-cranky old guy
I doubt the explanation.
I suspect that the real problem is that a really big engine is way hard to design. It is more reliable to re-use proven designs, even though complexity is greater, and the mass-to-thrust ratio is probably worse. At some point, that tradeoff doesn't work any more, and you have to design a larger engine, or end up with a far costlier vehicle.
All this debate from Slashdot rocket scientists over whether Musk is properly designing his rockets, whether he "understands" the finer points of his (imperfect!) metaphor, whether he really understands computers at all, and yet...
he launches rockets (and lands them) again and again and again and again.
How about a moderation of -1 pedantic.
can't tell if you're making a joke and I'm just to aspie to get it.
If an Australian is saying Falcon, to an American ear it would sound like they're saying Fuckin'
Bob Truax argued that the most cost-effective way to build a big rocket was using one huge engine, which is what was planned for the Sea Dragon booster. He said cost is driven not by size but by parts count. More engines equals more parts that have to be produced, inventoried, tested, assembled, etc., and that leads to higher cost.
However. . . The way SpaceX are returning their boosters to earth wouldn't work with one huge engine. There would be no feasible way to throttle it down enough for the return and landing burns! Each Falcon 9 core typically lights up just one engine for the landing.
yet still no "bag o' glass"...
the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff
He announced the next Bigger Falcon Heavy lunch vehicle will be made entirely of the rocket engines sold to high school science projects. 3.2 million of these rocket motors glued together will form the launch vehicle.
He said, "We know Elon got his start by building an electric car by duct taping 8000 laptop batteries together. Same thing here, back to the basics. man!"
sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
It took the US and the USSR years to solve instability problems with engines larger than the V2 rockets.
Also, why would anyone trust the opinion of a software developer with regard to hardware development?
The flight of the Falcon Heavy likely bodes well for SpaceX's next rocket, the much larger Big Falcon Rocket (or BFR), now being designed at the company's Hawthorne, California-based headquarters. This booster will use 31 engines, four more than the Falcon Heavy. But it will also use larger, more powerful engines. The proposed Raptor engine has 380,000 pounds of thrust at sea level, compared to 190,000 pounds of thrust for the Merlin 1-D engine.
Does anyone know what this means relative to the lift capacity of this new rocket they're working on? The Falcon Heavy was already a huge leap over the competition and this doubles the thrust with a few more engines (understanding that some of that thrust is going to come at the cost of carrying additional fuel too).
Weight of Rocketdyne F-1 engine on Saturn V (moon rocket): 18,000 pounds, thrust:1,500,000 pounds
Weight of Merlin 1-D engine on Falcon 9: 1,000 pounds, thrust: 190,000 pounds
The specific impulse of a Merlin engine is 282 seconds, the specific impulse on an F-1 engine is 263 seconds.
TL;DR:
18 Merlin engines weigh the same as one F-1 moon rocket engine but only 8 of them are needed to provide the same thrust.
There is a standard method of calculating this sort of safety case in an industrial setting - Safety Integrity Level (SIL) reference IEC 61508 and 61511 standards. Aerospace uses a similar method.
The idea is that the following are combined to determine actual risk of failure
Duration of exposure
Risk of failure (per time interval)
Number of things at risk.
In the case of spacex it is possible that for a short period near the pad, there is only a small reserve thrust margin, and a single engine failure could cause a mission loss with a max payload. Once one engines thrust equivalent fuel quantity is burned you now have N+1 redundancy of engines. Farther into the flight as fuel is burned even more margin is available, providing all of the control issues that arise from loss of engines can be managed.
The Soviet space program needed an equivalent heavy lift rocket and began development on the multiple engined N1 rocket - v. https://en.wikipedia.org/wiki/N1_(rocket) Its 1st stage sported 30 engines arranged in two rings; the 2nd, a single ring of 8 engines. However, it suffered from multiple failures and was eventually canceled.
Not the same thing
1.2 volt batteries are put in series because
Battery management of cells in series is simpler that the same kWhr of batteries in paralell
The weight of cables to handle 500 hp at 1.2 volts will not fit in a car sized vehicle.
Actual voltage in a Prius is about 200 volts
Tesla model S is 375 volts
A "catastrophic" failure is almost certainly at the turbopump. A vane falling off and the unbalanced rotor tearing itself apart is pretty forseeable. And it's quite possible to prevent (or hugely reduce the chance of) this affecting other engines in a Super-Kamiokande style cascading failure.
Turbines are routinely designed to contain rotor failure (ever see the jet engine tests?), and although the engines are tightly packed, the spacing is set by the engine bells, which are much larger than the pumps. There's lots of room for shielding between the pumps.
For example, the main strength members of the "octaweb" sit between the turbopumps.
And guess what else? Falcon 9 first stages have lots of shielding at their bottom ends, specifically designed to prevent hot flaming gases damaging the engines. It's heat shielding for re-entry, but it also serves to contain violent impulses by the engines. If it's destroyed by an engine failure, the stage might not survive re-entry, but that's not a mission failure because re-entry isn't part of the contracted mission.
:{ is
Only this time with reliable technology and safety systems.
These engines are lighting off at the *front* of the rocket, in the sense that they're open end is in the direction of travel (down), correct? Naively I would think that the air blowing into them would tend to blow them out, like blowing air at a candle you're trying to light. I suppose when it's supersonic, there's a shock wave that prevents that; but the final burn is surely at subsonic speed. But maybe the engines are designed so that this doesn't happen, even subsonic?
"modern cars have a LOT more parts in them than cars from 40 years ago but they also are demonstrably more reliable": Interesting point. I'm sure the reason for this is known by somebody, but it isn't known by me. I always assume it has to do with more QC in the factory, and that this QC effort was started by the Japanese in the 70s or so, and only taken up by Detroit (and the European mfgs) because they were loosing market share so badly. Can you (or someone) enlighten me? Why are cars more reliable now?
PURRRGE.... :P
[($)]
Nobody builds rockets and says "pounds". Nobody.