Slashdot Mirror
Inside the Tech of SpaceX's Homegrown Rocket Engine
An anonymous reader writes with this excerpt from a look at the engine behind SpaceX's Falcon rocket, the Merlin: "The rockstar of SpaceX may be Elon Musk, but the lead man behind the fire power is Tom Mueller. He is the Vice President of Propulsion Development and founding employee at SpaceX. Musk sought Mueller out in 2001 when Musk decided to build his own rockets instead of buying some from the Russians. Musk caught wind of a rocket engine Mueller built in his garage and 'apparently had a religious experience' once he saw it. If you didn't know, Elon Musk used $100 million of his Paypal money to start SpaceX. That money was used to build the Merlin engine Mueller had designed. The Merlin engine is the first new American booster engine in ten years and only the second in the last 25 years."
One of these days...
I'll meet you at the intersection of "Should be" and "Reality"
And not with a paid team working on a pay check.
You are talking bollocks.
There's nothing "cheap" about antimatter
Nuclear propulsion has this nasty side effect of fallout opposite the direction of travel. That's ok for space outside of Earth orbit, but I wouldn't use it at launch.
The highest density of accessible natural antimatter in the solar system is Earth orbit. One source of generation is thunderstorms. We need to have a way to store antimatter in quantity longer than 16 minutes for antimatter to be useful.
As far as SpaceX goes, I'm looking forward to seeing their BFR, Big F-kin' Rocket, a.k.a. Merlin 2. Something has to get us off this rock and a space elevator isn't going to be built for a long time. If they publish the date of their first launch, I'd like to see it in person.
There's nothing "cheap" about antimatter.
Maybe for you filthy savages that are still trapped on planet Dirt.
You've got a 3.8 x 10^26 Watt fusion reactor, why don't you fucking use it?
the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff
No religious experience here (then again, never seen in person) but everything I've read is the Merlin series is all about Chapman's "simplify and add lightness" which a lot of the old time aerospace pioneers used to use before they became profit munching incumbent contractors.
Pintle injector for throttling, stability, and some wall cooling. Damn good idea.
Don't wanna run a completely isolated hydraulic system and include a zillion new single points of failure? Hmm how bout using the fuel as the hyd fluid. How bout pressurize the hydraulic "fluid" using the main turbopump. Damn good idea.
The vacuum model uses radiative cooling. I'm sure a fat cat modern contractor would try for regenerative just to boost the contract cost / profit, but they're the "simplify and add lightness" people so simple radiative. Hardly a new idea for vacuum nozzle cooling, but a damn good one anyway.
They also show great judgment in knowing their own limitations, they buy their turbopumps from a specialist. Things that need to be custom they do, things that can be COTS are COTS.
I hope they can stay on task with the whole "simplify and add lightness" thing. The X and XX sound a little more like something you'd see from the incumbents rather than startups. Unless they have secrets up their sleeves, which is certainly possible.
Maybe the standard /. car example is the Merlin is as minimal as can possibly be made that'll work, like a 60s muscle car engine or a race car engine, whereas the incumbents are more like a modern engine which is mostly an elaborate emissions control system, oh and with an engine bolted onto it almost as an afterthought.
"Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
Controlled fusion power has not really been practical. There is certainly some program with project like ITER where a whole bunch of money is being poured into that kind of research (a total waste from my viewpoint, but some physics research may happen in spite of all of that money dumped down that rat hole).
Some other much more promising approaches include the Polywell and Focus Fusion concepts that seem to have some real theoretical potential, but none of these devices have been able to work out all of the engineering issues in terms of getting them to be producing usable energy of any kind. About the only real approach that might work and has been at least proven in terms of engineering is the original Project Orion concept. Unfortunately that uses thermonuclear warheads and is something only very large spacecraft would ever use. If you don't mind having nuclear weapons as a propulsion system, I suppose it could work.
I don't know much about the system you are suggesting here, which I think is in a similar shape to other nuclear fusion power devices of any kind. A nice theory and perhaps a different approach that could be useful. Nuclear rocket engines of any kind (fission or fusion) have the potential of a very high specific impulse (the amount of thrust they can produce given a certain amount of mass for fuel + engine) and in the long run I think most interplanetary spacecraft will be using nuclear engines of some sort or another. Chemical energy is just too inefficient to be practical. The raw physics for nuclear propulsion has been more or less worked out, but coming up with a practical design that actually works is where the real problem lies.
What the hell reflects gamma-rays anyway? Neutronium?
Also, any news on a gamma radiation reflector, a possible prerequisite to a propulsion with gamma rays from "cheaper" antimatter?
Prereq to a lot of weird stuff, including certain .mil toys you probably won't like very much. Probably would make an interesting hard sci fi novel.
"Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
Thanks a lot, that's what I was asking for. Hopefully, SpaceX will earn enough money from the tourist billionaires, or be it asteroid mining, to look for the nuclear propulsion as well.
"Merlin" is an engine brand of Rolls-Royce, a V12 piston engine from the 30's onwards used in a wide variety of aircraft. I can imagine raised eyebrows in their offices, but would they actually sue? I hope not, that would show these lawsuit-happy Yanks what British class really is.
I'm surprised you're so down on tokamak research which has actually produced large quantities of energy in tests (22MJ thermal in 1.5 seconds from the JET run in 1997, for example) while describing the con-artists like Polywell and Focus (zero joules in several years of funding and self-promotion) as "promising". At least you're not carrying a (fusion) torch for Fleischman and Pons.
Says the physics PhD.
And here's hoping that ITER will shut up these tedious cranks before 2020...
Musk sold Paypal back in 2002 and if I recall it was rather decent then or at least I never had any problems.
(22MJ thermal in 1.5 seconds from the JET run in 1997, for example)
22MJ/15 years gives an average power output of only 46 milliwatts.
Not a very impressive power station.
SpaceX is supposed have about 10-20% ex-NASA people, mainly younger folk. Dont think of NASA as automatically bloated - it was the only game for aspiring rocket scientists before the 2000s. Recycling NASA people preserves some of their experience. One of the problems with the Orion program is that a lot of the good Apollo ideas had been lost due to retirement of those engineers and loss of record.
Goddard wanted to build spinning engines which used the rotary pressure to increase thrust-to-fuel ratio; visible in his posthumous patents.
Some basic info at (follow the links):
http://www.halfwaytoanywhere.com/
I suggest you to look up TRW and the Low Cost Pintle Engine (LCPE) on the internet. Guess who was head of liquid rocket propulsion development there back at the start of the century.....
Well fusion is easier for electricity since it does not really matter if it masses a 1000 tons. So since we can't do the later we can't do the former.
Antimatter does not need gamma ray reflectors. When a proton and and a antiproton combine you get Pions. Not gamma rays. About 1/3 is each of the pi+,pi- and pi0. The pi0 does almost immediately decay into gammas and there goes about 1/3 of the energy. But the pi+ and pi- live for long enough and are moving pretty close to the speed of light. Since they are charged they can be directed by a magnetic nozzle. So we can even without much magic (other than generating and storing antimatter) have a antimatter engine with about 50% or even fairly close to 66% efficiency.
Even with 50% efficiency you can get away with pretty reasonable mass ratios for start stop upwards of 50% the speed of light.
However the only gamma ray reflector we could do a a dense block of Tungsten that is heated to close to its melting point by all rearward directed gammas. And a mirror system to send the radiated energy (from the white hot W block) backwards. It would need a rather thick block of Tungsten... so your probably going to be slower than not having it. An interesting calculation to work out for the evening perhaps.
If information wants to be free, why does my internet connection cost so much?
It's a scientific instrument, not a power station. The instrument (JET, amongst others), has done its job, and they're now building the proof-of-concept power station (ITER). Nonetheless, if you know anything at all about magnetic fusion and JET, you'll realize that what it's done (16MW for 1 second -- and more importantly -- 4MW for 4 seconds) are actually massive achievements, especially for 30-year-old hardware.
What a con man Robert Bussard was! He only spent most of his life working for NASA and the Atomic Energy Commission. Clearly he didn't know what he was doing, and only tried to make huge amounts of money from the government. Which is why he died broke and still looking for funding for his project. Weird how even after he died, the US navy picked the project up and is still working on it - with regular reviews by independent physicists.
The Polywell has taken something like 20M USD in its entire history, where ITER is supposed to cost about 20 billion. Funny how no Tokamak has managed to sustain ignition, or survive it's own radiation flux.
Polywell might not work, but Tokamaks don't, either.
I didn't know about this... I guess the Falcon Heavy is still not enough for a manned Mars mission so they're gonna build the Big Fucking Rocket.
Glad I actually RTFA
Oh, one other comment about the JET test. It produced a pretty high peak output power, yes. But what was the percentage of input power? According to Wikipedia, it was about 65% of input. The reaction was nowhere near ignition.
That's what every ray-proplled ship could be constructed from!
Since they both end in "um" can we assume neutronium weighs about as much as aluminum?
What the hell reflects gamma-rays anyway? Neutronium?
A black hole's event horizon, from the inside? Kind of impractical for a nozzle, though.
Ezekiel 23:20
It's a scientific instrument, not a power station.
That's what I was pointing out, for the benefit of those among us who might think that 22MJ somehow indicates any kind of significant progress towards an economically viable fusion power station.
Intelligent folks often do crazy things outside their own specialist fields. Newton, the author of the Principia among other great works was an alchemist in his spare time, "studying" strange Kabbalistic nonsense. Linus Pauling went all weird in his later days too, the list goes on and on.
Just because Bussard was a smart guy in one field doesn't mean Polywell or focus fusion will work; the "and then a miracle occurs" stage is a big hurdle to pass. I class it in the "carburettor that runs on water" family of ideas that a few overenthusiastic folks think will work with just another little tweak here or an adjustment there along with some out-and-out con men riding the coattails. I will be condescending here and assume you're a True Believer rather than the other alternative.
When you say no tokamak has "sustained ignition" I assume you think JET's 1.5 seconds at 16 megawatts thermal output is not sustained or not ignition, I don't know which. JET and other tokamaks were never designed for sustained running or even to reach a Q>1 (more energy out than in) return, they are fusion testbeds to develop methods of creating, focussing, heating and maintaining plasmas. JET was rebuilt yet again recently to support the ITER project, concentrating on construction and materials. ITER is meant to run to Q>10 for tens of seconds and more but it's a materials and operations testbed too, more engineering than physics though and intended to narrow the design parameters of the first prototype power reactors.
Bussard was a nuclear physicist. This was not outside his field, it WAS his field. He worked on Tokamaks long before he began his adventures into electrostatic (electrodynamic? polywell is weird) fusion.
I don't 'believe' that Polywell will work. I think it's worth investigating, and the US navy is doing so. I am satisfied. They will decide if it works or not, and I will accept their conclusion. I am NOT a nuclear physicist, and do not pretend to be one.
But I have read enough, from Bussard and others, to doubt Tokamaks. They have not yet reached ignition, producing more heat than it took to start the reaction - they cannot run on their own. ITER is _supposed_ to do so, but then they expect it to fail in short order because (as you just said) the materials science is not good enough to resist the effects of radiation bombardment on the structure. It's insanely expensive, and has been '20 years away' for longer than I've been alive.
Do they make a pion-pion sound?
Ignition != breakeven.
Unobtainium, obviously.
I got a malware alert from the office trying TFA's link.
Do not mock my vision of impractical footwear
"Dabbled in Perpetual Motion a bit? Almost got it a couple of times?"
-- Professor Harold Hill, "The Music Man"
Do not mock my vision of impractical footwear
Prereq to a lot of weird stuff, including certain .mil toys you probably won't like very much. Probably would make an interesting hard sci fi novel.
I suspect you're referring to "Footfall" by Larry Niven & Jerry Pournelle.
Do not mock my vision of impractical footwear
What would a magnetic nozzle look like?
Ahh, nvm -- GIYF.
Do not mock my vision of impractical footwear
http://en.wikipedia.org/wiki/DEMO
I might not survive the extra apostrophe flux you generate.
All of that energy after pouring how many billion dollars into the project? It sounds real impressive when you are talking megajoules, but that figure is only 6 kilowatt-hours if you turned it into more familiar power units that ordinary people use every day.
I would also say that Polywell and Focus have produced some energy... and in terms of a dollar per dollar rate of return it is pretty much equivalent. Fusion energy research is really going nowhere fast regardless of what technology you are talking about, and seems to be hitting dead ends all of the time. That could be said about the "cold fusion" work of Pons and Fleishman, which seems to be a lousy neutron emitter at best (and the Farnsworth-Hirsch fusor seems to do a much better job of accomplishing I might add).
I just see the Tokamak research as yet another dead end. Perhaps something could be rescued from all of the research that has been accomplished with the billions of dollars dumped down that rat hole. I'm not hoping for too much other than a bunch of PhDs to be generated by all of that money, and I think it suffers the same problem that almost all government funded research seems to face: There is no real desire to actually build something practical, and the funding is an end unto itself. If you think the other forms of fusion research are all confidence games, why are you so supportive of Tokamak research?
They're not breaking any records or anything but not bad. The MerlinC engine is around 300 Isp (specific impulse, the engine efficiency for those who don't know). That's not blowing away the Space shuttles specs (~400 Isp) but it also doesn't cost $40 Million per engine or use Liquid Hydrogen.
Tokamaks have produced fusion energy -- the single run on JET produced a significant amount of energy (22MJ) for a macroscopic period (over a second) with a Q (energy return) of 0.65, and that's with a tokamak that was never designed to produce high Qs in the first place -- JET was and still is a research tool, limited to short bursts of plasma heating and magnetic confinement.
What's the best Q-factor result from Polywell fusion at the moment? I can't easily find figures on numbers from experimental hardware on the Internet.
The folks designing ITER expect it to reach a Q of 10 or so, 50MW heating energy in resulting in 500MW (thermal) fusion energy and that sustainable for hundreds of seconds. Plasma stability, the bugbear of confinement systems is better understood than it was thirty years ago when JET and the other big tokamaks were first built. There will be other problems but that's why ITER is experimental -- a prototype power reactor wouldn't be very much bigger than ITER or so folks think and there was some temptation to build it to that scale.
If Polywell or Focus can demonstrate long-duration high-Q fusion and energy recovery before ITER's first light (expected in 2019 or so) then it will probably become a white elephant. I wouldn't hold my breath though.
More like Pie pie
If information wants to be free, why does my internet connection cost so much?
Case in point, there was a recent paper that showed (theoretically) that a 90+% magnetic nozzle should be possible.
If information wants to be free, why does my internet connection cost so much?
You know a experimental power station that cost about 10-20B over the 20-30 year lifetime is in fact not that expensive at all. Note that a plain old nuke plant cost about 10B to build. And a cool 1B even for a 1GW coal plant. Hydro cost even more.
Did you also know that plasma containment is getting better *faster* than Moore's law? In the last 20 years there has been a increase in containment of something like well over a million (can't be bothered looking it up). Also did you know that the ITER crowed have *never* been given the budget they expected. So the joke of 20 years always is strongly based on the fact they have not had the finding they said they would need for that to be true. Also even new fission is going to take 20-30 years to validate, before wide scale deployment.
We are in this energy thing for the long haul. 20 years even 60 years of sustained investment is the kind of long term planning we need to be doing. Sure we should have a more diverse portfolio, even the ITER guys have at times strongly supported that (other times they have been told to keep their mouths shut). But the whole "my fusion idea is better and cheaper but the baaad ITER people are taking all the money" is false. A bunch of scientist killed the SCC years back because they thought the money not spent on that particle accelerator would be available to them. It wasn't. They got no increase and set back particle physics 20 years. Fact is these R&D items are tiny line items in government funding programs, its not a zero sum game. kill one and there is *more* reason to kill another. Not less. ITER is not stealing anyone's funding. Kill it and that money goes away.
Finally the Buzzard Polywell really tries to pretend a whole bunch of both theory and experimental results are wrong. The facts are that the probability of a T nuclei and a D nucli undergoing fusion is >1000x more than the chance they scatter off each other. Thus they *always* thermalize faster than they fuse. Add electrons, and again energy sapping ion electron collisions are 1000x more likely giving there energy up as x rays. For any other reaction those numbers are far far worse.
Yes i am a nuclear scientist. Or at least i was.
If information wants to be free, why does my internet connection cost so much?
That's fine. I didn't say 'defund ITER,' the materials science and plasma research IS valuable. What you said was 'defund Polywell and focus fusion' when they are taking a tiny portion of the funding pie. And like you said, if you kill those, it's not like ITER is going to see its budget increase.
I would appreciate it if you would focus on arguments instead of insulting the dead guy. Buzzard? Really?
I have read a bit about those experimental and theoretical results. Bussard & co seemd to think that they did not apply. Different plasma types and such. I know enough about the theoretical results to say they could easily be wrong - didn't you just say that plasma physics is learning new things all the time? The experimental results concern me - but again, the Polywell research is constantly being reviewed by independent physicists, and if it looks like it's not panning out, it will be canned. It's been running several years now, and still going. So... maybe the experimental results really didn't apply?
As an aside: you want AWESOME ridiculous fusion? Look up these guys: http://en.wikipedia.org/wiki/General_Fusion :P
I don't think it will work, but I WANT it to work
I did not insult Bussard. I pointed out that both experiments and Theory is against the idea, if you read that as an insult then science is not for you. Being dead changes nothing. Also i said nothing about Focus fusion. Point in fact Eric has addressed the specific criticisms that I pointed out. While Bussard just ignored them, or claims can't just not apply, but then never says why they wouldn't when they clearly do apply. Nor did i say defund anything (though *I* wouldn't throw a penny at the polywell). I said a diverse portfolio is a good idea. That is fund many approaches. Polywell has had very little independent reviews and almost nothing public or peer reviewed, this is in strong contrast with the focus fusion team. I do keep up with the literature. Or at least try too.
Legitimate dark horses in fusion (ie don't require magic): General Fusion and other pulse compressed magnetized concepts. DPF (aka focus fusion, but then people have been working on this for 30 years too!), Colliding RFC. Note none of them have shown operating parameters even close to either Tokamaks or ICF. The mainstream is still closer on every metric that matters.
Note that if plasma's did what we expected we would have had fusion 60 years ago. All these methods could easily come undone as so many previous attempts have or come across problems that take time to fix like main stream concepts. Its a lot more tricky than people outside the field understand. But we have come a very very long way and have a much better understanding of how it works.
Another interesting fact about Tokamaks. When the Russians published their results in the west, no one believed them because it was just so much better than anything else (something like 100x better!). It wasn't until a team of scientists went over to Russia to independently verify their claims that folks on this side of the iron curtain believed them.
If information wants to be free, why does my internet connection cost so much?
You called him Buzzard. That's a heck of a typo if you were not trying to be insulting. Saying it's against the theory and experiment is fine, and I gave you my reply to that.
Polywell has had several reviews, though we don't know the specifics. If I remember correctly, the project is only funded on a year-to-year basis, so the navy must think they're on to _something_. I wish that the results were public! Then we'd actually have something to argue over.
I'm a little bit surprised that you consider General Fusion to be even a Dark Horse. I cannot say anything about the fusion reaction, but the mechanical engineering would be a true nightmare.
Quoting the JT-60 wiki page:
During deuterium (D–D fuel) plasma experiments in 1998 plasma conditions were achieved which would, if the D–D fuel were replaced with a 1:1 mix of deuterium and tritium (D–T fuel), have exceeded break-even—the point where the power produced by the fusion reactions equals the power supplied to operate the machine. JT-60 does not have the facilities to handle tritium; currently only the JET tokamak in the United Kingdom has such facilities. In fusion terminology JT-60 achieved conditions which in D–T would have provided Q = 1.25, where Q is the ratio of fusion power to input power. A self-sustaining nuclear fusion reaction would need a value of Q that is greater than 5.
So i am bad at spelling or remembering names and make lots of typos. Sue me. The Navy has also funded cold fusion and anti gravity. It really is *not* a feather in the hat of legitimacy.
If information wants to be free, why does my internet connection cost so much?
1) Watch your mouth.
2) JET -- scientific development/feasibility (done). ITER -- engineering development/feasibility with some added basic science to nail remaining issues. Power plant scale although will not feed power to the grid. DEMO -- prototype commercial power station.
Don't be a fucking idiot.