144 points by spking 9 days ago
A similar design (the space shuttle main engine) brought me some peace of mind at work one day.
Nowadays, I do deep learning for a living, before that I was doing computer vision. One day I am really pissed off at some compiling issue, with a parallel code and I am like "fuck this, this is a plumbing issue! I should spend my time on algorithms and shit. Not on ordering the damn bytes of the raw pixels. I have done that for years, that should be a solved problem by now."
I go on reddit for some well-deserved outrage-slacking and stumble on the shuttle engine. "Yeah, that's what I am talking about. Now that is engineering." Then I realized it is basically a maze of pipes and pumps.
Uh, yeah, 95% of plumbing (and some metallurgy feats as well). Ok, back to coding my data pipelines then...
"Aerospace is plumbing with the volume turned up" -- John Carmack
I think somewhere in this sci.space archive of rocketry usenet wonders is a quote along the lines of "Rocket science is just extreme plumbing". Cryogenics with powerful oxidizers is hard!
"All engineering is plumbing - some fields just manage to avoid working on broken sinks & toilets"
Even the arts aren't necessarily safe. A story told by the composer Philip Glass in 2011:
> While working, I suddenly heard a noise and looked up to find Robert Hughes, the art critic of Time magazine, staring at me in disbelief. ‘But you’re Philip Glass! What are you doing here?’ It was obvious that I was installing his dishwasher and I told him I would soon be finished. ‘But you are an artist,’ he protested. I explained that I was an artist but that I was sometimes a plumber as well and that he should go away and let me finish.
My day job is plumbing. I really very badly want to be able to make a decent living through the skills I've been practicing outside of my job, and this quote is motivating to me. Thank you! :)
There's some serious chemistry, thermodynamics and fluid dynamics involved as well.
I suppose someone is thinking that fluid dynamics is plumbing, but it would be a stretch to make that claim over issues like the size and shape of the combustion chamber, its throat and the bell, which are dependent on the characteristics of transonic and supersonic flow.
I guess one glaring difference might be that they got this right when it mattered. Not a single Saturn V ever failed during launch. The aerospace engineering culture is vastly different than the "engineering" that goes on in ML/AI, and software development more generally. I have a lot of respect for it.
Saturn V suffered several engine failures. It had persistent pogo oscillations that put Apollo 6 into the wrong orbit and forced much of its test plan to be canceled, and nearly destroyed Apollo 13 during launch. Vibrations severely damaged Skylab during its launch. I would not say it never failed. It never went kaboom in the sky, but there are other ways to fail.
It was an exaggeration, but the comment's underlying point is spot on: the "engineering" that we do doesn't hold up in terms of rigor or cleverness that AE and other related engineering disciplines exhibit.
Counterpoint: if we tried to build all software using the same rigorous and "clever" processes that aerospace engineers do, almost nothing would ever get built. It's a different process because it's a different field, not because aerospace engineers are fundamentally more clever.
Signed, Not an aerospace engineer ;)
Have an aerospace engineer cow-orker; one favorite quote for why he works in enterprise software is, "Because this flies and that (pointing at an Ares-1 mockup) doesn't."
On the other hand, there's nothing "engineering" about software engineering.
Hahaha, harsh. I like your coworker.
And I would tend to agree with your second point. We might occasionally borrow some formalism and organizational structure from our engineering cousins, but it's different in enough ways that it probably shouldn't bear the same moniker.
While I generally agree with that underlying point, if our computers were as unreliable as the Saturn V, we’d never get anything done with them.
The cost of failure is higher in aerospace. But I must say that in my job, when you manage to make your modified code pass all the thousands of unit tests and see the benchmark results increase and the 80 cores of the test machine firing up, feeding GB/s of data in their pipeline at the optimal fillrate, there is something of of a "ignition aaaaaand liftoff!" moment that is deeply satisfying.
I am just allowed to try pressing the red button several times a day instead of once every two years.
80 modern processors have around four orders of magnitude more "moving" parts than a Saturn V, according to what I can Google. And the Saturn V, it is said, had a particularly large number of parts because it was hand-made to an extent that modern machinery is not.
Comparing the number of parts underestimates the complexity of a computer system, of course, because the software is a machine with a vast number of parts, on top of the hardware.
In other words, I'm on the side of those who say software is hard to get right for good reason, and not because the practitioners are less competent.
Here is 500 fps film of the Saturn V launch for Apollo 11, narrated by Mark Gray. It explains the sequence of events over the course of 8+ minutes (30 seconds real time).
Thanks for that, great video that led me to this, https://www.youtube.com/watch?v=ImoQqNyRL8Y, sound recording organised by Dustin from SmarterEveryDay of the latest SpaceX launch, the sonic booms of the boosters coming into land are amazing, listen with headphones!
Just as a Side-note: that video is from the Falcon Heavy demo launch last year.
The recent launch was the first commercial Falcon Heavy launch (Arabsat 6a)
Cool. Shame they couldn't have a microphone by the landing site so the sound comes at the same time as the landing - maybe on a future one.
A portion of flown engine #5 from the Apollo 11 mission on display at the Kansas Cosmosphere and Space Center. It was retrieved in 2013 put on display in 2015.
I had a chance to see an F-1 at the Destination Moon exhibit which just opened at the Museum of Flight in Seattle. It is astonishingly large. If you're in the area, or near one of the other (few) locations where they are on display, I would definitely recommend going.
(The Apollo 11 command module is there, too!)
> The Apollo 11 command module is there, too
Temporarily though. Normally it's exhibited at the Smithsonian in Washington DC (Which is also absolutely worth a visit).
This is excellent information. I'm headed to Seattle from Norway in May, and will definitely take the opportunity to check this out :D
If anyone's near Huntsville, AL, the Space and Rocket Center has stacks of stuff for your gawking pleasure. Although I think the only F-1 you can go up and rub on is outside the MSFC main building, which you wouldn't get to unless you take a MSFC tour.
Using the fuel to both cool the nozzle and pre-heat the fuel is just genius. (The method first appeared on the V2.)
Concorde used fuel as a coolant also, not to cool the jet nozzle, but ancillary things.
The use of a fuel as a coolant is pretty common in modern high performance aircraft.
And motorcycle engines. Evaporating fuel happens to be a great coolant.
Cool. Any examples or details?
Most “v” block engines even if they are liquid-cooled direct more fuel to the rear cylinder(s) to cool the head, either with larger carburetor jets in older engines or with software in newer ones. If you think about it all ICE are cooled by fuel; if you fed them with boiled fuel they would overheat.
"if you fed them with boiled fuel they would overheat"
It isn't clear to me that they would.
I suppose if your cooling system was highly optimised the extra energy could tip the balance, or is that what you're saying? But then all ic engines would be cooled by their fuel, its kind of inherent to the process.
The SR-71 used it to cool its skin and keep it around 300°C during Mach-3 flight. It also used it to cool its engines.
Isn't it also the SR-71 that has to expand to become fuel tight, and so leaks fuel all over the runway at takeoff.
I wonder how do this kind of crazy engineering ideas happen. What is gained by not making the thing fuel tight unconditionally? Leaking fuel at takeoff sounds extremely dangerous!
> Leaking fuel at takeoff sounds extremely dangerous!
The fuel that was used had to withstand fairly extreme conditions before reaching the engine to be burned, which also made it pretty hard to ignite under ordinary conditions. https://en.wikipedia.org/wiki/JP-7
Expansion due to heat, the GP mentions a temperature of 300c.
Edit to add:
Different materials expand by different amounts due to heat, and different parts of the aircraft are reaching different temperatures.
If you know all the bits are going to expand x amount in normal flight, then you work back to what size they are when cold, if the shrinkage/expansion is too great you possibly could go with a different material (with various tradeoffs) or you just accept the shrinkage as they did here.
It also helps that the SR-71 fuel was almost impossible to ignite. It couldn’t be spark ignited, instead they used triethylborane, an pyrophoric liquid that burns in contact with air. It’s also used to ignite rocket engines as its reactive enough to burn on contact with liquid oxygen.
"almost impossible to ignite" .... "pyrophoric liquid that burns in contact with air"
Those seem mutually exclusive?
Do you mean that triethylborane ignites the fuel?
I don't know why I'm getting down votes for this, its a reasonable question.
To answer my own question triethylborane is indeed used to ignite the main fuel.
You’re getting downvoted because you’re calling out an inconsistency that doesn’t exist. The fuel that doesn’t burn is separate from the pyrophoroc stuff used to start it.
If you know everything is going to expand significantly due to heating, I’m sure making it fuel tight at room temperature would involve significant engineering trade offs.
It's interesting that at the dawn of rocket engines the regenerative cooling was thought to be a bad idea. The reasoning was the hot side will expand too much and disconnect from cool side because of stresses. It's good von Braun actually used one...
It's lighter and you're inherently accommodating for thermal expansion that would happen anyway and possibly cause issues.
> I wonder how do this kind of crazy engineering ideas happen
Three words: the Cold War. See also: Project Sunshine, the Oak Ridge Experiments, SDI, etc. Once you start writing blank checks to the military-industrial complex the crazy ideas start pouring forth.
I suppose on a supersonic aircraft, you can't reject heat into the surrounding air, because that's already been heated up too far by your passage through it. Is this technique still in use in supersonic aircraft?
The new raptor engine actually passes all the fuel and o2 through preburners to vaporize it. Pretty ingenious.
I am finding it astonishing that they could construct them at all, never mind get them to work and not explode.
There is a great deal of this in modern life that we successfully ignore until management failure makes it unmaintainable any more.
And they did it in with slide rules and manual machining.
And they explode quite a lot.
I don't have any data for the F-1, but the typical high power rocket engines of these days had an explosive temperament that needed to be tamed on the test stands. I recall that the SSME had a tendency to self-disassemble vilently on its test stand due to turbo pump failures that took a long time to solve. Other rocket engines probably were very similar in that regard.
At least as insidious to diagnose and overcome, if not more so, was the transient behavior of the engine components outside of even the turbopumps (e.g., valves, preburners, thrust chamber) and the high sensitivity to all the time-dependent characteristics. Robert Biggs, one of the fathers of the SSME, wrote a great book on it, most (all) of it posted online at: http://enginehistory.org/Rockets/SSME/ssme.shtml
IIRC, the injector plate on the F1 had to go through quite a few iterations to keep the engine from exploding.
There were serious problems with unstable combustion causing the motor to break up, that were fixed by putting baffles on the injector plate to divide up the volume, where the combustion started, into smaller independent regions . Once it was running smoothly, the engineers were still concerned that it might be on the edge of instability, so they ran some tests in which a small explosive charge was detonated in the combustion chamber of a running motor, to check that it settled down to running smoothly afterwards.
and according to a documentary I watched, they never knew WHY it stopped exploding, just that they found a plate that made it not explode anymore.
Von Braun: "So what was the problem?"
Ignition! by John Clark mentions it, although I don't think he tells that anecdote. Just that they were going to look at fancier designs, but the showerhead (with baffles) worked just fine.
> ... and not explode.
But it does explode! It's just that the all the explosion power is arranged to go in the same direction
Uh, well, more to the point, the propellant does not instantaneously go off as a single, entire quantity, expending all potential energy in the blink of an eye.
Does a person need to clarify this point, in order to convey intended meaning in this context?
Things explode in a car engine all the time, so it's not really anything special.
It's more of a gradual thwump that progresses through the combustion area than an explosion, unless it's a diesel.
Detonation in a car engine is bad. What's actually happening in a normally operating engine is deflagration.
Deflagration = subsonic combustion
Detonation = supersonic combustion
When people say "explode" they tend to mean detonation.
What I find most amazing about these engines isn't the complexity but rather that they operated successfully despite multiple errors. We are often told that a single failure will destroy the craft, but that really wasn't the case. Shuttle make it to orbit with multiple cracked tubes in one engine. Some bits from the recovered F-1s show clear manufacturing errors. Building something that will work at these energy levels is hard, but creating something that can survive multiple random failures and still keep ticking is spectacular.
The performance and track record of these engines is even more impressive when you realize that all of the design and modeling work was largely done by hand!
This is perhaps apocryphal, but I briefly worked with a guy who’d worked on the F-1 engine, and he said they’d use surplus WW2 grenades to simulate distortions/aberrations in the combustion chamber.
Wrapping the turbine exhaust around the fuel return manifold would seem to be counter-productive---exhaust gasses would heat the fuel, reducing its cooling efficiency as it returns to the top of the engine.
...Unless the fuel at that point was already hotter than the turbine exhaust...
Rocket engines are weird.
The turbine exhaust forms a barrier between the uncooled portion of the nozzle and the chamber exhaust, which is significantly hotter. It's not colder than the unburned fuel being used for regenerative cooling, but the bottom of the nozzle isn't regeneratively cooled anyways.
This engine is still outstanding - after all these years - for the thrust per single chamber it achieves. May be only the RD-270 approached similar value for the parameter. Which is not the most important parameter, but not an insignificant one as well.