Slashdot Mirror
How NASA Brought the F-1 Rocket Engine Back To Life
First time accepted submitter Martin S. writes "How NASA Engineers have reverse engineered the F1 engine of a Saturn V launcher, because: 'every scrap of documentation produced during Project Apollo, including the design documents for the Saturn V and the F-1 engines, remains on file. If re-creating the F-1 engine were simply a matter of cribbing from some 1960s blueprints, NASA would have already done so.
A typical design document for something like the F-1, though, was produced under intense deadline pressure and lacked even the barest forms of computerized design aids. Such a document simply cannot tell the entire story of the hardware. Each F-1 engine was uniquely built by hand, and each has its own undocumented quirks. In addition, the design process used in the 1960s was necessarily iterative: engineers would design a component, fabricate it, test it, and see how it performed. Then they would modify the design, build the new version, and test it again. This would continue until the design was "good enough."'
Bernie Eccelstone is suing for trademark infringement
the design process used in the 1960s was necessarily iterative: engineers would design a component, fabricate it, test it, and see how it performed. Then they would modify the design, build the new version, and test it again. This would continue until the design was "good enough."'
take note modern IT managers - this is agile, not that bastardised process-heavy "agile" scrum-style crap you do today.
I think the article and summary refer to the engineering process. In the 60s you did not have CAD applications. Sure they had computers, on the ship, on the ground and all, but not in the engineering department. Engineers where able to make technical drawings and hand that of to workers building the actual thing. Oh yea and they assisted and oversaw the work done, to correct any misunderstandings. It worked, why add computers.
I mentioned this in a comment last week. Manned spaceflight in the USA is essentially a matter of history, not something we know how to do today. If we wanted (for whatever reason) to go back to the moon, we'd bascially have to start over from scratch. It would probably take as at least as long as the original Apollo program, and cost far more.
After the fall of the Roman empire, knowledge of concrete was lost, and for about 500 years Europeans were walking around Roman buildings and upon Roman roads that they had no idea how to recreate. Right now all our Apollo engineers are dead or dying, and the Astronauts will soon follow suit. Soon there will be no living human who has set foot on another world. Then we will know just how those Medieval Europeans felt when we go look at our old Apollo relics in the museums.
Yes, if one is doing a one off project, or a prototype that will then be given to someone else to redesign, the perhaps this is the a good method.
Every mid to large sized IT shop I've ever seen or worked in (dozens) has basically been a "one-off project" when viewed as a whole. Yes sure, every one is basically built out of off-the-shelf hardware and OSes, but there is so much customization and scripting, customized apps and databases and communication software, and other various "glue" bits holding these microcosms all together, but after you examine the innards of any decent sized IT shop that's been running a while, the place as a whole is actually a giant hodge-podge Rube Goldberg contraption that has evolved and taken final shape over time and iterative development.
We've not building Henry Ford assembly line Model Ts here.
Compared to what we were doing at NASA in the 90s - much less by today's standards - the 60s really were lacking in the barest of computer aids. In hindsight, the assistance of computers was amazingly rudimentary. The ability to do structural analysis was being built "as they needed it" and independently in each group or center - NASTRAN, even in its earliest state, didn't exist yet. These are the people who started developing tools which didn't exist.
You have to remember - this was a time when Battin was using discrete math to plan missions, and a general n-body problem was considered unsolvable (and, afaik, still is in explicit form - but is trivial on modern computers for relevant values of n).
Is it just my observation, or are there way too many stupid people in the world?
This flies in the face of at least history. It also flies in the face of the usual mythology that NASA invented the computer. Which is it?
I never heard that myth. But NASA and its contractors were pioneers in some CAD tech, like FEM (finite element modelling), and the computers for Apollo spacecraft designed at MIT/Lincoln labs were marvels of miniaturization for their day.
But by 1963 manufacturing, at least for the money-means-nothing military, was already computerized.
Maybe a tiny amount of it. Don't confuse NC (numerically controlled) with CNC (computer numerically controlled). NC was developed largely in the late 40's and was widely used by the 50's. It used relay logic and so forth. CNC was too expensive until "inexpensive" minicomputers came along later in the 60's, and didn't take off until micros came along in the 70's. The video probably shows a futuristic "we tried it who cares what it costs" type of setup, like Doug Engelbart's WIMP interfaces in the 60's. Good forward looking stuff, but not necessarily ubiquitous, even for NASA.
The biggest engines we could buy for our model rockets was the D. This F is awesome!
And it's just the F1 !
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
A typical design document for something like the F-1, though, was produced under intense deadline pressure and lacked even the barest forms of computerized design aids.
Thank goodness for that. People still know how to read paper drawings. If it was computerized, we might be able to read the media if it survived (1/2" mag tape or punch cards) but would probably have to spend a lot of time reverse engineering obsolete CAD formats.
They explain why, so that the engineers can get an understanding of large liquid fueled rockets. Understanding the latest attempt seems to be a reasonable step before designing the next one. Also, since they have an engine with known qualities and are building a computer model of it, this will verify that the model simulation is basically correct. If it does not predict known facts (unstable exhaust gas without baffles, expected thrust, etc.) then they cannot trust the simulation on new designs.
I have a news flash for you, young man. Numerical solutions, on computers, for the n body problem were being done in the 1950s, S. von Horner being a notable person in the field.
Yes, analytical math can be used to plan orbits, even done today for first passes. my senior year physics project was orbital calculations by both numerical and multi-variate calculus. No reason what I did couldn't be done on say an IBM 701 or 7000 in the 50s...
These are the types of Articles I still come to Slashdot for ... and for the comments, which have (sadly) diminished in quantity in the last decade. Amazing engineering work, amazing science.
We emerge from our mother's womb an unformatted diskette; our culture formats us. - Douglas Coupland
This flies in the face of at least history. It also flies in the face of the usual mythology that NASA invented the computer. Which is it?
NASA didn't invent the computer. However, in the 1950s computers were room-sized assemblies of hardware. NASA and the Air Force were the only two entities that needed computers that were smaller than that (the Air Force to put in missiles, NASA to put in spacecraft). The Block I Apollo computer was the driver for integrated circuits, and hence the grandfather of all of today's desktop computers (called "microcomputers" back in the old days, when "non-micro" computers meant the Univacs and 1103 and the other big iron of the day.
http://www.computerhistory.org/semiconductor/timeline/1962-Apollo.html
They had no computers, or they invented them?
Both.
It's neither, actually. But by 1963 manufacturing, at least for the money-means-nothing military, was already computerized.
http://www.youtube.com/watch?feature=endscreen&v=_1g1b_EeVHw&NR=1
Why do you think it's called "numerically controlled" and not "digital"? It's because the whole concept is so old that the wording has had time to become obsolete.
The comment you're responding to was about computer design tools--CAD--not about numerically-controlled milling machines. And for that matter, the numerically-controlled milling machines of 1963 weren't really what you would call general purpose computers.
http://www.geoffreylandis.com
They had CAD applications, just not what you think as CAD. Anyways, this is interesting, because when do you think CAD applications started? Did the whole thing just pop into existence fully formed, or were there intermediary steps? Just on the electronics side, look at something like SPICE. It didn't pop into existence with a GUI on a personal computer, it started as a punch-card reading batch application on a mainframe.
SPICE dates to 1972. The Saturn V had been designed, built, flown, and out of production for years by the time SPICE was released to the public.
To be fair, SPICE derived from CANCER ("Computer Analysis of Nonlinear Circuits, Excluding Radiation"). But that was also not released to the public ready until the early 70s (the paper describing it was dated 1971: http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1050166 )
Boom, computer aided design.
"Boom," just in time to be ten years too late to be used in the Apollo program.
http://www.geoffreylandis.com
The Saturn V was designed with paper and slide rules, and very little computer work.
As far as circuit simulation goes, ECAP is probably the first full circuit simulator. I have a copy of the manual for the 1620 version, but it's dated 1970, but there are earlier versions. All written in FORTRAN. I'll bet that almost no computer simulation was done for the electronics on the Saturn V.
For thermal and mechanical FEM analysis, NASTRAN started in 1964 and was delivered in 1968 Not used for Apollo, but used for Shuttle.
Drafting wise, Sutherland's Sketchpad was in 1963, and was pretty much the first "drawing" tool on a computer, and it used custom hardware, and was hardly usable for actual drafting.There were some specialized tools for splines and such in the car industry, but "real" drafting on a computer probably didn't exist until the 70s. (Drafting or Technical Drawing was still a course you took in high school in the mid-70s.)
Banks were among the earliest commercial users of computers.
And tea shops.
They needed them smaller, but banks and businesses needed them cheaper and more reliable.
Correct. NASA was the driver for small computers, where "small" meant "smaller than a room." Pretty much all other applications-- such as the banks and businesses you mention-- used timeshare on big mainframes. Or, for the early 60s, sent the punch-cards to the mainframe to be entered.
By the way, in 1963 banks mostly didn't use computers. You youngsters are too young to remember when a bank "passbook account" meant a physical object that the teller wrote in by hand.
How can NASA be a "driver" for ICs when they were using generic commercial ICs????
They paid the companies to develop those products in the first place, because they didn't exist until the NASA contracts to develop them. The IC was developed with Air Force and NASA funding, because at the time, those were the two customers for whom integrated circuits were an enabling technology.
The comment you're responding to was about computer design tools--CAD--not about numerically-controlled milling machines."
They designed the parts on computers.
Wrong.
They fabricated the parts as part of a computer-driven process.
Wrong.
Look, learn something about 1963 before posting so confidently about how engineering was done with computers back in the early 60s, OK? Do you even know what a slide-rule was???
http://www.geoffreylandis.com
Why? Because sometimes reverse-engineering something that has been shown to work is cheaper than going through the entire development process from scratch. The fact that F-1 engines were built manually and designed without the use of electronic computers doesn't mean that they are much worse for it. Why the heck do you think a lot of Chinese products were reverse-engineered western designs?
A successful API design takes a mixture of software design and pedagogy.
SPICE was a combination of earlier programs...
Right. Specifically, what I said was "SPICE derived from CANCER ("Computer Analysis of Nonlinear Circuits, Excluding Radiation")."
So, you think they didn't use computers to solve numerical problems in 60s? Problems related to design? Is this what you are claiming?
The article said that the F-1 engines were not designed with computerized design aids. That is correct. CAD was just being developed-- in fact, it used to be called "Computer Assisted Drafting", long before it became a design tool-- and was not being used at MSFC back then.
http://www.cadbuilt.com/cad-drafting.html
I don't think you have much of a memory of what engineering was like in the 1960s. You might try some of these:
http://history.nasa.gov/monograph45.pdf
http://history.msfc.nasa.gov/saturn_apollo/documents/F-1_Engine.pdf
http://www.geoffreylandis.com
Don't be brainwashed by all this "process" crap. These days you have to talk to guys in their 60's and 70's to get the full oral history, but they wistfully recall days when the emphasis was on getting things done and making them work, rather than mindlessly following "process".
All that "process crap" is exactly how any successful engineering project is done. The space program in the 60's and 70's was no exception. Do some reading about the actual engineering that went on and you'll quickly realize it was ALL about developing working processes. A process is nothing more than a set of procedures used to accomplish a task. If the task has to be communicated to someone else or cannot be overlooked or is just plain complicated, documentation becomes a vital aspect of the process. You can't build something as complicated as a space ship without a huge amount of extremely robust processes and accompanying documenation. Developing effective production processes isn't mindless busywork - it is among the most challenging and important things we do. The best manufacturing companies spend a tremendous amount of resources on process development because without them they would be unable to function.
If you want to ensure that a rocket blows up, by all means ignore developing processes and don't worry about documenting or communicating the procedures used. Just be a cowboy and "get it done". When you have no way to discover what went wrong, who was responsible, when you were supposed to do it or how to do it again you might begin to understand why process is important. My company makes wire harnesses and we've made products that have gone into space. For even the simplest cable with a crimped terminal on one end we typically have about 15+ pages (and often much more) of assembly instructions, QA instructions, machine setup instructions, QA logs, shipping and packaging instructions, manufacturing orders (how many to build and when to build them), bills of material, training documentation, defect logs, packing slips, and invoices. And every bit of that documentation is genuinely important. Without robust processes in place it would be complete chaos to try to make even the most basic products, never mind something as complicated as a F1 engine. All that "process crap" lets us build a high quality product (repeatedly if needed), diagnose and correct any problems that may arise, and make sure everyone knows what they are supposed to do and when they are supposed to do it.
Sure they had computers, on the ship, on the ground and all, but not in the engineering department. Engineers where able to make technical drawings and hand that of to workers building the actual thing. Oh yea and they assisted and oversaw the work done, to correct any misunderstandings. It worked, why add computers.
Why add computers? Because it makes the lives of us engineers who do what you are describing (I am one) VASTLY easier. I've worked on a drafting table and I know how to use a slide rule. I've designed products and overseen their production. While it can be done without computers I don't really relish the thought of going back to the days without them. People who pine for the "good old days" when we didn't have computers to help with the work almost invariably never had to actually do real engineering without them. Trust me, it sucked.
Do you have any idea how much labor is involved in updating a set of work instructions and ensuring only the most recent version is distributed? Have you ever done a complicated product drawing on paper and then had to do it over again because of a revision? It's possible to do these things without computers but I can assure you from first hand experience that you don't really want to. It's much easier to edit a CAD model of a part and then print out the new revision. It's is FAR easier to use a versioning system to keep documentation up to date and distributed to the right places. We don't use computers just because we can. We do it because it makes us far more effective and faster at our jobs.
And the Navy (to put in missiles as well as shipboard), and the Army (to put on missiles and mobile launching equipment), and all four services to put on aircraft...
Well, yes... and no... Integrated circuits were originally developed for the DoD, which refused to become an early adopter for various reasons, so their availability to NASA was something of an accident. Not to mention the Block I AGC was based directly on the Polaris MKII Guidance Computer...