Reverse engineering seems like a good skillset for the DoD to nurture, anyway. Yes, the US is going to be at the cutting edge of weapons design at least through the medium term, but the leadership isn't arrogant enough to think that all engineering innovations will come domestically. If the US acquired a next-generation Chinese fighter, wouldn't it be handy if someone had the expertise to tell you exactly how its parts were made and machine replicas if needed?
So, the tinfoil hat interpretation of this story would be, of course the Pentagon has the original blueprints in quintuplicate, but it's identified reverse engineering as a key strategic advantage China has over the US (pretty reasonable so far, right?) and has announced this as a way to shuffle some money into building up domestic RE expertise in the defence industry.
I don't think this kind of thing is all that uncommon for military airplanes that have been long out of production. I recall reading a decade or two ago that our local air guard wing had to fabricate some of its own replacement parts for its F-16s (IIRC, then some of oldest that were still in service).
I also recall that Lockheed had to take special effort to preserve production know-how and tooling for the F-22 when they shutdown production early, so it would be possible to restart if needed. You can't capture everything you need to know in blueprints and documentation.
Preventing the loss of manufacturing capability is one of the reasons for offering military aid to friendly countries. We give them money to buy weapons, but they are obligated to buy certain weapons systems with that money.
For example, the Saudi's really don't need all those M-1s they buy from us (with money we give them), but we offer the it keeps our production lines open and active.
It's a devious way of laundering military readiness expenses.
A friend of mine looked into building a P-51 from scratch, as buying a real one was too expensive. He ordered the blueprints, which are readily available. But then he talked to some P-51 owners, who told him the blueprints wouldn't be much use. The airplane was defined by the jigs used to build it, and the blueprints were just drawn up later as documentation and never checked. The jigs were scrapped when the airplane went out of production.
During the escalation of the Vietnam War, the Air Force ran short of jet transports and decided to pull a bunch of WW2 era propeller transports out of mothballs and put them back in service. The mechanics, however, complained that they followed the instructions to the letter on tuning the engines but could not get them to develop the power they were supposed to.
The AF finally hired a bunch of old WW2 airplane mechanics. Those mechanics never looked at the manual, they just listened to the engine, made few adjustments, and the engines were soon singing at full power.
US DOD and their industry partners definitely has this capability already. If they're putting this call out there then there's a real need, not a covert effort to promote the discipline of reverse engineering.
One fears we'd be shocked by how much is actually going to General Smith's brother-in-law's toilet-seat dropship company. We'd like to think that Pentagon auditors could find that sort of thing, except that Pentagon people tell us continuously that they are unauditable.
As someone who works with the government a fair bit, this kind of forward thinking is extremely unlikely to be the case. This is the kind of government competence that only exists in the most wild conspiracy theories.
The idea that it was made by some long defunct subcontractor and nobody knows what happened to their blueprints and production notes is far more likely. Also, even if you had the blueprints you still need to be able to make the tooling, which is at least half of this project.
That said, if some contractor does come through with this and does a good job (on time, on budget, parts work, etc...) they could likely find plenty of other programs with similar needs. In some cases the original contractor does still exist, but they're asking way too much money to reconstitute an old production line so a scrappy startup like this could prove themselves valuable.
It's also possible it was "Hey! The foam injection machine isn't working, I've tried everything", "There's this string trick I used once, try that", "Great, that fixed it, we should document that", "Sure, once we get all parts made that have backed up"
Or, maybe I'm reading bad practices in my industry to everywhere else.
I would be very interested in reading case studies where a company both does documentation well and does not get bogged down in vast amounts of paperwork.
As yes, plenty of fixes for things are not written down anywhere.
Unfortunately, every company that I have been at that has tried to implant the Toyota Way just bolted additional processes onto their legacy ones, adding complexity.
I've seen that way too many times. "Why doesn't this link in the internal console work?" "Oh, after you click it and get the error, you have to remove the trailing dash character and refresh." "Weird, I wonder how this slipped by Jared in QA when he tested it. We better let him know." "Don't bother, Jared is the one who showed me how to remove the dash...."
No, the problem is that while you are writing documentation you are not making money for the company. I remember when I and Dave as the senior tech told our boss that we should put the most common repairs into a database so new hires would not have to ask us questions all the time.
He asked how long the job would take, and we answered about a week for all the common stuff. And his answer was, "That is a week you will not be making the company any money.".
We tried to explain the long term savings, but he was not interested.
I think the best way to think of this part of a wiki is as a cache for everybody's help. No need to pre-populate the cache, just start using it, and the most frequently used information will tend to be there.
It might have been protectionism, but also life is just complicated. Documentation is always incomplete and much of it is out of date as soon as you write it down. Humans have so many cognitive biases, we don't even notice half the things we do.
Many scientific studies cannot be replicated either due to errors or ambiguous steps.
You only have what you test. If you want working documentation, you have to do a blind test. That can be pretty dang expensive, and most people don't think that way. Running production is heavily motivated to just keep running, and not worry too much about all of the potential problems - there are just too many potential problems.
Conversely, a lot of management have zero interest in knowing this kind of detail. It may not be possible to get the info fed back. People may be punished for not following the process, even if it's impossible to get the job done by following it exactly ( this is a very common dysfunction!)
Yeah this sort of "why didn't the lowest-paid employee do more work without permission so her boss's boss's boss wouldn't have to think too hard" suggestion surprises me a lot more often than it should surprise me. Sure, it would benefit shareholders, but companies are run by executives.
As some siblings have commented, it's probably the failure to get back and document it. But the desire for job protection is not uncommon. I saw it when I worked on testing satellite power converters at a large aerospace company. There were procedures and paperwork for everything, but the test procedure for one series of power converters did not work. Only one technician knew how to actually test them. He led me through it, and he had plenty of time to update the procedures, but refused to do it. It was gross and obvious what he was doing, but somehow he was allowed to get away with it. There was no union.
So many problems could be solved by giving every adult $1,000,000 in a one-time, tax-free, no strings attached, lump sum.
The technician refused to document the procedure because he was afraid of becoming replaceable. He was afraid of becoming replaceable because he was afraid of losing his job. He was afraid of losing his job because he would lose his sole source of income.
I'm pretty sure this is sarcasm. But I never know these days because a lot of people actually avocate things like this. After seeing what happens to most people after they receive a lot of free stuff, especially large sums of money, I would never advocate something like this. There's also the rather obvious issue about how it would be paid for but, again, there a lots of people who think you can just print money out of thin air forever and nothing will ever happen.
>As someone who works with the government a fair bit, this kind of forward thinking is extremely unlikely to be the case. This is the kind of government competence that only exists in the most wild conspiracy theories.
No, but we'll totally sponsor/fund research into things like mind control and astral projection. Funding something totally useful like reverse engineering skills is totally anathema. Just really sad commentary
> No, but we'll totally sponsor/fund research into things like mind control and astral projection. Funding something totally useful like reverse engineering skills is totally anathema. Just really sad commentary
I don't get your point. The government does invest in reverse engineering, and much more than they invest into mind control and astral projection. It's nonsense to say that investing in useful skills is "totally anathema".
What GP was saying was not that they don't invest in reverse engineering, but they wouldn't try to backdoor it or secret the experience into the industry with some project like this. They wouldn't try to "trick" the industry into developing the experience by putting out smaller contracts like this. They do invest in it, but directly. That one program office has put out a contract for reverse engineering of one system does not mean that the general capability is ignored, within the government or the industry.
Small departments in poor cities occasionally wind up highly competent because they wind up staffed with a high enough ratio of workers to get things done and everybody above them has bigger problems than micro managing so they have the freedom to get things done. And it runs for awhile then comes under scrutiny and gets made back into junk because politics.
Why does it seem to be that we always assume that our own government is incompetent and unable of forward thinking, but all other governments (China, etc.) are?
People assuming the worst mostly. US intelligence used to always overestimate Soviet capabilities too. Usually by a large margin.
It’s a bit terrifying when you realize that the people at the top don’t know what they are doing either. Much of the world runs on guesswork and inertia. This is also true of most companies.
"It’s a bit terrifying when you realize that the people at the top don’t know what they are doing either. Much of the world runs on guesswork and inertia."
Because we can't see what's going on inside the Chinese government so they get benefit of the doubt. In contrast we know exactly how the local sausage is made.
China doesn't have to deal with a lot of the "complications" that the Western governments have to deal with, like human rights, etc. This gives those in power the potential to achieve their goals more efficiently.
Not saying this is how it actually ends up, but they certainly are prepared to use a lot of leverage tools that are not available to more free countries.
“...[The Russians] have an immense advantage. I think it’s a dangerous advantage and a temporary advantage. Twelve men decide what to do. They don’t sit here like I do and have twelve others question me on why I’m doin’ it... In the end they destroy themselves.”
—General George Marshall
https://youtu.be/Y7eVaVUESCo
It sounds like it could be they have the original blueprints of the first ordered plane but by the time you received the 100th they were different and then it’s had overhauls and other modifications since. About as useful as a ten year old X-ray for a 25 year old.
>but they're asking way too much money to reconstitute an old production line so a scrappy startup like this could prove themselves valuable.
The air force is currently going on an in-housing and "agile all the things" binge because they're sick of 30yr of getting screwed by contractors.
My money is on Northrop giving them a sky high but not without precedent price and someone saying "screw it, this is the perfect thing to learn how to do"
I don't think we need the tinfoil hat. This sort of thing happens sometimes. Like the time we forgot how to may part of our thermonuclear weapons and had to reverse engineer that [0].
> I don't think we need the tinfoil hat. This sort of thing happens sometimes. Like the time we forgot how to may part of our thermonuclear weapons and had to reverse engineer that [0].
The really interesting thing about that is one of the main things they had to reverse-engineer was an impurity in one of the ingredients that the original designers didn't even know they were depending on.
Reminds me of how NASA had to design completely new rocket engines for SLS because while they still had the plans for the Saturn-V rocket engines, back then plans weren't followed to such high accuracy as nowadays a 3D model would be followed. Each engine was custom tailored with small modifications here and there. The knowledge to custom tailor them was lost, so they went with designing new engines from scratch instead.
> Each engine was custom tailored with small modifications here and there
This was true of lots of things outside NASA. This was one of the problems that killed the UK attempt to modernise the Nimrod maritime patrol aircraft [0]. They wanted to fit new wings to the old planes, but only discovered after the contract had been signed that the different airframes had not been built identically, but were different interpretations of a common plan. Each of the nine planes was in effect a completely new refit challenge.
It reminds you of a thing that never happened? That's a neat trick. The SLS uses the RS-25 engines, the Space Shuttle Main Engines which are in no way an engine "designed from scratch".
Additionally the F-1 engines were redesigned using more modern techniques around 2012 with the F-1B [0].
The SLS went with the RS-25s for a number of reasons. While the F-1s provide high thrust they're not as efficient. The RS-25s can burn longer and with the SRBs providing massive amounts of thrust to get the stack off the ground the RS-25s can provide more delta-v to the upper stages. An F-1 reboot would have required a lot of extra work to add in modern computer controls which already exist in the RS-25. They have also flown in the past few decades and are well understood.
The comment you are replying to is "lying by omission" (if you want to call it that) but the video talks about your point so the snarky tone puts you in a worse spotlight than the original comment.
Basically, someone created a F1 engine inspired design with modern manufacturing technology but ultimately it was never used.
I heard this story as well. But in the end they picked something else they had and could make. I think the original idea was to start with the same as what went to the moon and build on that. At one point we made several thousand of something like them. They found they could not build that anymore, and it was not as good as stuff they could do, so they redesigned to something else that they could. That is as far as most people take that story.
In fact most 'stories' and 'news' is like that. The SLS is known to be in a congress sub committee hot potato game. So that there are different versions floating around of what happened and is happening is not surprising. Teasing out the 'truth' of what happened is not always easy. As many have interests in distorting it, or just remember wrong.
There were not thousands of F-1 engines manufactured. There were a bit over a hundred manufactured between flight and test models.
The whole "we can't make them anymore" meme has a tiny kernel of truth but it's been blown way out of proportion the more it's been repeated.
When the F-1 was manufactured a lot of the actual assembly of parts was done by hand. That meant Rocketdyne's engineers and machinists had a lot of tribal knowledge that was lightly or partially documented at best or not documented at worst. NASA and Rocketdyne maintained all of the designs and documentation and knew everyone involved in the engine manufacturing.
In the early 90s a lot of these Rocketdyne people were interviewed [0] and asked to review F-1 manufacturing techniques as part of the Space Exploration Initiative [1]. Restarting production of the F-1 would have been practical but more expensive than the alternate plan to go with an SSME derivative.
While the F-1 is an impressive engine it has downsides. It's enormous and heavy. The Saturn V's first stage (S-1C) was 10m in diameter while the Space Shuttle's external tank was 8.4m in diameter. The Shuttle's ET was going to be the core of the proposed SEI rocket (and SLS). With a 8.4m tank diameter it would be more difficult to mount more than two F-1s.
With only two engines it would be more dangerous of a vehicle for manned missions as a flame out in one engine could be disastrous. It would also make it more difficult to throttle the engines for different launch profiles. Moving to a larger core would have required all new manufacturing infrastructure, flight qualification, and unknown follow-on issues.
Tl;dr The F-1 was an awesome engine but not perfect or even desirable for all purposes and the "we can't build them" meme is stupid.
Wikipedia calls it an "aerogel", but rumors have it originally being closer to ordinary "closed-cell extruded polystyrene foam", which is often sold under the brand name Styrofoam. It might be something else entirely now. It's role is said to be absorbing the x-rays generated by the fission stage and turning into a plasma to help ignite the fusion stage. Edit: https://www.armscontrolwonk.com/archive/201814/fogbank/
Even if you're not innovating everything you want to reverse engineer enemy systems because you want to discover vulnerabilities. This is common practice.
I've heard from various acquaintances that it's not unusual for the contractor to be required to destroy documentation and tooling for certain classified projects upon completion.
It is absolutely standard practice for this to happen. At contract completion, the contractor must destroy or return all classified materials: these sorts of designs would certainly be included.
It's pretty interesting, but a lot of this stuff is structured specifically so that the contractor organization doesn't retain knowledge that it's "not supposed to have", as wasteful as that ends up being. Even stuff like a list of personnel who worked on a project or production line might be included... so that a few years after the project is completed, the management of the contractor might not even be able to figure out who to ask, because they've destroyed even their knowledge of the list of people who worked on the component's design and manufacturing.
I wish my tax dollars were going to some galaxy-brain forward thinking group of men and women capable of making such reverse engineering challenges for itself; but it almost certainly is the case that what looks like sheer incompetence is just that.
Read about the history of NASIC and the other service intelligence centers [0]. Understanding foreign technology is one of their core missions. I'm pretty sure they have already been reverse engineering stuff for a long time.
Nah, I think the one guy who knows the thing is gone, the data was on a hard drive that crashed and the backup system didn't work, and the papers copies are probably somewhere except no one knows where and it is faster to reverse engineer the damn thing.
I have no reason to believe that defense projects are not plagued by bureaucracy, office politics and incompetence like in a lot of government entities and large companies. If anything, it is worse because secrecy is a convenient way to hide incompetence. And when information is shared on a "need to know" basis and we realize that no one "needs to know", then it gets lost.
I have touched a bit on defense projects (not in the US) and generally, compared to civilian, open projects, if feels slow and primitive. I didn't get to the super-secret stuff but I have a feeling that the only parts where civilians don't do much better are the ones that only the army is interested in.
Yeah no. Back in the 2000s when Iraq war was happening and marines were using MSN Messenger and I asked whether the unencrypted nature of the chat program would adversely impacted on their missions etc, and he basically said "hell they can spy all they want WE don't even know what we're doing next"
> If the US acquired a next-generation Chinese fighter, wouldn't it be handy if someone had the expertise to tell you exactly how its parts were made and machine replicas if needed?
I suppose they could always call the engineers at Boeing or Lockheed who designed it in the first place. Like getting tech support for your Huawei networking equipment, just use the Cisco guide! [0]
For anyone unaware Lockheed was hacked in 2007 and the hackers got away with technical documentation related to the F-35 Joint Strike Fighter. The hack was attributed to hackers in China. So it is thought that the development of the Chinese J-20 was at least helped along by information from the hack.
I had to check, it's not as ancient as some other bombers with B-2 first flight occurring in 1989. I met a pilot who flew on the same B-52 as his father, and his grandfather had flown some of the first B-52s. This was in the 00s, about 15-18 years ago.
It's impossible to overstate how long these things have been in service. The oldest B-52 airframes still in active service were built longer ago than the first flight of the Wright brothers was when they were built.
I worked at Boeing Wichita (which is no more), and an engineer told me that the B-52 was getting heavier over time because the old wires weren't removed for avionics upgrades. Of course, they were always trying to put one over on me as a newbie (I worked on avionics S/W in Ada for the tankers).
He also said a big problem was the expense of doing a modern engine upgrade. Anyone know if they ever re-engined it? Was he joking again?
Having one of these fly over my apartment on Rock Rd. was a dish rattler--I used to see B-2s every once in a while too (fairly loud too, IIRC).
The introduction of the B-2 in service was ~24 years ago, this is one of the most recent jet in the US air force. By military standard it's definitely young.
There are only 2 planes in service in the US more recent than that, the F-22 and F-35 (not counting transport aircrafts).
Those planes are designed for lifespans of 50-100 years.
My dad was at the Skunk Works for 20 years. He then went on to work at Northrop for a good chunk of time as well.
The B-2 was started in the 70s and the maiden flight occurred while I was still baking in the womb.
I realize aircraft can live a long time and can be retrofitted and upgraded with modern technology - but all things considered it is old.
Not saying ancient is bad, either, but as far as organizations go, org knowledge, handoffs of info and books and writing etc... I am not surprised to see the need for reverse engineering.
> Those planes are designed for lifespans of 50-100 years.
No, they weren't "designed" for a century. Think about it for a minute - would a military contractor even admit it if they did? There would be no new programs. :)
The reason the B-52, F-16 (and T-38 for training, F-5 for aggressor training, etc.) are still around is because they were made in quantity.
By contrast, the F-22 had a run of about 190 airframes, and the B-1/B-2's dozens each, with only a handful (like 3 - 5) of each model operational at any given time - it's really that bad.
All they had to do for the B-52 was to update the avionics and add a cruise missile rotary launcher in the bomb hold. (Believe it or not, that is the main thing that keeps the B-52 in operations.)
I read once that if you tried to build WWII-era aircraft according to specs and mechanical drawings, they wouldn't fly. The real design was found in the dies and jigs that were adapted after many iterations and test flights.
Also sounds like a git repository where people were working on local branches to do production work and then when they left the company, never pushed their branch upstream.
The analogy to out-of-date software documentation is very apt. As problems or inefficiencies popped up in the factory, they'd change things and just not write them down (because of deadline pressure, often). If you wanted to build new ones you'd have to either get the factory workers to show you how the thing was actually built, or tear down a working example to reverse engineer it. (Which has been done!)
If you find this stuff interesting, pick up a book aimed at hobbyist metal lathes. Just reading a little about how you set up a tiny home version can give you a whole new level of understanding and respect for this kind of stuff. Structural components are different but in that era and even to some extent today, most precision parts were machined with two primary tools: the lathe and mill. Today both would be CNC and you'd add other modern tools but the CNC lathe is still a lathe and they work the metal in the same fundamental way.
Due to the dispersed mass production in WW2 it's probably not totally true that they wouldn't fly without a custom tooling setup (it'd be a maintenance nightmare if two examples of the same model from different plants diverged dramatically, altho of course it happened) but you can be certain that it's of critical importance.
Statistical quality control was not widespread at the time outside of a few enlightened industries. The Army was much more likely to buy a lot of parts where they could expect most of them would fit, rather than having all of the parts in statistical control within the tolerance limits.
I only worked tangentially in the field, but I get the feeling that modern semiconductor manufacturing is sort of the same, except with more excel sheets and the like?
Semiconductor manufacturing is essentially the interchangeable parts engineering taken to the extreme. The tolerances are so large (relative to the feature size) that instead of trying to eliminate them you exploit the relations between various manufacturing tolerances such that the errors cancel out.
A huge amount of industry all over the place ends up like this. The engineers produce the theoretical form of the design, and the tooling up process refines it into a practicable-at-scale reality.
I've done installs of video streaming tech on USAF-managed NIPR net and we've had to reverse engineer some of the mechanisms on their own network... This does not surprise me at all.
This happens even outside Defense in all specialized industry. I was once asked if I could reverse engineer a part for a 30 year newspaper stacking machine in a printing press. Very interesting stuff and a consequence of being niche.
- Limited production run. These aren't Ford Mustangs or Toyota Priuses with hundreds of thousands and millions being produced. There were 21 built over 13 years. This also leads to the consequence that each one should be considered a bespoke creation and not a "copy" of the others (even ignoring the 20+ years of individual maintenance work they've each had).
- Time. It's been 21 years since the last one was produced in 2000. Whatever facility produced this has long since lost that capability.
- Aging workforce. Whoever designed it is likely retired, and could even be dead at this point. Certainly the senior engineers who may have been 40+ when the project started in the 70s/80s. Even if they weren't retiring and dying, the people on the project have been doing other things for 20+ years.
The last part is extremely true. I have known many engineers that worked on the B-2 and have since retired. There's very little chance many of them would come out of retirement and honestly I don't think the government would pay for that.
Large scale printing machine setups are insane. I've seen very old installations go for prices that are higher or comparable to a new setup just because you know the throughput and you know the civil engineering is sound.
> ... Air Force’s Rapid Sustainment Office (RSO) was looking to industry for a “cutting-edge, automated 3D scanning system,” specifically intended to replicate aircraft parts that are no longer in production, including at maintenance depots.
You see the real issue? Even if they have the blueprints, it will be difficult to even manufacture the parts here.
There's so many more problems like this coming down the pipeline. We've effectively lost (or rather acquiesced) almost all manufacturing skill to other countries. This was a result of the largest corporations taking advantage of labor arbitrage and then flooding their home market out of business.
If we really want to reverse this, we need to do what it takes to bring manufacturing back to the United States. It's a painful reversal of the open-trade policies of the past. Honestly, Trump was right to fight so hard for this, even going so far as to remove us from NAFTA.
I don't have a good strategy for this, but probably some set of minimum manufacturing of everything should happen within our borders, say 10%, and accomplishing that target with a combination of sticks and carrots.
Unfortunately, your assessment is very correct. There will be no solution and only a slow decline as previous systems become obsolete and there no on to fix or upgrade to better systems.
This country is producing more Tik Tok stars than engineers that can sustain civilization.
Which country built the pocket computer/communicator? Which country built the electronics, circuit boards, cables, batteries, housing and fastening screws?
Which country built the components and systems and subsystems that go into manufacturing an electric car?
Which country has lithium, rare earth metal processing capabilities critical in manufacturing all modern electronic systems?
Which country built the components and subsystems in the vaccine manufacturing plants?
Decades ago, US had the education system to move millions of people through manufacturing and engineering pipeline to build all above things.
Now, that US education system does not exist. The manufacturing base does not exist. The engineering pipeline does not exist.
We produce more dropout Tik Tok stars than Professional Engineers that develop critical cooling cores, that can withstand heat and stress thresholds in modern military aircraft, and master welders with 20 years of experience that can weld specialty metal alloys in compact designs.
It’s not just aircraft. Do you know who works on water treatment plants that provide clean drinking water to millions of people in US? These guys will be retiring in few years. Who are going to replace them? Who are going to do the hard and difficult job of maintaining these critical systems?
This brings up an important principle: the higher the tech, the shorter its lifespan. A cellphone from the 1980s won't function anywhere but in a Faraday-caged lab equipped with a suitable cell site, while 2G and 3G systems are fast disappearing. Only recently have landline telephone exchanges begun to drop support for pulse dialing, but otherwise, a rotary-dial desk phone from 100 years ago would still work today, and DTMF dates to 1963.
> a rotary-dial desk phone from 100 years ago would still work today
This still amazes me, even though every time a call comes in on the landline, the 50-year-old rotary-dial phone in my living room rings alongside the cordless phone. It doesn't get used much, except for the loud ringer and during extended power outages.
You can get hundred year old manual lathes/mills and they still do their job. It's because the basic principle upon which they were built is still valid today. The only difference with modern machinist tools is that they are fully automated and you can just retrofit CNC parts to the old equipment.
Somethign I haven't seen mentioned here is that the technology used to design this thing is waaaaaaaaaaaay old. I know for a fact it was designed on an IBM mainframe (dad ran the one that the YF-22 was designed on). I can't remember what CAD package they used there, but its definitely conceivable that the software doesn't even exist anymore. How would you go about building a new one if you can't read the CAD files, if you have them.
Hell, the building in hawthorne where dad worked isn't even owned by northrop anymore. Its spacex.
Your story reminds me of what I read about the McLaren F1, that it had a particular 90s Compaq that had to be kept running in order to service them. As it turns out, this is recently no longer true.
He also has "the laptop"—a mid-1990s Compaq running custom DOS software created by McLaren when the F1 was new. This gray brick was, for many years, required to access the car's engine control and body control modules. Today, McLaren uses a modern Windows computer running a software emulator for day-to-day computer maintenance. Hines keeps the vintage Compaq around just in case.
Good callout. However, it might not be all that bad. CAD file formats are known to last for decades to maintain backwards compatibility, since the packages are expensive and upgrade paths of large companies slow.
CAD files (at least planar modeling as opposed to parametric or solid modelling) tend to have pretty straight forward data structures for geometry. Find the point or vertex struct and you're halfway there. The complexity is association and metadata which were nowhere near as bloated then as they are now.
As someone who works regularly for large asset owners, asset life-cycle information (digital and analogue) and the proper management of that information is often ignored because organizations pass that risk to the people working with the asset maintaining their own working knowledge to get their job done. We've had international Building Information Modeling standards since the turn of the century but there's often no business case or standard enterprise process to make leverage technology for asset management.
AI and ML are starting to appear more and more in the operations space, however many pilot projects have to start with digitizing the asset in the first place, which becomes a cost impediment for many getting off the ground.
It really frustrates me as there are great things that can already be done today but the initial step financially usually scares off anyone who has to know 5 years in advance what their operating budget is.
This country simply does not have enough skilled engineers to design and produce original parts to requirements. Let alone reverse engineer a worn out part.
B2 is product of its time and the amount of engineering resources that went into design and development simply can not be replicated with current society.
Tik Tok stars make multi millions of dollars, so why would any smart person spend years in school and solve complex problems to build state of art Stealth Plane.
Boeing can’t even produce a simple jet without having it fall out of the sky few times, 737 Max.
I would disagree on this point. I work in aerospace with a a lot of smart folks. Boeing's 737 issues are a product of many years of poor management choices, but I don't think represent the talent of the industry as a whole.
737 Max is just the poster child for issues in aerospace. There are many other examples, F35, 787, F22, are just recent examples.
The flow of smart folks in the pipeline are dwindling. You may be working with smart folks, but how many are going to retire in few years? Who are going to replace them? Does the current education system produce the number and quality of candidates needed?
Project failures cannot be simply attributed to a workforce problem without a detailed analysis especially when the top example is a widely reported management failure.
It’s an industry failure, society failure, not just a single project failure.
This country can not build a new reliable airplane, like it did decades ago.
It’s failed project after project. Ask yourself why this is the case? What has changed in this country? What has changed in the industry. What has changed in society?
It’s not just aircraft. Where are US high speed rail systems, that other countries built 30, 40 years ago?
Where are modern infrastructure?
Who works in water treatment plants, providing clean drinking water for millions of people? What happens when all these guys retire in few years?
Tik Tok is all fun and games, and no one wants to do the hard messy work of treating water for consumption.
The F35 seems like a project management issue more than engineering. No one stopped the scope from creeping out of control as new tech and capabilities were added.
> No one stopped the scope from creeping out of control as new tech and capabilities were added.
The problem with F-35 wasn't scope creep, it was the initial premise which operated under a mistaken belief that getting a single “do everything” airframe to maximize commonality would drive down unit cost, rather than driving it up through the conflicting needs of distinct missions.
> EFF and other pro-reverse engineering organizations should sue and demand that the DOD follow the same laws that citizens follow.
They'd have to have standing for such a suit and they have none. NG and others involved in the design and manufacturing might have standing, but they're also unlikely to sue. This is not a novel activity for DOD.
Yes, as the US cedes the economy to China, they will also surely become stronger militarily. And with the US (and other Western countries) in a perpetual lockdown in the coming years, which will accelerate the transition - we had better make sure we are able to copy the tech!
One of the proposals for future upgrades to SLS (ugh) was liquid-fueled boosters using new-build F-1 engines (i.e. Saturn V main engines). Unfortunately, the specs were not complete; not only was lots of stuff different in the factory than in the drawings, but parts of the engine were literally hand-built. Think API documentation built on a tight deadline, and published while software is still under development.
Luckily, some NASA engineers were already taking one apart and modeling the parts for previous SLS work. The new Dynetics version ended up much simpler and cheaper because of new manufacturing methods, but the reverse engineering process required to get there was a serious project in its own right.
The US nuclear weapons program had to do something similar in the 2000's with a material known as Fogbank. This is used in the W76 warhead as an interstage material, and they needed more of it to keep the weapons in operational state.
Unfortunately, they'd forgotten the manufacturing process, decommissioned the plant that made it, and the people that knew how to do it weren't around any more. They figured it out, but it doesn't necessarily look easy to do (nor does the material seem easy at all to make in the first place).
My grandfather spent the last part of his career doing this sort of work—often on classified projects, I'm told. Usually there was an assumption made on the project about the useful working life of any part, and the expected duration of the program, and they made the "right" number of spares.
Things change, and the factory/tooling/people are long gone. Also, often the underlying tech isn't available. But the part needs to weigh the same amount, meet the same guidelines, and fit in the same hole.
It's cheaper to pay someone to spend the time to figure out how to make the replacement then to mothball the entire airplane—almost regardless of how expensive it might be.
This is exactly the sort of problem that Jon Blow talked about in his talk about preventing the collapse of civilization [0] -- the human race forgets how it does amazing things.
One of my aerospace professors in college used to be the secretary of the Air Force and B2 development started during his time in office there (back when Carter was president). I remember him saying that one of the missed opportunities in his opinion for the U.S. was not buying more B2 bombers. They can reach anywhere on earth and we only have 20 of them. But Congress didn’t want to keep making them after the Cold War.
Advances in additive fabrication, plus other advances in substractive machining and the use of sintered metal molded preforms, composites etc., explain most of this. The time saved with the newer methods yield a net efficiency.
264 comments
[ 4.5 ms ] story [ 261 ms ] threadhttps://news.ycombinator.com/item?id=26321523
https://news.ycombinator.com/item?id=26322477
---
As an aside, does it still qualify as a dupe if there wasn't any significant discourse?
Generally no. Some submissions just don't take due to poor timing (not lining up with when an interested audience will catch, upvote, and discuss it).
So, the tinfoil hat interpretation of this story would be, of course the Pentagon has the original blueprints in quintuplicate, but it's identified reverse engineering as a key strategic advantage China has over the US (pretty reasonable so far, right?) and has announced this as a way to shuffle some money into building up domestic RE expertise in the defence industry.
I also recall that Lockheed had to take special effort to preserve production know-how and tooling for the F-22 when they shutdown production early, so it would be possible to restart if needed. You can't capture everything you need to know in blueprints and documentation.
The difference in technology likely plays a large role - these parts were designed without CAD and modern fabrication protocols have changed.
It reminded me of NASA's endeavor to reverse engineer the Saturn rockets
https://arstechnica.com/science/2013/04/how-nasa-brought-the...
Actually, the B-2 was one of the first military aircraft designed with CAD.
https://www.northropgrumman.com/wp-content/uploads/B-2-Spiri...
(search for "NCAD")
For example, the Saudi's really don't need all those M-1s they buy from us (with money we give them), but we offer the it keeps our production lines open and active.
It's a devious way of laundering military readiness expenses.
The AF finally hired a bunch of old WW2 airplane mechanics. Those mechanics never looked at the manual, they just listened to the engine, made few adjustments, and the engines were soon singing at full power.
> You don't actually think they spend $20,000 on a hammer, $30,000 on a toilet seat, do you?
The budgets they use for secret reverse engineering aren't booked as any kind of reverse engineering.
https://www.thenation.com/article/archive/pentagon-audit-bud...
The idea that it was made by some long defunct subcontractor and nobody knows what happened to their blueprints and production notes is far more likely. Also, even if you had the blueprints you still need to be able to make the tooling, which is at least half of this project.
That said, if some contractor does come through with this and does a good job (on time, on budget, parts work, etc...) they could likely find plenty of other programs with similar needs. In some cases the original contractor does still exist, but they're asking way too much money to reconstitute an old production line so a scrappy startup like this could prove themselves valuable.
Anyway the process as such did not work as documented.
You had to add strings to the inside of the mold to get it to set correctly. This was not written down anywhere.
Part of it may have been job protection. The union was bad about that. New hires would not be able to complete the work.
Always stuck with me that at some point the ability to make that part would be los-tech.
Or, maybe I'm reading bad practices in my industry to everywhere else.
As yes, plenty of fixes for things are not written down anywhere.
Have others come in and follow the directions.
It is totally possible that the line workers have no idea what the documentation on the tooling says.
Every time i am dealing in new hires, i realise how much we just take for granted
He asked how long the job would take, and we answered about a week for all the common stuff. And his answer was, "That is a week you will not be making the company any money.".
We tried to explain the long term savings, but he was not interested.
Many scientific studies cannot be replicated either due to errors or ambiguous steps.
You only have what you test. If you want working documentation, you have to do a blind test. That can be pretty dang expensive, and most people don't think that way. Running production is heavily motivated to just keep running, and not worry too much about all of the potential problems - there are just too many potential problems.
https://www.sarna.net/wiki/Lostech
Management gets exactly the behaviours from employees that it rewards and incentivises.
The technician refused to document the procedure because he was afraid of becoming replaceable. He was afraid of becoming replaceable because he was afraid of losing his job. He was afraid of losing his job because he would lose his sole source of income.
Let's just give people free money already.
No, but we'll totally sponsor/fund research into things like mind control and astral projection. Funding something totally useful like reverse engineering skills is totally anathema. Just really sad commentary
I don't get your point. The government does invest in reverse engineering, and much more than they invest into mind control and astral projection. It's nonsense to say that investing in useful skills is "totally anathema".
What GP was saying was not that they don't invest in reverse engineering, but they wouldn't try to backdoor it or secret the experience into the industry with some project like this. They wouldn't try to "trick" the industry into developing the experience by putting out smaller contracts like this. They do invest in it, but directly. That one program office has put out a contract for reverse engineering of one system does not mean that the general capability is ignored, within the government or the industry.
And, possibly, before the 1960s.
Small departments in poor cities occasionally wind up highly competent because they wind up staffed with a high enough ratio of workers to get things done and everybody above them has bigger problems than micro managing so they have the freedom to get things done. And it runs for awhile then comes under scrutiny and gets made back into junk because politics.
I agree. And there's a massive depth unwinding to that.
It’s a bit terrifying when you realize that the people at the top don’t know what they are doing either. Much of the world runs on guesswork and inertia. This is also true of most companies.
In my early 20's this 'discovery' shocked me
Not saying this is how it actually ends up, but they certainly are prepared to use a lot of leverage tools that are not available to more free countries.
People are just people. Organizational psychology is universal. Bit rot is inevitable.
The air force is currently going on an in-housing and "agile all the things" binge because they're sick of 30yr of getting screwed by contractors.
My money is on Northrop giving them a sky high but not without precedent price and someone saying "screw it, this is the perfect thing to learn how to do"
0 - https://en.wikipedia.org/wiki/FOGBANK
The really interesting thing about that is one of the main things they had to reverse-engineer was an impurity in one of the ingredients that the original designers didn't even know they were depending on.
https://www.youtube.com/watch?v=ovD0aLdRUs0
This was true of lots of things outside NASA. This was one of the problems that killed the UK attempt to modernise the Nimrod maritime patrol aircraft [0]. They wanted to fit new wings to the old planes, but only discovered after the contract had been signed that the different airframes had not been built identically, but were different interpretations of a common plan. Each of the nine planes was in effect a completely new refit challenge.
[0] https://en.wikipedia.org/wiki/BAE_Systems_Nimrod_MRA4
Interchangeable parts don’t.
Additionally the F-1 engines were redesigned using more modern techniques around 2012 with the F-1B [0].
The SLS went with the RS-25s for a number of reasons. While the F-1s provide high thrust they're not as efficient. The RS-25s can burn longer and with the SRBs providing massive amounts of thrust to get the stack off the ground the RS-25s can provide more delta-v to the upper stages. An F-1 reboot would have required a lot of extra work to add in modern computer controls which already exist in the RS-25. They have also flown in the past few decades and are well understood.
[0] https://arstechnica.com/science/2013/04/new-f-1b-rocket-engi...
Basically, someone created a F1 engine inspired design with modern manufacturing technology but ultimately it was never used.
In fact most 'stories' and 'news' is like that. The SLS is known to be in a congress sub committee hot potato game. So that there are different versions floating around of what happened and is happening is not surprising. Teasing out the 'truth' of what happened is not always easy. As many have interests in distorting it, or just remember wrong.
The whole "we can't make them anymore" meme has a tiny kernel of truth but it's been blown way out of proportion the more it's been repeated.
When the F-1 was manufactured a lot of the actual assembly of parts was done by hand. That meant Rocketdyne's engineers and machinists had a lot of tribal knowledge that was lightly or partially documented at best or not documented at worst. NASA and Rocketdyne maintained all of the designs and documentation and knew everyone involved in the engine manufacturing.
In the early 90s a lot of these Rocketdyne people were interviewed [0] and asked to review F-1 manufacturing techniques as part of the Space Exploration Initiative [1]. Restarting production of the F-1 would have been practical but more expensive than the alternate plan to go with an SSME derivative.
While the F-1 is an impressive engine it has downsides. It's enormous and heavy. The Saturn V's first stage (S-1C) was 10m in diameter while the Space Shuttle's external tank was 8.4m in diameter. The Shuttle's ET was going to be the core of the proposed SEI rocket (and SLS). With a 8.4m tank diameter it would be more difficult to mount more than two F-1s.
With only two engines it would be more dangerous of a vehicle for manned missions as a flame out in one engine could be disastrous. It would also make it more difficult to throttle the engines for different launch profiles. Moving to a larger core would have required all new manufacturing infrastructure, flight qualification, and unknown follow-on issues.
Tl;dr The F-1 was an awesome engine but not perfect or even desirable for all purposes and the "we can't build them" meme is stupid.
[0] https://www.thespacereview.com/article/3724/1
[1] https://en.m.wikipedia.org/wiki/Space_Exploration_Initiative
Also it is no longer known how to make the Saturn V rocket which flew to the moon - https://www.nytimes.com/1987/05/26/science/hunt-is-on-for-sc...
It's pretty interesting, but a lot of this stuff is structured specifically so that the contractor organization doesn't retain knowledge that it's "not supposed to have", as wasteful as that ends up being. Even stuff like a list of personnel who worked on a project or production line might be included... so that a few years after the project is completed, the management of the contractor might not even be able to figure out who to ask, because they've destroyed even their knowledge of the list of people who worked on the component's design and manufacturing.
Similar: See Dilbert, the cartoon version had a funny episode about company names through acquisitions.
[0]https://www.nasic.af.mil/About-Us/Fact-Sheets/Article/611728...
I though they already had enough practice reverse engineering the alien tech at Area 51?
I have no reason to believe that defense projects are not plagued by bureaucracy, office politics and incompetence like in a lot of government entities and large companies. If anything, it is worse because secrecy is a convenient way to hide incompetence. And when information is shared on a "need to know" basis and we realize that no one "needs to know", then it gets lost.
I have touched a bit on defense projects (not in the US) and generally, compared to civilian, open projects, if feels slow and primitive. I didn't get to the super-secret stuff but I have a feeling that the only parts where civilians don't do much better are the ones that only the army is interested in.
Which I thought he was joking at the time.
I suppose they could always call the engineers at Boeing or Lockheed who designed it in the first place. Like getting tech support for your Huawei networking equipment, just use the Cisco guide! [0]
[0] https://blogs.cisco.com/news/huawei-and-ciscos-source-code-c...
He also said a big problem was the expense of doing a modern engine upgrade. Anyone know if they ever re-engined it? Was he joking again?
Having one of these fly over my apartment on Rock Rd. was a dish rattler--I used to see B-2s every once in a while too (fairly loud too, IIRC).
[PDF] https://crsreports.congress.gov/product/pdf/IN/IN11413
I guess he was right about the engines.
The introduction of the B-2 in service was ~24 years ago, this is one of the most recent jet in the US air force. By military standard it's definitely young.
There are only 2 planes in service in the US more recent than that, the F-22 and F-35 (not counting transport aircrafts).
Those planes are designed for lifespans of 50-100 years.
The B-2 was started in the 70s and the maiden flight occurred while I was still baking in the womb.
I realize aircraft can live a long time and can be retrofitted and upgraded with modern technology - but all things considered it is old.
Not saying ancient is bad, either, but as far as organizations go, org knowledge, handoffs of info and books and writing etc... I am not surprised to see the need for reverse engineering.
No, they weren't "designed" for a century. Think about it for a minute - would a military contractor even admit it if they did? There would be no new programs. :)
The reason the B-52, F-16 (and T-38 for training, F-5 for aggressor training, etc.) are still around is because they were made in quantity.
By contrast, the F-22 had a run of about 190 airframes, and the B-1/B-2's dozens each, with only a handful (like 3 - 5) of each model operational at any given time - it's really that bad.
All they had to do for the B-52 was to update the avionics and add a cruise missile rotary launcher in the bomb hold. (Believe it or not, that is the main thing that keeps the B-52 in operations.)
B-52 Rotary Launcher Video https://www.youtube.com/watch?v=-hCjO43k-1Y
Do you remember where you read that? Sounds like an interesting bit of history.
The analogy to out-of-date software documentation is very apt. As problems or inefficiencies popped up in the factory, they'd change things and just not write them down (because of deadline pressure, often). If you wanted to build new ones you'd have to either get the factory workers to show you how the thing was actually built, or tear down a working example to reverse engineer it. (Which has been done!)
Due to the dispersed mass production in WW2 it's probably not totally true that they wouldn't fly without a custom tooling setup (it'd be a maintenance nightmare if two examples of the same model from different plants diverged dramatically, altho of course it happened) but you can be certain that it's of critical importance.
For mass production, they had to be re-tooled with larger tolerances which also allowed for interchangeable parts.
How things came to be and are, are often at times not documented. Dimension and tolerancing is lost over time too.
- Limited production run. These aren't Ford Mustangs or Toyota Priuses with hundreds of thousands and millions being produced. There were 21 built over 13 years. This also leads to the consequence that each one should be considered a bespoke creation and not a "copy" of the others (even ignoring the 20+ years of individual maintenance work they've each had).
- Time. It's been 21 years since the last one was produced in 2000. Whatever facility produced this has long since lost that capability.
- Aging workforce. Whoever designed it is likely retired, and could even be dead at this point. Certainly the senior engineers who may have been 40+ when the project started in the 70s/80s. Even if they weren't retiring and dying, the people on the project have been doing other things for 20+ years.
You see the real issue? Even if they have the blueprints, it will be difficult to even manufacture the parts here.
There's so many more problems like this coming down the pipeline. We've effectively lost (or rather acquiesced) almost all manufacturing skill to other countries. This was a result of the largest corporations taking advantage of labor arbitrage and then flooding their home market out of business.
If we really want to reverse this, we need to do what it takes to bring manufacturing back to the United States. It's a painful reversal of the open-trade policies of the past. Honestly, Trump was right to fight so hard for this, even going so far as to remove us from NAFTA.
I don't have a good strategy for this, but probably some set of minimum manufacturing of everything should happen within our borders, say 10%, and accomplishing that target with a combination of sticks and carrots.
This country is producing more Tik Tok stars than engineers that can sustain civilization.
SpaceX makes working spacecraft.
This country built Saturn 5 rocket, 50 years ago. Landed astronauts on the moon, with moon buggy and golf clubs.
Now we can barely launch basic satellites. How is this progress?
Which country built the components and systems and subsystems that go into manufacturing an electric car?
Which country has lithium, rare earth metal processing capabilities critical in manufacturing all modern electronic systems?
Which country built the components and subsystems in the vaccine manufacturing plants?
Decades ago, US had the education system to move millions of people through manufacturing and engineering pipeline to build all above things.
Now, that US education system does not exist. The manufacturing base does not exist. The engineering pipeline does not exist.
We produce more dropout Tik Tok stars than Professional Engineers that develop critical cooling cores, that can withstand heat and stress thresholds in modern military aircraft, and master welders with 20 years of experience that can weld specialty metal alloys in compact designs.
It’s not just aircraft. Do you know who works on water treatment plants that provide clean drinking water to millions of people in US? These guys will be retiring in few years. Who are going to replace them? Who are going to do the hard and difficult job of maintaining these critical systems?
https://www.propublica.org/events/the-flint-water-crisis-an-...
No amount of rosy thinking is going change the course of decline in this country.
> Manufacturing, assembly and integration operations are located at Decatur, Alabama, and Harlingen, Texas.
https://www.ulalaunch.com/about
Being unable to manufactures things we did in the past is extremely worrying
Its some of the little manufacturing the country has left
Older planes and other hardware have LONG had to be reverse-engineered (including microchips).
This still amazes me, even though every time a call comes in on the landline, the 50-year-old rotary-dial phone in my living room rings alongside the cordless phone. It doesn't get used much, except for the loud ringer and during extended power outages.
Hell, the building in hawthorne where dad worked isn't even owned by northrop anymore. Its spacex.
What are the odds someone said the phrase "those parts? I could mock up copies in a weekend!"
He also has "the laptop"—a mid-1990s Compaq running custom DOS software created by McLaren when the F1 was new. This gray brick was, for many years, required to access the car's engine control and body control modules. Today, McLaren uses a modern Windows computer running a software emulator for day-to-day computer maintenance. Hines keeps the vintage Compaq around just in case.
https://www.roadandtrack.com/car-culture/classic-cars/a14453...
CAD files (at least planar modeling as opposed to parametric or solid modelling) tend to have pretty straight forward data structures for geometry. Find the point or vertex struct and you're halfway there. The complexity is association and metadata which were nowhere near as bloated then as they are now.
As someone who works regularly for large asset owners, asset life-cycle information (digital and analogue) and the proper management of that information is often ignored because organizations pass that risk to the people working with the asset maintaining their own working knowledge to get their job done. We've had international Building Information Modeling standards since the turn of the century but there's often no business case or standard enterprise process to make leverage technology for asset management.
AI and ML are starting to appear more and more in the operations space, however many pilot projects have to start with digitizing the asset in the first place, which becomes a cost impediment for many getting off the ground.
It really frustrates me as there are great things that can already be done today but the initial step financially usually scares off anyone who has to know 5 years in advance what their operating budget is.
This country simply does not have enough skilled engineers to design and produce original parts to requirements. Let alone reverse engineer a worn out part.
B2 is product of its time and the amount of engineering resources that went into design and development simply can not be replicated with current society.
Tik Tok stars make multi millions of dollars, so why would any smart person spend years in school and solve complex problems to build state of art Stealth Plane.
Boeing can’t even produce a simple jet without having it fall out of the sky few times, 737 Max.
The flow of smart folks in the pipeline are dwindling. You may be working with smart folks, but how many are going to retire in few years? Who are going to replace them? Does the current education system produce the number and quality of candidates needed?
This country can not build a new reliable airplane, like it did decades ago.
It’s failed project after project. Ask yourself why this is the case? What has changed in this country? What has changed in the industry. What has changed in society?
It’s not just aircraft. Where are US high speed rail systems, that other countries built 30, 40 years ago?
Where are modern infrastructure?
Who works in water treatment plants, providing clean drinking water for millions of people? What happens when all these guys retire in few years?
Tik Tok is all fun and games, and no one wants to do the hard messy work of treating water for consumption.
The problem with F-35 wasn't scope creep, it was the initial premise which operated under a mistaken belief that getting a single “do everything” airframe to maximize commonality would drive down unit cost, rather than driving it up through the conflicting needs of distinct missions.
Nevertheless, these things may become more common if we are at the apex of our civilization.
I could imagine a RFP in Rome from 300 AD or so soliciting bids to figuring out how the hell to make an aquaduct, amphitheater, or even cement!
Obviously the answer is that it is legal to reverse engineer something if you own it.
That said... I think there's an opportunity here:
EFF and other pro-reverse engineering organizations should sue and demand that the DOD follow the same laws that citizens follow.
If they lose the lawsuit, it be great ammunition (so to speak) for stopping anti-reverse engineering laws.
If they win... bonus!
They'd have to have standing for such a suit and they have none. NG and others involved in the design and manufacturing might have standing, but they're also unlikely to sue. This is not a novel activity for DOD.
Or the contract allows it.
One of the proposals for future upgrades to SLS (ugh) was liquid-fueled boosters using new-build F-1 engines (i.e. Saturn V main engines). Unfortunately, the specs were not complete; not only was lots of stuff different in the factory than in the drawings, but parts of the engine were literally hand-built. Think API documentation built on a tight deadline, and published while software is still under development.
Luckily, some NASA engineers were already taking one apart and modeling the parts for previous SLS work. The new Dynetics version ended up much simpler and cheaper because of new manufacturing methods, but the reverse engineering process required to get there was a serious project in its own right.
https://arstechnica.com/science/2013/04/how-nasa-brought-the...
https://youtu.be/ZSRHeXYDLko
Included is an anecdote about Texas Instruments and loss of knowledge between generations of silicon hardware.
Unfortunately, they'd forgotten the manufacturing process, decommissioned the plant that made it, and the people that knew how to do it weren't around any more. They figured it out, but it doesn't necessarily look easy to do (nor does the material seem easy at all to make in the first place).
https://www.armscontrolwonk.com/archive/201814/fogbank/
p20 here says more: https://www.lanl.gov/science/weapons_journal/wj_pubs/17nwj2_...
Edit: I somehow missed this other thread in this story, despite a search: https://news.ycombinator.com/item?id=26334367
Things change, and the factory/tooling/people are long gone. Also, often the underlying tech isn't available. But the part needs to weigh the same amount, meet the same guidelines, and fit in the same hole.
It's cheaper to pay someone to spend the time to figure out how to make the replacement then to mothball the entire airplane—almost regardless of how expensive it might be.
[0] https://news.ycombinator.com/item?id=25788317