Why making military tech so hard?

I’m pretty sure it landed in the ocean, so no.

Why do you think anyone would be able to develop 1950s US missile without much effort when no one, including the US and USSR, were able to do so without much effort? It cost a lot of money and time for the US and USSR to make working missiles, and missiles are mostly highly specialized equipment, you can’t just buy off-the-shelf stuff today. There’s a ton of engineering work needed for the missiles, and while Iran is not a basket case like North Korea, they’re not remotely in the same league as the US and USSR for technology.

One huge question is Iran would WANT to develop 1950s missile tech - ICBMs are barely even 1950s tech (US and Russia didn’t deploy any until 1959). And what was deployed in 1959 were unreliable missiles that could not be kept ready and required a huge, vulnerable facility. Building 1950s missile tech would piss off the US, Russia, and Israel, all three of whom have conventional aircraft within range to easily bomb such a facility off the map. NK makes a habit of pissing off other countries, but Iran is not in the same situation and really doesn’t want to reach NK levels of hostility with the rest of the world.

I think there is also a curious perception that our technology has advanced greatly in the last 50 years. On a great many fronts it simply has not. Computers is where the action has been. And an extraordinary amount of the advances we see in other areas have been built on the back of available compute power. But in any area there are intrinsic limits on what can be achieved and how hard it is to do something.

50 years ago the the US was sending man to the moon. It can’t do that today. It is re-creating, at large cost, a capability that in many ways is mark-2 of that same 50 year old capability. Some of that work has involved taking a 50 year old rocket engine design (the J2)and reworking it (into the J2X).

We take for granted some astounding advances in digital electronics, and seem to assume that all technology has advanced just as fast. My favourite example is the 747. The first 747 flew about 50 years ago. Today, it remains one of the most common sights in the sky for international travel and Boeing still make one. The only reason it has fallen from favour is that the economics of how air travel works now favour large twin engine planes. But if you step onto an international flight, the plane you fly in will very likely still be a 747. Back when the first one flew, computers were building filling behemoths made from discrete transistors. now you have vastly more compute capability in your phone.

Building an ICBM is no easier for the US than it was 50 years ago. It needs just the same level of care and effort. It is no longer bleeding edge. We know it can be done. But it remains very difficult. As noted earlier, a big part of what enables it to be done at all is access to an industrial base than can supply precision machinery, metalergy, electronics, and the like.

Your phone has a navigation capability because the US spent rather a few billion lofting the GPS system. Civilian GPS receivers are designed to stop working if they are travelling too fast, with the explicit intent of preventing them being used in a missile. It is possible to work around this, but it isn’t trivial. An inertial guidance platform is not an easy thing to build. However if your only intent is terror, rather than specific targetting, the requirements are lessened.

The main thing is that a lot of tech is conceptually simple but requires almost unbelievable levels of quality control to work properly. Take for example a solid rocket stage. To a first approximation it’s a giant firework: a metal tube filled with a solid fuel/oxidizer mixture. Simple? No, actually. For starters the fuel segments must be made to near-perfect standards of composition and uniformity, and be cast without any defects whatsoever. The burn rate has to be built into the propellant itself- a variable burn rate as the propellant is consumed, and it has to yield a thrust exact enough that a rocket that will travel thousands of kilometers will insert its payload into an extraordinarily narrow window of space, time and velocity. If you don’t do it right, either your solid rocket blows up or your payload goes wildly off course. It’s worth noting that the Soviet Union continued to rely on liquid-fuel ICBMs and SLBMs long after the USA developed the Minuteman and Polaris missiles.

Without the necessary level of quality control, you’re like a blacksmith trying to build a bicycle by hand.

As someone who knew pointed out to me:

Often, the manufacturers are both civilian and military and treat any project with similar safety levels as if it was military; while their systems are in development, they’re as much of a closed box in both cases. It isn’t until they become available for sale that you can get hands on them.

One of my coolest projects was a process analysis for an aeronautics company that needed to have some parts subcontracted out; they needed to make sure that everybody got the necessary documentation in the appropriate formats, the appropriate materials with the exact specs, and at the same time that nobody got anything they didn’t need. It was real neat (we got to see what would be the whole process for how to design and tweak a new product, from idea to production), plus the factory was full of airplane models, airplane posters, airplane… :slight_smile:

Well, we hope the booster section of an ICBM doesn’t explode during operation; that would definitely be…problematic. The point is well taken, though, that with rocket launch vehicles, you are rarely able to recover the vehicle for later analysis and almost never in an intact state. That leaves you with just telemetry from flight–which is limited in both bandwidth and fidelity–and ground test simulation to correlate to whatever analytical models and estimates you have of the vehicle performance and the loads and environments that the vehicle and its payloads are exposed to. Testing is expensive, often difficult, and ground testing can never fully simulate the combined environments (thrust load, temperature, near vacuum or plasma, vibration and shock, et cetera) that are seen during flight. So, it’s hard to develop a new vehicle without prior experience with the design, and that experience is typically gained through previous failures. Witness that SpaceX had three sequential failures of their Falcon 1 vehicle before they got a successful launch.

However, there are a couple of even more fundamental physical limitations to rocket launch vehicles that prevent the performance from improving on the kind of power law that microprocessors or software systems have advanced in the same period. The first is materials; we’re still limited in the materials that we use to build both rocket launch vehicle structures and the propulsion systems that power them. Sure, we’ve had incremental improvement in strength, corrosion resistance, high temperature limits, et cetera, but ultimately there are limits in strength that are fundamental to the materials that we use; no alloy of aluminum is ever going to give a 200 ksi yield strength and an operating temperature of 1500 °F. Fiber wound and broadcloth composites were once thought to overcome the mass ratio limitations of metallic structures for the rocket primary structure, permitting adquate performance for single stage to orbit (SSTO) or spaceplane vehicles, but it turns out that composites are really difficult to manufacture without flaws, undergo progressive damage over their operating life, require reinforcement in joint areas and heavy “fill & drill” repair when they delaminate (as they inevitably do), and are very sensitive to environment conditions. At best, they offer modest improvements in weight perfomance at high cost. Until we have some kind of technomagical material that is simultaneously stronger than steel, lighter than carbon fiber, and can withstand temperature better than Ni-W-Ta superalloys, we aren’t going to see revolutionary advances in the mass performance of launch vehicles.

Similarly, we are limited by the energy sources we have to power a propulsion system. All ascent propulsion systems are currently powered by hydrogen- or hydrocarbon-fueled engines carrying their own oxidizer (typically liquid oxygen, nitrogen tetraoxide, fuming nitrix acid, or high test peroxide). These have limited energy per unit mass, as well as limits on the maximum temperature achievable through combustion at a given chamber pressure, which itself is limited by the material strength and temperature limits of the chamber pressure, although we overcome the temperature limits by various means of active cooling such as evaporatie film, regenerative chambers, or ablative chamber walls. Most modern high performance rocket engines achieve within a couple percent of the maximum power output for a given fuel and oxidizer combination, even though the energy efficiency (as measured by the waste heat of propulsion passing through the nozzle exit plane) is not very good, and the higher the propulsive performance typically the lower the energy efficiency. Using chemical combustion, we will never see much higher performance than we get now; there are some very exotic fuels that give somewhat modest improvements over RP-1/LOX or LH2/LOX, but their hazards associated with handling and use outweight the small improvements in performance they offer. Going to a detonation regime instead of combustion offers a significant jump in both thermodynamic and propulsive performance by as much as a factor of two or slightly more, but until we have some kind of revolutionary power source that is compact and can through propellant much faster, combined materials and methods for dealing with much higher temperature in the chamber and plume, we’re stuck with the relatively limited performance we currently see, and no amount of engineering or testing is going to change that.

For the same reason we’re worried about a bunch of Islamic extremists whose major victory in the last couple of decades was hijacking poorly secured airliners and crashing them into landmarks that were known and unprotected targets, while vastly more people die every year in car accidents, lack of basic medical car, homicide and domestic abuse, et cetera than died in the World Trade Center and Pentagon attacks. It makes good theatre and justifies spending tens or hundreds of billlions of dollars to develop systems and fight ‘wars’ and maintain giant, job-creation bureaucracies.

Although this wasn’t the point of our argument, but the Global Positioning System is a great example of a nearly magical technology that seems commonplace in commercial use today but only exists because of tens of billions of dollars spent developing, deploying, and maintaining the systems for military applications. It is vastly more expensive than any nation smaller than a major superpower could afford to produce, much less a struggling, essentially Third World nation like North Korea or Iran, and definitely beyond what a commercial company could every afford to invest in development. Similarly, a nation like North Korea can’t afford a development effort for a ballistic missile booster with dozens of developmet test flights or the ludicrous amount of money on developing the requisite high precision guidance systems and reentry vehicles capable of surviving the erosivion of the hypersonic reentry environment in a predictable fashion.

In other words, every part of making a launch vehicle and strategic missile weapon system is hard, and all parts have to work together for the system to be successful. Validation and verification testing requires both expensive ground testing of individual components and subsystems, and ultimately repeated flight tests to be assured of adequate reliability or to uncover design flaws or discrepencies between analysis and reality. The fact that a few nations have managed to develop systems over the last five decades doesn’t make it significantly easier for other players unless they can directly access and use that experience. Even being given design information or flight articles to reverse engineer isn’t as big of an advantage as it may seem; there is a lot of tribal knowledge and lessons learned that goes into turning a paper design into a produceable item, and a critical mass of manufacturing and materials technology to support high quality production of flight items. You don’t get this from specifications and drawings; you get it from people, processes, and qualified suppliers.

Stranger

Because they are possibly crazy enough to use it, and definitely aggressive enough to use it as a threat. The former makes them dangerous, and the latter makes them good news copy.

If, say, China started threatening to nuke its neighbors they’d make the news too.

Nah. North Korea and Iran both want to demonstrate a credible nuclear threat capability so that people will take these otherwise backward and incompetently managed despotic governments seriously. In the case of North Korea, they’ve been very canny about using the promise of not developing their presumed nuclear capability as a bargining chit for needed foreign aid.

Stranger

It occurred to me that SpaceX and even Blue Origin have tech MUCH superior to anything the DPRK can even dream of having.

That’s the difference between having easy access to some incredibly intricate technology - lil Kimmy has to make do with rivets in high-strength Al (don’t laugh) when USA civilians can get welded skins of fantastic alloys.

Kim is much like Trump - how much of the demonstrated insanity and wild threats is real and how much is an act?

There is a KPA meeting shortly - the first time this group has met in 30-some years.
Presumably, Kim has something Special to reveal. There is speculation that he will announce that all these demonstrations of Military Might have reduced the US and its cronies to quivering gelatin so we can now address quality of life and spend money on food production, education, etc.

I really want to see what the purpose of this confab turns out to be.

As to PRC - Kim went to Moscow for Vlad’s poorly-attended party.
He has NEVER gone to Beijing to pay his respect to Xi.

Yeah, China just might be pissed.

And on top of that your house is surrounded by wealthy, powerful neighbors who are concerned about you building a car because you have told everyone you’re going to run them over. If they catch you working on a car they may stop you from going to the grocery store, cut off your phone & internet connections, randomly replace your steel ball bearings with zinc ones and stick a banana up your (car’s) tailpipe.

Early missile guidance systems relied on on “dead reaconning”, and used very primitive versions of acceleratometers and gyroscopes in a modern smart phone. They would serve perfectly well, if your only interest is to get your nuclear warhead in right city.

The IMU in your phone is tiny, cheap, and low-power, but otherwise inferior to those “very primitive versions”. See, for example, slides 6 and 8 of this presentation, or get the datasheet for a consumer-grade MEMS accelerometer and gyro, and integrate the specified typical error forward over the trip from North Korea.

Very interesting, thank you. I sort of knew that a phone accelerometer wasn’t that great, given how janky games that use tilt/motion control are. However, I didn’t appreciate that it was so many orders of magnitude inferior to what you need for missile guidance.

How good are the best off-the-shelf accelerometers available without restrictions due to ITAR or other munitions export controls?

The ITAR regulations cover a relatively narrow range of items, mostly (but not always) items with limited civilian applications. They’re what forbids, for example, selling a fighter jet to China. The EAR regulations cover a much broader range of items, including many dual-use technologies. They’re what forbids, for example, selling a high-performance computer to North Korea (but maybe not to China).

The EAR regulations call out certain limits on scale, offset, and other errors; so they answer your question exactly. I don’t know about ITAR.

For North Korea, that’s perfectly adequate. In fact, if the goal is EMP wipeout of 10% of America, wouldn’t the required accuracy be even less? “Detonate at 500km altitude over New England.”

The accelerometer and gyroscopes in a modern smart phone are flimsy and inaccurate crap compared to what’s in even a 1950s missile. Completely worthless for guiding a missile onto target, they’re too inaccurate and probably can’t survive the trip. I assumed you were talking about GPS since it is something that’s better than a 1950s missile (though I think civilian GPS is designed to fail at ICBM velocities).

One of the problems with accurately assessing how hard it is to build something like a rocket is that people assume that cheap, available modern consumer-grade products are massively superior to custom designed military tech from the 1950s when it’s often just not the case.

North Korea has neither the range nor the reliability to launch a missile over Eastern North America, and I’d be surprise if they could reach the Pacific coast even if their booster didn’t come apart like a cheap gold watch.

As far as a regional or continental-wide electromagnetic pulse, it is no where near as easy as many people think. Getting a wide area pulse requires using a radiation-enhanced high yield weapon optimized for high energy x-rays in a specific bandwidth, detonated at a fairly precise altitude and latitude. And although such a pulse could be potentially devastating to industry and commerce, it would in no way forestall the inevitable response, which would likely consist of some impressive number of ICBM and SLBMs targeting Pyongyang and every strategic military base and weapons production facility in the northern part of the Korean peninsula, followed by an internationally supported embargo and blockade, with China leading the way in the “Fuck you guys!” contingent, wiping their hands of any obligation or responsibility.

North Korea seeks to possess nuclear weapons and delivery systems because it gives them a seat at the big boy table, not because they have any real intention of using them. Accidents could still happen, but deliberate intent is not the hand they have shown to date.

Stranger

I’m not sure about the “probably can’t survive the trip” part. A typical consumer MEMS IMU is rated to survive ~10000*g, presumably because the acceleration when a phone gets dropped onto a rigid surface is very high. A missile shouldn’t experience that until the very end. Otherwise agreed, though.