Given the relatively primitive state of transistorized electronics in the 1960’s, how could the computer(s) in a 60’s era MIRV type missile be able to accurately pinpoint multiple targets on the fly and send off it’s multiple warheads to specific longitudes and latitudes within the range of the fuel capability?
The Wiki article doesn’t really say. Multiple independently targetable reentry vehicle
IANRS (I am no rocket scientist!) but two things occur to me. First, my memory is that proper MIRVs - multiple INDEPENDENT re-entry vehicle - were not deployed until the seventies and early microprocessors were becoming available at this time (when I was studying electronics!); and second, the complex computing was not done in real-time. The calculations were probably done on a main frame and then hard wired (or at least set up in some sort of programmable logic) into the missile - manoeuvre, separate warhead, manoeuvre, separate warhead … Only limited computation is need to respond to feedback input from the inertial navigation system in the missile.
Hopefully somebody with more knowledge will come along shortly.
IANARS either, but my Russian buddy’s dad is a professor of math in St. Petersburg. This is purely anecdotal, but he was telling me that his dad developed some marvelously elegant rocket trajectory algorithms that proved to be as accurate as the super computer simulations. Theoretically, these algorithms could be implemented by a relatively primitive state machine or computer. Well my friend has very high regard for his pops, and Stranger will be along to tell me that I’m full of shit.
In the online historical reference materials on MIRVs (which are surprisingly sparse) it indicated the MIRV was designed in the early to mid 60’s, and came on line in the early 70’s, so assumedly whatever guidance technology was to be used was available in the 60’s.
The point I’m confused about is that to choose a target pattern, input the coordinates, and then have the missile respond would be trivial if the missile were pre-wired with logic boards for specific pre-set locations. However, my understanding is that MIRVS were supposed to be immediately targetable just prior to launch depending on the threat location, so I would think this requires a CPU, IO sub-system, memory etc. that can calculate speed, height, longitude and latitude input etc. and make flight corrections and adjustments when fired. The 60’s computers I’ve seen that might be able to do this would take a space the size of up a small conference room.
Am I mis-understanding the nature of how these missiles were targeted?
The computers of the late 50’s, early 60’s were perfectly capable of doing the calculations required, and fairly quickly – not as fast as today, but software dedicated to simple lat/long/azimuth/distance problems would return answers very quickly.
We used those computers in flight simulators, for similar calculations (bombing, tracking) and they worked great. These were Univac 1205 and 1230’s, refrigerator size computers. The computers were big and heavy, but considering the weight and cost of the weapons and the launch vehicle, putting a huge, expensive computer onboard would not be out of the question.
I think you’ll find, however, that they put purpose-built, custom-fitted, dedicated-task computer systems on board each missle, as part of the inertial-guidance system. This would be the ‘guidance computer’.
In the 60s? What, built out of vacuum tubes? I find this highly unlikely.
MIRVs used analog computers? I find that difficult to believe.
Keep reading the article until you get to the part about electronic analog computers. No moving parts required.
ETA: I’m not saying that MIRVs used such. Just that not all computers were composed of transistors or vacuum tubes.
The Minuteman II, fielded in the mid-60’s, was one of the first practical applications of integrated circuits.
Yeah, I’m such an analog retard that I didn’t even consider analog computers. However, wouldn’t analog computers of the '60’s contain transistors in order to build the op amps?
I imagine that they’d have the various targets precalculated, after all, gunners of that era had precalculated tables for use in aiming their artillery. So it seems to me that what they would have done is grabbed the book with the necessary codes, punch those into the computer and launch the missile. Now, if they had to target something that they hadn’t previously crunched the numbers for, it’d take some time, but I can’t imagine that we wouldn’t have the necessary data for the various major cities in the USSR not compiled in advance.
I might be mistaken on this, but accuracy wasn’t all that high in that era. Certainly, it wouldn’t have been measured in meters as it is today, and with the large warheads of the time, just getting it in the right ballpark would have been “good enough.”
I know that the computers used on the LEM, were really little more than automated sequencers (and stored data on a wire), and that if you threw too much at them, or if the data was too far out of line with what it was told to expect, they were utterly useless. I’d imagine that the technology used for guidance on the missile systems was pretty similar.
And don’t underestimate the power of tube based computing for years after the US military switched to solid state electronics, the USSR kept on using tubes. There’s an account of US engineers pouring over a captured Soviet fighter and laughing that it still had tubes, until one of the engineers pointed out that the plane could survive an EMP with no problems, while a US fighter would have had it’s electronics fried.
Of course, the USSR, had some pretty weird non-electronic computers, as well.
I know an analog EE who thinks he was probably the last person ever to design an analog missile guidance system. I’ll try to remember to ask him next week if he can say when that was. But I’m pretty sure it was well after the '60s. He was doing antiaircraft stuff, though, not ICBMs.
He’s a cool guy. He’s got some funny stories about missile tests and whatnot.
So the missile is lying there on the ground, and the guys in the armored truck reach out and close their hatches. About that time, the VIPs in the grandstand start to think it’s a good idea to get out…
The Minuteman I used a spinning magnetic disk as the primary storage for its guidance computer. It had no RAM or other electronic storage. Supposedly this made it more resistant to electromagnetic pulses from nearby nuclear detonations.
MIRVs were basically ballistic vehicles. The firing solution was not dynamic, in the modern sense of the word. Multiple solutions were available, given the original trajectory of the missile, but the re-entry portion was not adaptive in any sense of the word, after launch. The heavy lifting, computation wise was accomplished long before the fact, but even that was done on vacuum tube computers. (oh, the barbarity!)
The small scale integrated circuit was available to the military in the 1960s, and had been expected since the fifties. Early missile guidance systems were among the first applications using large numbers of such circuits. A specific chip for each target set was one method of “uploading” targeting information to the missile guidance system.
The target circle of the ICBM of the sixties, and the seventies was a Kilometer or more in radius. Hence, the drive for bigger and bigger warheads. The size of the explosion was what made the weapon feasible at all. Missing by half a mile was probably not enough to leave the target safe.
Tris
Hmm. “Safe” was a very relative word in those days. I remember the evacuation plan for our part of Austin had everyone driving out 71, right past Bergstrom AFB. Didn’t seem safe to me.
Duck and cover. Yeah, right.
Almost only counts in three things.
- Horseshoes
- Hand grenades
- Atomic weapons
Well, it all depends on what you’re worried about. If you’re afraid that people outside ground Zero will be turned into radioactive mutants, then the safe thing to do is to make sure you get as many of them in the impact point as possible. 
And in fact the primary impetus for developing production grade ICs, although this was more from the standpoint of reliability and nuclear survivability than raw calculation power.
MarcusF is correct that true MIRV (Multiple Independently-targeted Re-entry Vehicles) were not deployed until the early 'Seventies with the American LGM-30G ‘Minuteman III’ and Soviet R-36M (NATO: SS-18 ‘Satan’) for ICBMs and the American UGM-73A ‘Poseidon C3’ and Soviet R-29R (NATO: SS-N-18 ‘Stingray’), the last actually deployed at the end of the 'Seventies. Previously some systems (Polaris A3, R36/SS-9 ‘Satan’, R27/SS-N-6 ‘Serb’) had used a MRV (Multiple Re-entry Vehicle) bus; however, unlike MIRVs these weren’t designed to hit independent targets but rather to allow for greater dispersal and to compensate for the high CEP of then-modern guidence systems, especially for mobile launch platforms like submarines. (The CEP, or Circular Error Probable–briefly, the radius about a target in which an individual RV would land half the time–was measured on the order of miles for these vehicles, whereas later systems would measure CEP in terms on the order of hundreds of meters or less.)
The basic technology that allowed for the use of smaller, highly accurate warheads, thus making MIRVs practical, wasn’t just (or even primarily) computers, though. A compact analog computer of the day was more than capable of computing the trajetory to release of multiple RVs (which is really not much more difficult than computing for one RV; you are, after all, just continuing along the same ballistic track); rather, the necessary developments were threefold:[ul][li]More accurate solid state inertial measuring units (IMU) combined with secondary stellar navigation correction, []higher throw weight boosters and post-boost rocket maneuvering systems for independent payload deployment, and[]high speed errosion resistant RVs that could more reliably be characterized for calcualting ballstic target trajectory.[/ul][/li]More compact and reliable computers were useful, certainly, and allowed for much faster retargeting, but they’re hardly the most important element in stationary silo, purely ballistic missile guidance, which, after all, mostly relying on Sir Isaac Newton and a couple of coordinate transforms for the bulk of the trajectories. (They are, however, essential for real-time embedded systems that allow for modern fly-by-wire avionics in unstable aircraft, self-guided rockets, and maneuverable re-entry vehiles (MaRVs)). Remember, the first Gemini flights (up to G-VIII and subsequent flights that used the Gemini Guidance Computer, a forerunner to the AP-101 used on Apollo and STS) didn’t even have digital computers, and the Mercury capsules had no real computer at all, relying strictly on a the pilot using analog instruments backed by a dumb analog autopilot and auxillary ground control, and these were brought back from orbital space, using blunt-body aerobraking re-entry,managing (with a couple of exceptions) to land within a few miles of the splashdown target (sometimes uncomfortably so). So, I’ll take Darryl Lict’s buddy’s father’s claim as being only very slightly hyperbolic if at all.
Stranger
Univac 1206. 1958.
Univac 1230. 1961.
The first flight simulators I worked on were, with the exception of some lat/long/azimuth/altitude calculations, exclusively analog computer systems. Thousands upon thousands of gears, servos, cams, microswitches, and specially-wound potentiometers.
For example, the airspeed ‘module’ had about 18 different potentiometers, 8 or so cams, and 20 some microswitches, mounted on 3 or 4 gearshafts. All possible calculations for airspeed were performed by this single analog computer. There was another module for altitude, yaw, angle-of-attack, etc.
Hugely complex, and heavy, but it’s not outside the realm of possibility that the purpose-built, custom-fitted, dedicated-task computer that acted as the guidance computer for early ICBM’s was fully analog. All logic functions and mathematical formulae can be replicated mechanically.