Here’s what I have for my space safety paper:
ICBMs were never designed to be highly reliable, because to do so would have dramatically increased their costs (many hundreds of them were built), and it wasn’t necessary for their mission. They were designed to be launched in massive numbers, and if a few out of a hundred didn’t make it through, that was all right, because they were often redundant in their targeting (that is, more than one missile would be aimed at a key target). Some estimates at the time of the reliability of the Titan II was only 80% or so (that is, one in five would not deliver its payload to the designated target), based on the fact that eight of its initial thirty-three test launches were failures. The early manned spaceflights were performed on modified versions of them (specifically, the Redstone and Atlas for Mercury and Titan II for Gemini). But what was good enough for a weapon as part of a fusillade of dozens or hundreds wasn’t perceived to be for a single flight carrying a human, particularly with recent memories of nationally televised ignominious failures of rockets on the launch pad. Thus was born the pernicious (and now obsolete) concept of “man rating,” which confuses the space industry and obfuscates policy down to this very day.
Is there anything inaccurate in that?
Clear as a bell and dead smack on target.
I have a question. What does it mean to say that an ICBM would miss its target? Does that mean it would explode somewhere else? Or just fail altogether? I assume adequate safeguards exist to prevent it from detonating near launch, but what about closer to the target?
Rand,
First, if I recall my history correctly, the reason they accepted a lower reliability was time, not money. The US wanted to close the “missile gap” with the Soviets as soon as possible and was willing to trade off reliability on order to deploy the systems faster. And since it was national defense, money was no constraint, as with Project Apollo. (Waste everything but time…)
Plus they still had the old gunnery officer mentality, (or old bomber pilot if you will), the more shells/bombs on the target the better the odds of destroying it so it wasn’t seen as that critical a tradeoff.
Second, to play the devil’s advocate. You argue that “Man Rating”came about because the early ICBM’s did not need to be reliable as that could be countered with multiple launches. But isn’t requiring a vehicle to be built to a certain level of reliability just “man rating” under a different name? After all, that was the basic function of “Man Rating”, to increase the reliability of systems that were originally built with a design philosophy was to trade off reliability for other factors regarded as more important.
IIIRC, part of the “man rating” of those early missiles was to reduce the failure rate but another part was to properly instrument the rockets to give some warning of failure early enough to use the launch escape system.
But isn’t requiring a vehicle to be built to a certain level of reliability just “man rating” under a different name?
In a sense, but that’s the point. The EELVs were designed to a 98-99% (I think) reliability rating, yet modern critics say they can’t be used as manned launch vehicles because they’re not man-rated.
I think Rand’s point is that man-rating made sense when taking an 80% reliable vehicle up to 98% reliability but loses its meaning when the launch vehicle is designed to 98% reliability in the first place.
Michael,
But isn’t the “Man Rating” of the EELV simply adding in the sensor systems needed to determine if a launch abort is necessary or not? Or are there other more extensive changes needed related to other issues? Does any one here know?
In any case since they were the launch vehicles intended for OSP the issue of using them for HSF shouldn’t be major.
I agree that adding the sensor system should be all that’s required for man-rating the EELVs, but it is apparently more than that. During the ESAS days I saw reports of NASA estimating that man-rating the Delta IV Heavy would cost $21-$24 billion (for man-rating a system that only cost $500 million to develop in the first place). Supposedly it requires complete redesign of the RS-68 engine and major changes to the structure of the common booster core.
It is also alleged that demonstrated reliability has no bearing on the man-rating requirements.
If what you’re saying is true (and I’m not disputing you), then that suggests that the “man-rating” process is completely bogus. Why exactly would a flight-proven engine have to be redesigned? Why would demonstrated reliability be ignored? Because the bureaucrats say so? When was the last time NASA flew a fully “man-rated” system?
Rand,
Here is a declassified memo that may provide you a definitive reference documenting redundant targeting as a policy. Specifically, see para 8 (“duplications”). I hope this helps. Link: http://www.gwu.edu/~nsarchiv/NSAEBB/NSAEBB130/SIOP-2.pdf
If you are referring specifically to ICBM’s of the 50’s and 60’s you are accurate. Beyond that, both reliability and accuracy became increasingly important, especially in the context of targeting hardened silo’s, etc.
If you are referring specifically to ICBM’s of the 50′s and 60′s you are accurate.
I am. I’ll prepend the word “Early” in the first sentence, though, to make it clear.
Rand,
Another note. As Astronautix noted the Titan II (used on Gemini) success rate is listed as 92.59 %, not 80%.
http://www.astronautix.com/lvs/titan2.htm
The first generation Titan I by contrast did have a success rate was 78.57 percent.
http://www.astronautix.com/lvs/titan1.htm
Also note this statement from Astronautix on the Titam II.
[[[The Air Force authorized Martin to proceed with the Titan 2 ICBM in June 1960. In January 1961 the Titan 2 was selected to boost the Dynasoar manned spaceplane on suborbital tests, and in July 1961, to orbit the manned Gemini spacecraft. Therefore during development the ICBM had also to be man-rated. This caused some difficulties, as a pogo problem encountered in flight tests was acceptable to the Air Force for a weapon system, but unacceptable to NASA for a manned booster. ]]]
http://www.astronautix.com/lvs/titan2.htm
So Man Rating was not done after the fact as with the Redstone and Atlas, but was part of the design process, which is why the operational Titan II missiles likely had such a major jump in reliability over the Titan I missiles.
Note especially that the pogo problem which was acceptable as a weapon delivery platform would create problems with astronauts on board, so even the reliability figures by themselves would not tell the whole story.
This is always been a personal opinion of mine in regards to atomic weapons. The value of them in WWII was because the damage radius made up for the lack of accuracy in bombing techniques and equipment. Now that are accuracy is near pin-point in reality; the need for atomic weapons has moved from a military weapon to a diplomatic weapon.
I don’t think this matters, but your paragraph opens specifically about ICBM’s and then mentions Redstone, which was only a short range ballistic missile.
I guess the main comment I would make is that no one knew how to design an ICBM for reliability in the early days, but they quickly realized that operational readiness was of paramount importance. Hence the evolution from the LOX/kero Atlas and Titan to the storable liquid Titan II and finally to Minuteman. They found all-solid rockets cheaper to produce, and built a slew of them to guarantee success. Along the way, they discovered that solids were also reliable, and figured out how to make them more so.
By the time Peacekeeper came along, they knew how to design for both life and reliability. PK had a design reliability of more than three nines, though I doubt that it ever demonstrated more than 98%. But that was good enough to restrict total force size to 50 missiles (down from the planned 100).
With up to 10 MIRVs delivered with astonishing accuracy, and reasonably high reliability, it was able to replace the 54 Titan IIs and their 9 megaton missile base busters and give us better capability.
Atlas never became a reliable ICBM, but did become a reliable space launch vehicle. Titan never became either, while Titan II became both.
Atlas never became a reliable ICBM, but did become a reliable space launch vehicle. Titan never became either, while Titan II became both.
After the Titan Is were withdrawn from ICBM service, they were mostly scrapped with only a few surviving in museums. I’ve seen no evidence that any of them were ever used for anything else. Does anyone know why? Were they that unreliable? You’d think they could’ve used them for test vehicles or for launching lower priority satellites but they didn’t.
That’s an interesting point, and I’ve never thought about it before. The Atlas became a satellite-launching workhorse, but the Titan I didn’t. Why?
Atlas had been converted to space launch in 1958, for Project SCORE (18 December 1958 first orbital launch). Titan first flight of any kind was 5 February 1959 (success), followed by 69 development and operational test flights. Out of 70 flights, 17 failed. Titan had the same orbital payload capacity as Atlas, so it would have offered no advantage. By the time Titan was retired (1965), Atlas was both an established space launcher, and a retired ICBM which had been built in greater quantity than Titan. It wasn’t cost effective to refurbish both the Titan and its ground support equipment just to get the same payload as Atlas carried — especially since the retired Atlas existed in such numbers that we didn’t run out until 1995.
It always puzzled me about the amount of refurbishment ICBMs (Atlas and Titan II) needed before being used for space launch. They decommission the missiles, remove the warhead, then spend a lot of money on refurbishment. If they were good enough to carry a whopping big nuke (9 MT in the case of Titan II), you’d think they’d be good enough to carry a satellite.
I believe that at least one was used to test whether a first stage could drop into the ocean, and be reused again with little clean-up. It worked for them, that time.
In other things, the Titan 2’s performance so eclipsed the Titan 1 that it was basically shoved under the carpet. It was the Martin Company’s ICBM learning rocket, and the experiences there served them well in designing the Titan 2. In spite of appending the series name, perhaps to get Congress to buy in as a small change to Titan 1, the Titan 2 ended up with about 85% new parts, from the engines to the guidance systems to the warhead.
I don’t think space industry is confused by the concept of man-rating, but they know that the politicians currently funding the lion’s share of manned space flight are, and that affects their proposals and arguments.
It seems to me that the different operational environments of ballistic missiles and launch vehicles must have played a role in reliability as well. A launch vehicle can sustain a three-day countdown filled with all sorts of system checks and analyses, but an ICBM can easily be in a use-it-or-lose-it situation. Better to get 100 missiles out of the silo pronto and lose 20 of them to failure than perform a long, drawn-out countdown and lose all 100 of them to counterstrike.
More or less, this is correct. What’s a bit misleading, as I recall, is while the AF gave the Titan II an 80% reliability rating,
it didn’t necessarily mean only 80% of launched Titans would reach their final destination and detonation point. Rather, the notion
was that 20% of Titans would fail to launch, period, simply because the missiles required a great amount of maintenance. After that,
the notion was that ALL launched ICBMs would probably reach their targets and release their warheads.
Sorry, I don’t have documentation to point to, but this is how it was presented to those of us in freshman AFROTC back in 1964.
I should add, the working assumption was that 100% of aircraft launched would reach thier targets and drop nuclear weapons — the
important feature of missiles was that they reached their targets in minutes rather than hours. I should also add, in retrospect,
that these figures were DOCTRINE rather than settled fact; I think, looking back, that the captain teaching my course had unvoiced
reservations.
A wise man once told me that an example of a rocket that was not man rated was a rocket with 8g of thrust from a SRB.
I’ve read that during phases of a Titan II/Gemini liftoff, they pulled up to 8Gs. I think it was shortly before second stage shutdown. The Mercury/Atlas rides were pretty harsh, too.
Astronauts were made of sterner stuff back then.
Neither the Redstone or early Atlas missiles were classified as ICBM’s. They were IRBM’s.
Grammar: “at the time of the reliability” is kind of confusing. How about “Some early estimates of the reilability of the Titan II WERE only 80%.” Also, “were” not “was.” “Estimates” is plural.
I’m not an engineer, but it makes sense to me.
“…that was all right, because they were often redundant in their targeting (that is, more than one missile would be aimed at a key target)”
I know why you wrote that AND the paranthetical, but how sad is it that you must explain it like that?