Mike
Excellent history. I worked in aircraft acquisition for 15 years and had
similar observations. Do you know of any reports, books, etc. going into more
detail.
Thanks,
Rick Wills
From: arocket-bounce@xxxxxxxxxxxxx <arocket-bounce@xxxxxxxxxxxxx> On Behalf Of
Michael Kelly ("mskellyrlv")
Sent: Friday, 24 March, 2023 9:38 PM
To: John Stoffel <john@xxxxxxxxxxx>; arocket@xxxxxxxxxxxxx
Subject: [AR] Re: Relativity Space launch failure
John:
The Small ICBM (or Small Mobile ICBM) was the next ICBM development project
after Peacekeeper. That was the last time TRW (the System Engineering and
Technical Assistance contractor to the Air Force Ballistic Missile office) was
able to sell a program to BMO, and then leverage it up the chain to become a
real programI worked for TRW, where I had started at the beginning of the
Peacekeeper Full Scale Engineering Development effort. Peacekeeper was the
most successful strategic missile program in history - anybody's history.
Small ICBM was the least.
SICBM was to be a single warhead intercontinental ballistic missile to
specifically be launched from a mobile platform. Mobility was thought to be
the key to survivability against the kind of massive attack the Soviet Union
could unleash against our fixed silo fields. As an aside, proof of that idea
came during the Gulf War, when Iraq decided to launch SCUD missiles against
Israel. SCUDs are mobile, and though we had every kind of tracking asset you
could think of to figure out where a particular shot had originated, we never
found a single one.
Anyway, the seeds of SICBM's doom were sown by a caveat in the Congressional
authorization for the program. The bill called for development of a mobile
ICBM carrying a single Mk-21 warhead to a range of 6,000 nmi; all fine
requirements. But then Congress put in its own caveat: the missile had to
weight no more than 30,000 pounds. That was a strict, not-to-exceed
requirement. Nobody at BMO/TRW could figure out where it had come from; we
only knew it wasn't from us. The only explanation I've ever heard, probably
apocryphal, was that one of the Congressmen asked his son, who was a
self-styled rocket expert, how much such a rocket should weigh. His kid's
answer was 30,000 pounds, and we were stuck with it.
Another factor that led to SICBM's failure was the military-industrial
complex's culture involving how to spread the spoils of the program. The
spoils dogma included the absolute requirement to have three booster stages.
Why? Because there were three solid propellant rocket houses, Thiokol,
Aerojet, and Hercules, and they all had to have something. (United
Technologies, which made the Titan III and IV zero stages among many others,
was a complication. But they and Aerojet were both California companies, and
only one needed to win to keep that state's congressional support). So, like
the Peacekeeper and Minutemen before it, Stage I went to Thiokol, Stage II to
Aerojet, and Stage III to Hercules. The Post Boost Vehicle (PBV), a small
hydrazine system which carried the guidance system and warhead, and provided
the precision state vector for the required accuracy, went to Martin Marietta.
Martin's bid was a lowball, and the booster companies followed suit. The whole
missile was underbid from the beginning.
This was a case where the military-industrial complex shot itself in the foot.
TRW had done all of the system engineering and requirements definition, and my
department (Propulsion & Ordnance Engineering) in particular had conducted the
optimization studies. The optimum configuration for SICBM was: two solid
propellant booster stages, and a bipropellant liquid third stage. Not only did
it have significantly more performance margin than a three stage booster and
small PBV, but the entire engineering development program would have been
vastly simpler and less costly. But it would have only kept two of the four
solids companies in business, and that was a no-no.
So the BMO/TRW scoped the cost of the three-booster-stage with PBV program, and
that was the baseline against which the proposals were evaluated. All bidders
adopted a "buy-in, bail-out" lowball strategy, and we would up with a missile
with significant technical risk that was seriously underbid. The final nail in
the coffin came when the BMO Commander sent the difference between our program
budget and the cost submitted by the bidders back to the Treasury (a career
move on his part). The money to "bail out" a failing program was no longer
there, but the contractors didn't know it.
The program was in trouble from the start, and as designs matured and hardware
was built, we started overrunning the budget and were unable to handle even the
routine problems such a program encounters. The most glaring defect soon
showed its ugly head. With the 30,000 pound weight limit, it soon became
apparent that we could never achieve the 6,000 nmi range requirement.
If money hadn't been the issue, we would have reoptimized the whole missile to
meet the to hard requirements of 6,000 nmi range and 30,000 pounds weight. But
we quickly realized that modifying more than one of the four stages was
impossible within the projected budget profile. So the decision was made to go
hat-in-hand to Congress, and request relief from the 30,000 pound weight limit.
The only way to regain the range by changing only one stage was to add 7,000
pounds to Stage I. It was the least cost, but by no means LOW cost. Among the
added engineering challenges were the control problems associated with the
increase in length to diameter ratio. It changed the bending frequencies, and
reduced the structural margins so that the flight control computer needed
significant redesign of its digital filters. The attempt to keep Stage I
length down led to a highly submerged nozzle. That, in return, caused the
failure of the first flight due to an effect no one anticipated. All of the
booster stages were static fired horizontally, so we thought we knew how
everything worked. But when fired vertically, Stage I would accumulate molted
aluminum oxide slag in the large stagnant volume aft of the throat entrance.
It filled with liquid aluminum oxide, and then when the nozzle gimballed to
perform load relief at altitude, the aluminum oxide spilled over the throat
inlet asymmetrically, causing a side force which was then opposed by an
gimballing in the opposite direction, and so on in a diverging induced
instability. A guy who worked for me, Paul Carman, and I figured out the
probable mechanism. No one believed us, but they did do a vertical static fire
of a Stage I in a test stand equipped with real time radiography equipment. We
had an X-ray video of the aft end, and watched as it filled with liquid AL2O3,
which sloshed around and spilled out every time the gimbal was commanded. The
solution was to take the length hit, and desubmerge the nozzle.
The second flight got through Stage I satisfactorily, but had a Stage III
control problem, and was blown up.
Around that time, the Berlin Wall fell, and soon thereafter so did the Soviet
Union. I remember being at work one night, on a commercial launch program, as
we all watched the TV news. George H.W. Bush announced that as a gesture of
trust and good will to the Russians, he was ordering the cancellation of the
Small ICBM. That was the cheapest goodwill gesture of all time.
The reason for the range loss at high temperature on Small ICBM was that it had
to use all Class 1.1 propellants to make the range requirement, and these have
a high burn-rate temperature sensitivity coefficient. That, and the fact that
it was a MOBILE missile with a much wider launch temperature requirement range
than silo (or submarine) based missiles have. A sub or silo each remain at
pretty constant temperature. There was no way to guarantee that with the Hard
Mobile Launcher.
As for ICBM ascent profiles, they vary widely with basic design approach. The
Soviet approach was way different from anything we ever did. They built very
high performance engines, and their vehicles had very high thrust to weight
ratios to minimize gravity loss. To deal with the resulting high dynamic
pressures, they simply built the structures like battleships. American liquid
ICBMs were built with ultra-light structures, and thus had to be babied through
boost. That was done through a combination of low thrust-to-weight ratio at
liftoff, a longer vertical rise time than any Soviet rocket, then an open-loop
tilt program as a function of time combined with guidance programming that
sensed any uncommanded angular rate - which only came about with an angle of
attack induced by winds aloft - and steered the rocket back "into the wind" to
eliminate the angle of attack. It was the guidance controlled analog of having
big tail fins, which aerodynamically null out any angle of attack.
The solid booster ICBMs had stage burn-time restrictions imposed by the nature
of solid propulsion - about a minute per stage. So they had to start to turn
pretty quickly. The Minuteman missiles were inherently robust structurally,
since their pressurized motor cases were also the missile primary structure.
Peacekeeper was designed as a mobile missile, though it went through a slew of
basing modes. One of these, the multiple protective shelter mode, had the
missile stored horizontally. If war occured, the shelter door would open, and
the missile and its launcher would be pushed out, erected, and fired.
Peacekeeper was canister launched, that is, it was ejected from a launch tube
by a steam generator. That applied to any of its basing modes. But in the
multiple protective shelter site mode, it was always assumed that there would
be some damage to the erector mechanisms which would limit the elevation angle
of the canister. The Stage I gimbal angle was therefore mandated to be more
than +/- 5 degrees, though in actual flight it almost never saw more than +/-
0.5 degrees. The high angle was to permit missile to recover from a launch
eject as low as IIRC 50 degrees from the horizontal. It may even have been
lower. Peacekeeper Rail Garrison and SICBM Hard Mobile Launcher were always
designed for off-vertical launch - something like 10 degrees from vertical.
The main problem with depressed trajectories in solids wasn't the structural
bending limit as it was the problem of stage separation in a high-Q
environment. That problem occupied a lot of my missile career. The
complexities are manifold, and beyond the scope of this tome.
I have left out an awful lot from this, both technically and program
history-wise, and hope it hasn't been boring. Let me know if you have any more
questions.
Best,
Mike
On Mar 24, 2023, at 1:06 PM, John Stoffel <john@xxxxxxxxxxx
<mailto:john@xxxxxxxxxxx> > wrote:
"Michael" == Michael Kelly <<dmarc-noreply@xxxxxxxxxxxxx
<mailto:dmarc-noreply@xxxxxxxxxxxxx> > ("mskellyrlv")> writes:
It was an excellent first test flight, though I have to wonder at
the criticality of max Q. Every launch vehicle with which I’ve been
associated has had a max Q-alpha limit that was easily handled by
load relief. Small ICBM, after they added 7,000 lb to Stage I to
make up for all our mistakes, was the biggest wet noodle missile of
all time.
Can you elaborate here please? This sounds like a really interesting
piece of engineering history.
It was the only solid propellant strategic missile to ever LOSE
range with increased grain temperature (requiring trajectory lofting
to lower max Q due to shorter burn time)
I would have thought that all strategic missles went stright up until
max-q was passed, if only to lower the drag before they started
heading to the target.
But we
know how to build for this, and mitigate the effects. There seems
to be something more in question about this rocket’s structural
integrity. Nevertheless, I applaud Relativity.