Jeffrey LewisBulavas Whirling Though the Sky

We’ve had quite a discussion in the comments over Russia’s recent failed Bulava SLBM test and the pretty spiral it made before crashing into the ocean.

Oooh. Aaah.

Having largely dispensed with the usual spate of extraterrestrial and paranormal hypotheses — is everyone on the internet crazy? asked one commentator — you, dear readers, have gotten down to brass tacks.

As far as I can tell, there are basically three non-insane explanations floating around, none of which are (strictly speaking) mutually exclusive:

1. Something went terribly, terribly wrong.
2. It was an energy management maneuver.
3. It is a countermeasure designed to defeat future boost phase missile defenses.

For the something went terribly, terribly wrong we have the official statement of the Russian Defense Ministry that the third stage became “unstable” and that the test failed. As we know from similar US mishaps, an unstable missile can make a couple of loopdy-loops before going kaput.

On the other hand, this is correlation not causation.

Which brings us to the second reason that solid-fueled missiles make pretty spirals — so-called “generalized energy management” manuevers or GEMS. You can’t just shut-off a solid-fueled missile in mid-burn (at least not easily), so you need to let it do a little dance to burn off some energy.

Here is a nice video of a US THAAD interceptor getting funky before getting down to business.

The Bulava was most likely aimed at the test ground in Kamchatka, so it would have needed to perform an energy management maneuver to reduce the missiles range from 8,000 km to about 3,000-4,000 km. The fact that a bunch of drunken Norwegians think that a StarGate has opened up is just a bonus, as far as the Russians are concerned.

Which, brings us to the countermeasure hypothesis. A missile in boost could use an energy management maneuver to attempt to evade a boost phase missile defenses.

It isn’t clear to me that the nice tight spiral is the optimum countermeasure, since the goal is to move unpredictably in ways that stress the kill vehicles maneuverability. One hears talks of dog-legs, lunges and jinking — not tight spirals.

Indeed, since GEMS are likely to be part of most ICBM and SLBM flight profiles, I presume our any future boost phase system would be designed against the typical array of spirals and so-forth.

Moreover, the problem with executing an energy management maneuver is that it reduces the distance that the missile flies — which is, of course, the point. With an 8,000 kilometer range, the Bulava wouldn’t want to manage too much energy on its way to North America unless Vladmir Putin has something against Pugwash, Nova Scotia.

I haven’t done any calculations, but fortunately Mosher et al discuss GEMs and other boost-phase countermeasures at some length in the Final Report of the American Physical Society Study
Group on Boost-Phase Intercept Systems for National Missile Defense
. If you are interested, I recommend pp S230-236, S264-266, S271-274.


  1. Fylla Har Skylda

    We not drunk, we sober…….ing up.

    🙂 yeah it created a UFO hysteria for some hours

  2. RAJ47

    A very interesting post.

    Could you clarify a small doubt, plz?
    If we consider countermeasure hypothesis then another ‘ boost phase missile defenses’ missile would have been launched in all likelyhood.
    Hence, GEMS theory seems to be the most plausible.

    The final report of APS is educative. Thanks.

  3. Jochen Schischka (History)

    “You can’t just shut-off a solid-fueled missile in mid-burn, so you need to let it do a little dance to burn off some energy.”

    Yes, you can!

    In essence, thrust termination in solid fueled boosters characteristically works like a bathtub-plug. There are, additionally to the main nozzle, sealed ports on the booster casing/thrust chamber which can be opened at the right moment by an explosive charge. Those ports (in essence a couple of small nozzles symmetrically arranged around the radius of the booster, in many cases at the upper end) are specifically formed and aligned so that the thrust of the main nozzle is neutralized.
    Just pull the plug at the specified velocity and voilà, thrust terminated in mid-burn. Simple as that (well, actually not that simple, as designing this system to sufficient reliability usually causes the most trouble on solid boosters, but simple enough in principle)!

    An alternative way of solid-booster thrust termination is used on the third stage of e.g. the Trident SLBM: a ring-like post-boost-vehicle carrying the MIRV-warheads surrounds the third-stage motor to which it is connected by explosive bolts; at the right velocity, those bolts are activated and the bus is released while the now lightened booster whizzes straight away and burns out uncontrolled (which doesn’t matter anymore, since all that counts is that the MIRV-bus has generally the right velocity and heading).

    BTW, lower stages ususally aren’t thrust-terminated (thus are allowed to burn out on their own – which generates a non-insignificant amount of uncertainty considering the respective burn-out-velocity, which can be compensated by the thrust-terminated upper stage and the post-boost-system) for matters of simplicity and economy (saves weight, maintenance hours and money).

    The energy-management method only makes sense if you want to have a lighter and thus more maneuverable missile later on during burn (aka anti-missile systems etc.).

    In case of the Bulava, we can clearly see that the booster burns out uncontrolled – there is no additional ‘flash’ from thrust-termination at burn-out (and the second method with the ring-like bus wouldn’t result in such a spiral, either. We’d see something flying constantly in a straight line, and making a visible leap at bus-release before burning out).

    Thus i’d definitely exclude the possibiltiy of the ‘norwegian spiral’ being anything but a booster failure. This is no russian ‘super-duper-ultra-sophisticated wonder-weapon’ – it’s just a dud!
    (I wonder if this ‘spiral misconception’ has it’s roots in the stories about iraqi Al-Hussain-missiles not being intercepted by PAC-2 because they flew spirals on reentry – but that was a.) before the times of the PAC-2(GEM) and b.) a totally different case: on reentry, not during boost, non-separated warhead, imperfect missile concept, liquid-fueled missile etc. etc. etc.)

  4. Jeffrey Lewis (History)


    Yes, you are correct about the bathtub plug possibility. (I ought to have added — “at least not easily.” I will clarify that in the post.)

    As you note, it’s not easy relative to an energy management maneuver.

    So, that leaves the interesting question of whether Bulava does a GEMS or not.

    I am not sure we can exclude that possibility — yet.

  5. anon

    “Having largely dispensed with the usual spate of extraterrestrial and paranormal hypotheses — …”

    Yeah, sure! Now we know what happens when you shoot at the ETs

  6. Jochen Schischka


    Well, i can’t exclude that possibility with 100% certainty, either (only with something in the 99% range…).
    Let’s just say that GEMS on a large surface-to-surface missile simply doesn’t make a lot of sense to me, especially since there are no boost-phase-interceptors out there (at least not yet, and probably also not in the next couple of years). And those would most likely not attack the upper-stages, but during the first-stage-burn.
    Only in case of missile-interceptors do i see some advantage in GEMS (lighter and thus more maneuverable missile while avoiding too high velocity end-game-wise, which would result in an excessively short intercept window aka impossible maneuverability requirements, lighter missile or not).

    Back to Bulava: An interesting detail of that ‘norwegian spiral’ is that on closer examination there seem to be two sources of gas-emission (see e.g. around 0:06 in the video), thus the ‘barred spiral galaxy’-look; It seems to me, too, as if those two sources wouldn’t be aligned and on different ends of the booster, so maybe what we see is a missile out of control because of a ruptured seal on one of the termination ports (this was also a big problem during the development of the Polaris A1 SLBM). If that should turn out to be the case, it would indeed be an indication for rather shoddy workmanship on the Bulava-program, since there are apparently no such problems on the in this respect rather similar Topol-M.

    Also very interesting is the ‘black hole effect’ (0:12-0:15 in the video) due to the uncontrolled burnout of the booster while still spiraling – so this clearly is no GEMS in the classical sense (which requires that the solid-booster still burns after energy management for the thrust-vectoring-system to work for final corrections – see the THAAD-video), while GEMS for ‘thrust termination’ purposes would without doubt result in a rather excessive inaccuracy (i’d estimate something in the 10-100km-range, instead of the usual 0.1-1km CEP for intercontinental ranges…but that is just a wild guess), since it’s impossible to determine the exact timing of an uncontrolled burnout – so it would result in a random heading plus a wildly rotating burnt-out missile (aka lateral speed at bus separation).
    O.k., admittedly a good post-boost-system (stellar- and preferably also Glonass-inertial) may be able to correct this flaw, but then a lot of additional propellant would be needed just for correcting the improper thrust-termination (which could otherwise be used more advantageously for mid-course maneuvering to avoid GBI), which would result in a larger upper stage (additionally strengthened for excessive maneuvering), which in turn would result in an even larger lower stage (or stages). All-in-all rather counter-productive for SLBM-purposes where size is a restricting factor, if you ask me.

  7. FSB

    Would not the Ruskis test their alleged possible new CM in a desolate part of their own large country instead of giving an open show to the Finns?

    I do not believe the CM hypothesis.

  8. FSB

    OTOH, perhaps it was Santa’s new and improved sleigh and he wanted to throw off all the bad little boys’ and girls’ missiles while enroute to the N. Pole 😉

    Merry Xmas!

  9. George William Herbert (History)

    Actually, thrust termination may be easier compared to energy management maneuvers.

    A relatively tight spiral requires a fair amount of attitude control to establish and maintain stably. TVC attitude control on solids, using flex nozzle seals of various sorts, is nontrivial. Those using fixed nozzles and separate monopropellant or hypergolic thrusters usually want to avoid having to use them much, and usually don’t bring much propellant for them, though using the same thrusters for third stage ACS and PBV ACS/velocity trim thrust is a possibility.

    From a reliability point of view, a small explosive charge and a front-facing exhaust port is a pretty easy thing to do, and not very heavy. You can do cute packaging tricks, like using an intentation for a nose aerodynamic spike as the vent nozzle. Various base nozzle options exist too.

    There are also various quenching technologies, using an inert liquid or gas dump to put out the solid fuel burn. Used in combination with venting to reduce chamber pressure, successful in flight solid shutdown on command has been demonstrated, but not flown on any production vehicle that I am aware of.

    The easy solution is and has been to simply pop the case somewhere and let it burn out.

  10. Jeffrey Lewis (History)

    I suppose there is an empirical answer on how misileers feel about GEMS v. others methods, but it is my understanding that GEMS is preferred. I can ask around.

    On a related note, perhaps FSB is onto something. Maybe Santa is hoping to deploy a new sleigh before the big night?

  11. Dennis Krenz (History)

    Against the GEM speaks the fact, that spirals are not that useful for long range missiles. Typically, you just fly a S-shaped ground track already during the first stages, since you have the flight time to do this and can get rid of a lot of energy by just a tiny offset from the azimuth early during flight. Doing this early means you can get rid of way more energy with less effort. And doing such a GEM in plane of the flight (pitch motion) is not so smart: It is simpler to predict the effect on the cut-off energy off-plane, then in-plane, since the in-plane deviations have a non-linear effect on the trajectory – which includes all the errors by the solid rocket motor performance, that you want to correct with the PBV. Thus a in-plane GEM would cost way more ensured performance in the final design, than a off-plane GEM, since you would have to reserve way more fuel in the PBV for correcting strong GEM maneuvers.

    And has the advantage that boost phase predictions of the target heading by early warning satellites are flawed: The missile defenses would be slightly surprised by your real trajectory. A small 2° deviation would be more than enough for some surprise elements. And if you would correct the 2° azimuth error during third stage flight, you would have a good chance to make boost phase interceptors miss. Though this would cost a lot of performance from the missile design as whole, since you then need a lot of more fuel on the third stage, etc… would be more effective to teach the PBV to do erratic maneuvers and surprising deployments, so you have more effect post-boost. And you could gain more protection from boost-phase interceptors by just reducing the duration of the boost phase. A single second of boost phase less results in about 5 NM less distance traveled after launch – the more seconds you save the better it gets. And the closer you are to your launching submarine at the end of boost phase, the harder it is putting ABMs into your way.

    Even the THAAD didn’t do such loops, but rather rotated around a cone – a less simple maneuver, which ensures that the sensor keeps on tracking the target.

    Stopping a solid rocket motor by opening the case at a second point works pretty well, but has a tiny disadvantage for ballistic missiles: You need additional heavy hardware, that can fail, for making sure you don’t put the following stages or PBV out of control. Effectively two smaller nozzles, so you have really zero net thrust – which always mean you have a chance of engulfing the next stages with hot exhaust.

    GEMs don’t need additional hardware – if your missile can deploy multiple warheads and/or has explicit guidance, its guidance system can also deal with off-plane thrust.

    And of course, it is also important to note that almost all inertial guidance systems for rocket applications are today capable of doing such GEMs – it is pretty doubtful the Russians go back to 1965 in technology, if their own industry produces light-weight 6DOF guidance systems (with weak quality control) for the Briz upper stage family and others.

  12. .

    The Russians reported a third stage failure, and looking at the photographs I count two plumes, blue then white – two stages. Given a squashed perspective and assuming a controlled and allowed for roll in the ascent stages, the dramatic spiral may not have looked so dramatic from other angles. Exhaust travels a lot more directly and freely in space.

    IMHO this was a relatively normal launch, perhaps but not necessarily with some spiral thrust bleeding but with proper functioning of the first and second stages, followed by a third stage ignition failure and no further exhaust.

  13. Jake (History)

    “From Polaris to Trident” says that GEMS was used instead of thrust termination on the third stages of Trident C4 and D5. But in the videos it looks like the entire third stage flight consisted of the vehicle spinning around in a circle, which sounds a lot more like a failure than an attempt at trajectory control. The high-res pictures also seem to show two plumes in slightly different directions, one smaller than the other, which also fits with a structural failure of some sort in the third stage.

    I’m also not sure what the benefit is to convincing the other guy that your SLBMs are less reliable than they actually are…

  14. Jochen Schischka (History)


    ““From Polaris to Trident” says that GEMS was used instead of thrust termination on the third stages of Trident C4 and D5. “

    I’m not familiar with that source (so i can’t say if this is some sort of misunderstanding or if it’s in fact misrepresented there), but i can exclude with certainty that either Trident used/uses GEMS for burnout-velocity control purposes. Both C4 and D5 clearly employ the “drop the post-boost-vehicle (with the warheads) at the right velocity (by explosive bolts)” method i described earlier (a ringlike bus carrying the warheads surrounds the non-terminated smaller-diameter third stage in the middle, joined by explosive bolts etc.). They simply release the burning booster at the pre-determined velocity.
    Both Tridents may additionally use GEMS to be less predictable (and this of course has a negative effect on the achievable maximum range…), but that has nothing to do with thrust termination (which is indispensable if an acceptable degree of accuracy without excessive growth of the launcher is desirable).

    But on all other issues, i completely agree with you, this looks more like a (rather classical) failure. It does not make a lot of sense to fly circles until the fuel runs out instead of shutting down the engine. And it does look like there are two exhaust plumes (like i wrote in my second comment on this thread: “barred spiral galaxy look”…), as you describe.
    My hypothesis is that we see an out-of-contol missile due to a seal failure on one of the thrust termination ports (which generates a pitch/yaw-moment because of the asymmetric add-on-thrust on the forward end of the missile).

  15. Jeffrey Lewis (History)
  16. Jochen Schischka (History)

    Thanks for the link, i’ll check that up.

    Happy new year, everybody!