Launch Escape

guest blogger john hare

There is an increasing discussion of launch escape systems in the last few months. Most of them seem to focus on developing some sort of high acceleration tower to yank the capsule free in event of booster malfunction. The financial numbers involved get quite entertaining, with the added dangers of a seldom used system a close second.

If the second stage could just do an early high acceleration staging event, it would seem desirable to do that and save the whole upper stage by having it land at the launch site under it’s own power. This would also  have the advantage of using regularly used engines that would almost certainly be more reliable than the every several year operation of the units now proposed.  Unfortunately, that would require oversizing the upper stage engines for a major cost and mass hit as they would be normally operated at far below their designed thrust.

Unless you steal borrow a page from the orbital refueling proponents. Stretch the tanks on the first stage some so they hold as much propellant as both stages normally would. At take off, the second stage has only 20 seconds or so of on board propellant. If the upper stage has a normal mass ratio of six, then if it has to stage early, it can have a thrust to weight ratio of well over five, which should be sufficient to clear the lower stage debris field with adequate failure warningLES

 Just before staging, the upper stage is fueled up from the first stage tanks. This puts the entire mass penalty of the LES in the first stage tank stretch and reusable high volume/low pressure transfer pumps. IMO, this would be faster and cheaper that an LES tower that is expended every flight or carried all the way to orbit. If it saves the upper stage as well as the cabin, it could be better as well. It might also reduce the insulation requirements for the upper stage tanks as they would be topped off just seconds before lighting the  engine(s).

The following two tabs change content below.


I do construction for a living and aerospace as an occasional hobby. I am an inventor and a bit of an entrepreneur. I've been self employed since the 1980s and working in concrete since the 1970s. When I grow up, I want to work with rockets and spacecraft. I did a stupid rocket trick a few decades back and decided not to try another hot fire without adult supervision. Haven't located much of that as we are all big kids when working with our passions.

Latest posts by johnhare (see all)

This entry was posted in Uncategorized. Bookmark the permalink.

34 Responses to Launch Escape

  1. Paul Breed says:

    I’ve always wondered why the launch escape system could not be used every flight to add a little bit of Delta V. IF fire it as the settling
    motor after stage 1 sep and befopre stage 2 ignition.

    That way it gets used and tested…

  2. Such a concept might be a great alternative for manned aerospace plane launches on top of a booster.

  3. Tom D says:

    One problem I see with it is that there would almost certainly be a block of time before the end of the 1st stage (S1) burn where the 2nd stage (S2) has become to heavy from the propellant being madly pumped into it to be used as a launch escape system. It simply cannot accelerate fast enough anymore. Since pumping the propellant from S1 to S2 will not be instantaneous this no-abort period could be sizable.

    However, this suggests to me a possible solution. Instead of trying to pump propellant between stages, how about having two or (more likely) three sets of tanks in parallel for S2 with the motors mounted to the center tank stack only. Normally the S2 motors would drain all three sets of tanks, but in case of abort, the outboard S2 tanks could be dropped during the stage separation to lower the weight of S2 for better acceleration performance.

    The big downside for this scheme is that an abort would involve not one, but *three* separations. A possibly smaller downside would be the extra big hammerhead configuration that this would present for aerodynamics and control. A possible plus is that you could drop the outboard tanks when they are empty for some mass savings.

    The first (1995) proposal for the Russian Angara launcher looked kind of like this with outboard tanks, but they were for both first and second stages. I rather suspect that they were proposed because large motors were available, but the possible tank diameters were severely limited. The designers seem to have changed there minds on the usefulness of these outboard tanks, though. They disappeared by at least 1999 from the designs. See:

  4. Wow – this is simple and obvious once stated! It seems to have all the hallmarks of a brilliant idea. If I was building such a rocket, I would give this serious consideration.

  5. A_M_Swallow says:

    How do the astronauts escape if the second stage has the problem?

  6. johnhare john hare says:

    Tom D,
    I think your idea is probably better than mine, and addresses A_M_Swallow’s point as well.

    Beware tachyon radiation from my handwaving. By the time the upper is too heavy for rapid escape, the lower will have very little explosive potential left being almost dry, maybe. I think either of these ideas might be worth a low level evaluation if someone is about to spend major bucks on a specialized escape system.


    I don’t know on my idea, I think they have a bad day. Do the current escape systems address that issue well?

  7. Rüdiger Klaehn says:

    I think this idea has some merit. It is true that in the last 20 seconds where the propellant transfer takes place, the upper stage would not have sufficient acceleration to get away from the first stage. But then an almost empty first stage is unable to produce a huge fireball that you need to get away from at >5g. In addition, the last 20 seconds would be with almost zero dynamic pressure. So a separation with lower g would probably still be survivable for the last seconds.

    You obviously have a problem if your second stage uses a turbopump, since a turbopump powered engine probably takes too long to start up to be useful as an emergency escape rocket.

  8. gravityloss says:

    There’s something to the idea. Abstracting a bit, the main aim is to keep the aborted mass as light as possible.
    There are multiple approaches here.

    This could mean a Soyuz style vehicle, keep the crew in the small re-entry module and only separate that on abort. Much much much lighter than Orion. (Even without the SM.)

    But then you get configuration headaches. You need a heat shield hatch.
    One idea is something like a certain Soyuz rocket upper stage where there is a ring of tanks and the motor is in the center. You could just fly out from the center with that and leave most of the tanks behind. Pretty much like Tom D:s tanks idea. But then you have to abort the whole payload, which is heavy. Service module propellant being the biggest. If your spacecraft was mounted sideways…

    In a sense, disregarding real flight conditions, a good abort system would consist of the second stage motor and perhaps tiny tanks coupled to the re-entry module. The rest of the hardware (orbital module, service module, second stage tanks etc etc) could reside beneath that combination somehow.

    And, actually, there is a vehicle with a configuration resembling that in parts – the Space Shuttle! How fast would it accelerate with just its own mass and the SSME:s burning? (The SLWT weighs only 26 t empty though so it’s not *that* much.) I get a T/W of about 5 so 5 gees. Of course the solids would race much faster with anything but a full ET and would fail much quicker anyway to make it moot… 🙂

  9. John,
    I had pretty much the exact same idea a while back (ugh I sound like Gaetano when I say crap like that). I’ve always been a fan of eliminating the big solid rocket of doom from the top of the stack. This is one of the more creative ways of solving the problem. It allows you to use a normal engine, and as you say it offloads the dry mass penalty to the first stage. If the first stage uses pump-fed engines, the penalty might actually be pretty small. Just divert some of the turbopump stream to fill up the upper stage and get simultaneous throttle down toward the end of the run (like many stages do to limit G’s).

    As it is, most stages start doing things like pressurizing the upper stage only late in the first-stage flight, so something like this would only be a tiny bit riskier than what people do now. And with an RLV, if something goes wrong, you’re in much better shape to do an abort to the launch site still because the upper stage is mostly empty at this point…

    I like it.


  10. jsuros says:

    This idea lends itself well to a popup first stage trajectory, where there is a long delay between the main engine shut down of the first stage and the separation of the stages.

    Used in this way, I can’t help wondering about the possible designs of propellant tanks that are filled entirely above the atmosphere. Metal foil bags hanging from booms, maybe? Cylinders hanging off the sides like a boat bumper?

  11. johnhare john hare says:


    It took me a while to figure out what was bothering me about your comment. You think it is best to keep the aborted mass as small as possible, while I was thinking in terms of saving as much ship as possible. Since I posted this idea concerning current escape systems, I am guilty of mixing concepts since they are all ElVs. Somehow I had RLV upper stage in mind which doesn’t apply to the current contenders. Maybe it should.


    Gaetano’s spams seem to disappear while stuff he puts thought into doesn’t. Odd.

    For the record, I borrowed this from you, borrowed I say, you can’t convict me of larceny here………….can you?

  12. donnie says:

    If VB were alive he’d reply to this and have something interesting to say

    I’d like to hear it, but in lieu of one rocket innovator, I’m happy to hear from five others

  13. donnie says:

    Skribit What should I write about?

    should hl-20 use solid abort motors, hybrid motors, or liquid motors with a) small full internal tanks just for abort, b) big full internal tanks for abort and transit, c) near-empty internal tanks that are refueled by an earlier stage on orbit for both abort and transit, d) small full internal tanks and an external tank on the side or bottom?

    if the primary goal is increasing safety by increasing the flight rate of the abort motor by using it as the stage motor, how does the safety of each configuration compare? does the viability of an option change when comparing capsule to winged lifting body crew vehicles?

    are inflatable external fuel tanks viable?

  14. Carl says:

    On the ELV side of things, Elon Musk has pointed out that you will use retrorockets, or an escape rocket; but not both. So why not combine the two?

    I don’t know if that’s what he’s planning for the escape system for his rocket; but it is a substantially good thought.

    The discussion about engines surrounded by tanks which may be jettisoned brought to mind images of a multi-stage Armadillo Pixel/Texel. 🙂

  15. Habitat Hermit says:

    The original idea is pure gold however I don’t like the suggested tweaks at all and there’s no need to limit this to RLV second stages.

    For concerns about turbopump dependency as well as a late explosion of the first stage my own tweak would be to include (perhaps internal, i.e. tanks within tanks) tank volumes in the second stage that are constantly pressurized from before launch (i.e. quick-start without further turbopump dependency). The size of these tanks would be determined by max volume needed for whichever situation requires the most thrust (depending on overall propellant mass and propellant choice would that be “on-pad-at-ignition” use or “explosion at max-Q” use or would there be a midpoint that beats both? There’s no way the last twenty seconds of the first stage flight would need more power than that to escape what would now be a small weak explosion and it could perhaps instead be too much –see the next paragraph).

    A great additional benefit to the original idea would be the ability to tune the internal LES dynamically in flight according to the flight profile to minimize unwarranted acceleration at higher altitudes thus minimizing g loads. Controlled gradual pressure release back to the same stage main fuel tanks should do nicely for this.

    Second stage failure must be far less difficult to mitigate, one could have a tiny external/”normal” tractor LES however I doubt it would be necessary to instead add anything but a little bit of additional blast shielding (letting the much smaller explosion rush away from you as opposed to outrunning it). At any second stage explosive failure the vehicle won’t need a LES to achieve parachute height as that will be somewhere below the current height.

    The only bad consequence of the original idea that I can think of is that one most likely won’t be able to use LM balloon-style tanks (i.e. tanks that attain structural rigidity from fuel pressure) like the Centaur for the second stage.

  16. Nathan says:

    How long would it take to transfer the propellant?

  17. jeff m says:

    A fundamental question you’ll have to consider is your upper stage engine selection. Generally an upper stage engine (high nozzle exit area ratio) will get torn apart if fired at sea level . Therefore, if you were to employ this design solution, you’ll most likely have to live with sub-optimal engine(s) with a big hit in efficiency.

  18. A_M_Swallow says:

    The service module also has an engine. It may not have produce much lift but I suspect that it can push the command module horizontally, possibly in an arc.

  19. ColinR says:

    I always regarded the klipper case with a ring of spherical tanks around a central engine as a neat solution..

    #15 @jeff m
    If you use an upper stage engine with a nozzle extension then the ‘small’ nozzle created when the extension is stowed can be sized to cope with sea level and obviously the deployed nozzle extension copes with the vacuum case.

  20. Joe Latrell says:

    In the event you have a problem and need to get the crew out, having the second stage tanks drop off is a great idea. Use the center mount engine/small tank set and hit the throttle, jettisoning everything not needed. It is a great idea that should be researched.

    BTW, f the objective is to save the crew (an not hurt the public), the vehicle can become over-sized tinker-toy pieces for all I care. Bring back only what will save them. Hardware is replaceable.

  21. Steve Wachowski says:

    Nobody has answered the question in post 5 by AM Swallow. What happens if the problem is with the second stage to begin with? Seems to be far too Rube Golbergish to be of any use. What I see is a series of very complex operations that need to happen quickly and perfectly just when things are not going perfectly.

  22. Jonathan Goff Jonathan Goff says:

    Steve, AM,
    Pretty much every LAS is jettisoned right after or right before staging anyway. Nobody carries them all the way to orbit. So I would imagine you could do what everyone else does–use the RCS or other engines for “escape” purposes. Sure, if the 2nd stage decides to spontaneously explode, it won’t save you, but neither will the traditional approach.


  23. A_M_Swallow says:

    We can use the second stage to escape an exploding first stage.

    There are methods of extracting capsules and people from launch pads – these will cover problems with both first and second stages.

    At what height can the service module be used to escape the second stage?

  24. johnhare john hare says:

    Brief thought not worth a post. Could the upper stage or service module have some plumbing run for rapid dumping of onboard propellant into their expansion nozzles for a TAN boost, trading very low Isp for a massive thrust increase under emergency conditions only?

  25. tom cuddihy says:

    So crazy it might be…crazy.
    There are serious thermal issues with trying to dump massive amounts of cryogenic fuel rapidly into a tank during flight aren’t there?

    I’m supicious TAN would also have problems in the milliseconds where you need escape thrust.

  26. Tom,
    TAN has to function in a timeframe of less than 100 microseconds usually…IOW, I think there’s a reasonable chance that an afterburning engine for escape purposes could actually be competitive with large solid tractor motors (which have startup/thrust build-up problems of their own).


  27. Of course, an interesting observation is that this concept works a lot better with likely RLV designs than with existing ELV designs. People have brought up the fact that in order to do a pad abort, you need the engine to be able to do sea level operations, while it still needs good high-altitude performance. Note that this is the very challenge that faces powered-landing RLV upper stages. It might unnecessarily complicate an existing ELV upper stage…but there’s no point in trying to save the whole stage if you’re just going to toss it away moments later.


  28. Brad says:

    For VTVL RLV it sounds like a plausible idea. But for ELV it sounds like it would involve too many compromises to simplicity and performance.

    When it comes to the more general issue of launch escape, pressure suits today are universal SOP, so I’m surprised that ejection seats have not made a comeback for consideration. They certainly have some advantages over conventional tractor escape towers. The primary advantage is an ejection seat only has to fire during an actual emergency, whereas an escape tower must fire every launch at some point during ascent to jettison it’s costly deadweight. That tower jettison event is another potential failure that can cause mission failure and/or loss of crew.

    Launch escape towers are quite the two-edged sword. It may save you if something goes wrong, but it might also kill you because it adds one more thing that can go wrong.

  29. johnhare john hare says:

    Ejection seat in a space suit at max q sounds suicidal to me.

  30. Brad says:

    Probably, though some interesting work in the X-15 seat was designed to mitigate high speed ejections. But no launch escape system is perfect and all involve tradeoffs. And what percentage of probable launch escape events occur at max q?

    So, how does a failure of an ejection seat at max q compare in probability to a potential failure of a launch tower jettison event?

  31. AlbertoT says:

    How about the 2nd stage motor starting with the nozzle in the interstage?
    Why not use an extended interstage as a sort of blowpipe (pressure coming from a big rupture valve in first stage)?

  32. johnhare johnhare says:

    Do you mean using the pressure in the upper stage to force the propellants through an interstage water rocket type thing?

  33. AlbertoT says:

    Not a water rocket.
    Imagine a giant pneumatic pusher driven by the same gas used for tank pressurization (first or second stage is the same).
    You can have a very big force with a small pressure due to the large area.
    The effect is that of a blow pipe (or an air compressed gun).
    Moreover you get more clearance from a doomed first stage by pushing it back.
    Basically this is a high power stage separation device.

  34. AlbertoT says:

    Sorry, my post #31 was misleading.
    First part is about your idea; it’s ok for me except for 2nd stage motor ignition sequence with nozzle in the interstage.
    This could lead to serious problems, isn’t it?

    Second part is another idea I had about using the interstage as a high power separation device, extending it down the first stage and creating a sort of very big pneumatic pusher with the interstage being the cylinder and the first stage the rod.

    Curiously the water rocket idea that you put out in your post could be a real boon for a pressure fed second stage; nothing could be simpler and faster, yust some pyro bolts on a flange and on the thrust structure of the second stage.

Leave a Reply

Your email address will not be published. Required fields are marked *