Crazy Idea 319 Ejector/injector Tribrid

guest blogger john hare

Development cost is a major player in anything concerning space launch vehicles. Why not heavy lift, RLV, laser, maglev, turborocket, scamjet, rotovator, BDB, and on and on and on, is largely a function of development cost at this point in time. If you can’t afford to develop a system, it really doesn’t matter how good it is going to be, it just won’t fly in any sense of the word. You can find proponents of any of those systems and more that claim that theirs is the one that will take mankind to the stars, or at least the solar system. However, the bucks start with development, and no bucks……..

Solid rocket proponents make a fairly good case for low relative development costs for new launch systems. Almost good enough to think about ignoring the many drawbacks of solid systems. Drawbacks like low Isp, combustion instability causing a rough ride, nasty failure modes, and other operational headaches. Der Griffenschaft has made many of these problems far more public than ever before. Still, there is that low development cost to think about when development funds are light.

Big Dumb Booster proponents make some pretty good arguments also. Simple construction, modular ease, throttling and shutdown capabilities and others. There are drawbacks here as well, like fairly low Isp, fairly low attainable mass fractions, development problems and cost of large low pressure engines. The first two are annoying, the third can kill you and your program. Large liquid engines are notorious for combustion instability. Fixing it can cost money and time, assuming you have the right team to do it in the first place.

It would be nice if either of these two could be modified for better Isp, stability and mass fraction. Safety in the solid is also a major concern. It may not be possible to bring either of them up to snuff, but it may be possible to use them  synergistically to improve all the requirements while dropping development cost. The strengths of one can support the weaknesses of the other. 

One strength of the solid is that it can operate at high pressures without complex equipment. These high pressure gasses can be used to pump the liquid propellant in a venturi section (like a paint gun) or ejector in a variant of the ejector ramjet. Here it is high speed gas pressurizing a liquid instead of  low pressure air. With the liquids having a density of about 1,000 times that of air, it should be possible to increase propellant pressure to a far greater value than the air breathing version. A high pressure solid rocket should be able to entrain two to four times its’ gas mass to a pressure two to four times the liquid initial value.

What I am suggesting here is a solid rocket of one mass unit with BDB tanks at low pressure with two or more mass units of liquid propellant ejector pressurized to medium pressure. The ejector becomes the injector simultaneously as the propellants are pushed into the lower combustion chamber. The hot fast gasses of the solid shear the liquids rapidly and evaporate them in the ejector section just before low chamber injection. The low chamber uses a hot gas-gas injection, which should have relatively low instability problems caused by variable droplet evaporation and mixing. A second consideration is that hot gas-gas injection should have a very low L*, which lowers mass and wall heat flux.

With the liquids having better gas properties than the solid, and higher pressure than the BDB, average Isp should be higher than either stand alone engine. With the solid providing ejector pumping action, the BDB tanks should be lighter for better mass fraction. With the liquid tank mass, and the predominantly liquid propellant, the solid engine vibrations should be heavily damped to an acceptable level. With liquid tanks perhaps surrounding the solid, they provide a partial flak blanket in case of solid spontaneous disassembly.   

I am suggesting that it might be possible to develop this tribrid engine system to an operational level for less than either of them standing alone, and get better performance in the bargain.


If this makes it feasible to build Falcon 1 class rockets with low development costs and a fast schedule, it might be possible to beat SpaceX in the marketplace, as well as Orbital, ULA and Proton. For sub-orbital RLVs, this might not make as much sense with a solid rocket motor in the middle of your airframe. For early entry orbital though, it might just be a valuable interim engine class.

The ultimate expression of this concept would be a Shuttle tank full of hydrocarbon and LOX with two SRBs modified this way. The five million pounds of liquid propellant would double that of the solids with about seven and half million pounds of propellant in an eight million pound lower stage. Take off thrust of twenty million pounds would let this lower stage lift a fully fueled Saturn 5 as an upper stage.

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)


About johnhare

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.
This entry was posted in Uncategorized. Bookmark the permalink.

18 Responses to Crazy Idea 319 Ejector/injector Tribrid

  1. Tim says:

    Interesting idea. Could this work as an add-on to an existing solid motor? I’m thinking if you added OTRAG style tanks to an existing ICBM, development costs would be confined to the injector/ejector section.
    I wonder if this might fit better with a hybrid rather than a solid motor. You could eliminate the oxidiser portion of the solid (which I understand is difficult to deal with) in favour of better throttling, and a common oxidiser tank; and the problem of the hybrid exhaust being oxidiser rich towards the end of the burn could be dealt with by enriching the fuel of the liquid section. There would be the complication of pressurising the oxidiser for the hybrid section, though.

  2. john hare says:

    I was thinking in terms of the solid being SEP rather than developing anything like a any type high pressure upper in house. Throttling could be with the liquid components alone. By using ablatives in the ejector section, the material regression in that aera would gradually open up during the flight, reducing the upper solid fuel pressure and solid regression rate. This would allow some tailoring of the thrust to match the desired flight profile.

  3. jsuros says:

    Can you choose solid fuels that will burn quickly enough and completely enough to avoid hammering your fuel injectors with lots of high pressure particles?

  4. John, thanks for pointing out how important development costs are! That is exactly why Universal Transport Systems went with them in the first place. In the early stages of an industry like this, startup costs are critical.

    On this idea, although the development costs would be far lower than an equivalent turbo-pumped liquid rocket it is still going to have a fair amount of development. I’m especially worried about the oxygen pumping side. If you use an easy to develop/standard solid, the fuel rich gas will destroy your engine when mixed with LOX. If you customize the fuel grain to have an oxidizer rich exhaust I fear that you will harm the Isp (or pumping power, in this case) and at the same time greatly increase your development cost. As I said though, probably still less expensive than a turbo-pumped liquid.

    Jsuros: First, yes there are solid formulations that provide a virtually particle free exhaust. But secondly, with this design you do not need normal injectors. Injectors in a rocket perform 2 functions: First, they mix the propellants together. Second, the propellant flow rates are disconnected from the chamber pressure. On the mixing, gas/gas mixing is much easier in general than liquid/liquid. For the propellant flow rates, as long as the solid’s exhaust is supersonic the liquids injected will not see the chamber pressure at all anyway. Honestly, this should have better isolation than a standard liquid injector.

  5. john hare says:

    If you are wanting the fuel rich solid gas to burn with the LOX, I don’t see how that would destroy the engine more than the normal burn cycle. If you design it that way of course.

  6. Habitat Hermit says:

    I can’t find any real flaws, like it a lot. It’s completely sane solids! ^_^

    Not sure if I’m reading too much into the drawing, do the venturi injection system cross from side to side? If so I would instead place all the Venturi injectors along and close to the perimeter/circumference with the inlets facing towards the perimeter/circumference. And whether or not there is a second stage I would extend and combine the first stage tanks upwards above the solid motor casing and join them. If done sufficiently to protect whatever is above it opens up possibilities for using the second stage as a launch escape system.

    Beyond those small tweaks I can’t think of anything much.

    Well about the design of the venturi segment I described above one could do it like this individually for each pair:
    Think of it as a big X only curved, the left and right half are separate tubes (oxidizer/fuel) that arch out from the wall and meet at a dip back towards the wall creating the venturi effect for the point of injection (this would be at the center of the X). The space between each arc and the wall is filled forming each side of the channel. After the point of injection the tubes are solid and curve back out to the sides as well as back towards the wall. Hopefully I described it well enough not to have to draw it ^_^;

    Maybe I should describe it as an hourglass shape split in half along the major axis? Or as halfway embedded in the perimeter/circumference? (The embedded half wouldn’t exist).

  7. Habitat Hermit says:

    Well there is one thing kind of, one has to be very careful with what the Venturi segment (or anything else for that matter) does with the pressure and flow characteristics of the solid fuel engine, maybe that goes without saying though? Since it will affect the solid fuel burn rate and/or increase the need for pressure containment around the fuel one has to find sweet spots that still have plenty of safety margin. Not really a flaw just something to keep in mind.

  8. john hare says:

    Habitat Hermit,

    My drawing was supposed to represent an idea rather than a true design.

    For the very small proof of concept units, I see a sugar rocket above a pinched section with oxydizer and fuel inlets on opposite sides of the pinched venturi. That is the sub thousand dollar “look at this crap” experiment just for the fun of it. Some hobbyists might even fly a few to see how it would really work off the bench vise.

    A bit larger would need injectors into the chamber as you suggest. A bit of low level CFD would most likely validate the method you describe. Also at this medium level of simulation or test, gotchas would be found to invalidate the idea if they exist. If they don’t, it seems possible that light LEO projects could be flying on something of this nature in a few years.

    For heavy configurations, the multiple ejector units would need to be seriously engineered. I see the ejector channels as alternating oxidizer and fuel preventing the mixing until the end of the elements. The best large ejector configuration might be concentric rings alternating oxidizer and fuel, fed through large structural support feeders from the walls. The bottom of the ejector units could be corregated (^v^v^v^v) to facilitate mixing the hot gas-gas combustion.

    For upper stage protection, a cap as you suggest would be good. For performance though, the Outrag style tanks as Tim suggested would be better. 6 tanks surrounding the solid could be dropped in pairs to enhance performance. The first two could be quite heavy and relatively high pressure for off the pad and low altitude. Second pair medium pressure for the mid level burn. Third pair light and low pressure for vacuum operation. This way the pressure tanks could provide considerable throttling with relatively low mass penalty. Also, if it was desired to change fuels in the engine during assent, it could be done in discrete tanks without a lot of pump discontinuities. The three pairs might want to be kerosine>methane>hydrogen, depending on the company involved.

    I see the protective cap as being a distinct second stage with slightly sub-optimal tanks as shrapnel catchers. OTOH, the cap tanks could be used on the first stage after the drop tanks are gone and the solid is nearly out of fuel. When the solid runs out, the remaining cap tank propellants would still provide enough thrust to allow lighting the next stage engines if they are the type that require gravity for the ignition sequence. (SSME)

  9. So, can the combustion chamber be described as a kind afterburner for the liquid propellants?

  10. jsuros says:

    Your earlier comments show that controlling the feed pressure of the fuel/LOX into the venturi ejector/injectors allows throttling of the engine. As always, I’m curious as to how far you could push this idea. Could you raise the pressure of the “lower” combustion chamber to a level near or above that of the solid gas generator by feeding in the propellants at high pressure?

  11. Eric Collins says:

    I think the idea may have merit, but I am concerned about the back pressure generated by the secondary combustion events. If I am understanding you correctly, the purpose of the down stream combustion of the liquid fuels is to further raise the chamber pressure, above and beyond that already generated by the solid motor exhaust. If that is the case, then there will necessarily be a drop in fluid velocity in this chamber as it encounters the rise in pressure. If the pressure increases significantly, it may reduce the flow velocity through the venturi, which will reduce the amount of propellant entrained in the flow.

    It is possible that the whole process may be self-regulating. If the chamber pressure grows enough to slow the incoming exhaust gas/fuel/oxidizer stream, then less fuel/oxidizer will be entrained by the exhaust gasses. This will reduce the amount of downstream combustion, which in turn, reduces the chamber pressure. If the system behaves in this manner, then it is likely that some form of dynamic equilibrium could be established.

    As David Summers points out, if you can keep the flow supersonic past the fuel injectors, then there is no way for them to be affected by the combustion events down stream. If you use a converging-diverging nozzle to force the primary exhaust gasses to go supersonic, then the rise in the secondary chamber pressure should not exceed 2-3 times the pressure of the gas as it leaves the nozzle1. Otherwise, the flow through the throat of the nozzle will not be able to attain supersonic flow speeds.

    Here’s a question; if the flow remains supersonic through the secondary combustion, then is there really a need for a second nozzle? If not, then would this look more like a solid rocket with a thrust augmented nozzle?

  12. john hare says:

    The concept as I see it would have a solid component at around 150 atm, the lower chamber at about 50 atm, and the venturi/tanks at 10-15 atm. This would allow fairly low pressure liquid tanks to use an engine pressure in the 750 psi range for decent Isp. More important is the possible ease of development. A bit of freelance flow simulation work would give some interesting answers here.

    I don’t see how the lower chamber pressure could be higher than the higher chamber. That would prevent flow in the right direction. I think you need the nozzle after the liquid burn to bring it back up to supersonic.

    It could be thought of as an afterburner.

  13. John Bossard says:

    Crazy idea 319, indeed! The fecundity of your ideas is a marvel to behold, John.
    The use of the solid-rocket for ejector pumping is quite novel. It will require some care to get a ejector/injector design that provides a net increase in total pressure and can survive the environment just downstream of the solid motor. Perhaps using a fuel-rich solid can help control gas temperature. Ejector pumping is quite inefficient, but perhaps the simplicity of the design more than makes up for that in terms overall system mass. The hot combustion gases from the solid will definitely help the atomization and vaporization process for the liquids. That makes the liquid injector simpler and less demanding on its performance.
    What is the anticipated liquid tank pressure and solid rocket motor Pc that you might anticipate?
    You can do a great deal of useful testing of possible injector/ejector designs using shop air from your compressor, plastic tubing, and PVC pipe fittings, and perhaps 2-liter PET bottles. That’s what I used on my vortex test apparatus.

  14. john hare says:

    Well you admited to 318 crazy ideas and I had to try to beat that. So when another nut comes along, we can use the Jack Nicholson line, “Sell crazy someplace else, we’re all stocked up here.”

    If this can be made to work, I see solid pressure at 3-4 times that of the main combustion chamber and perhaps 10 times that of the liquid tanks. Hopefully that scales. I would see a progression from shop air to sugar rocket to higher amateur sizes before any realistic space craft engine development. If it is all workable, I would expect a steady progression in sizes used IF it is a trully cheap development process.

    The inefficiency is a given. The results might not be. If it takes 10 times as much enthalphy to pump the fluids as a normal pump, that just means that solid mass times temperature will have to be 10 times that of a rational pump. The extra mass and heat is used for reaction in the main thrust chamber and through the nozzle and is therefore not wasted. The inefficiency shows up as higher dead mass for the solid because it has to be larger and probably higher pressure to make up for the pump performance losses. Since I am suggesting pumping flow on the order of 20% of the total MDOT, it seems possible that the pumping ineficiencies will be overcome with the pure brute force of the solid component.

  15. Habitat Hermit says:

    I think I like the concept too much to be objective but yes specific impulse should go down however system-specific impulse should leap upwards. Gut feeling is it should do so for the safety-conscious sugar fuels as well (although less of course).

    You all probably know about this man but just in case you don’t:
    (Warning: one can easily spend a lot of time at that site).

    I don’t know him and he seems like a very busy person but maybe he’ll have time to look at the idea and/or possibly recommend someone else for it?

  16. Axel says:

    Very good, for a crazy idea. It’s fun to think about. Some comments did confuse me. What I basically understand is: you are considering to replace the turbo pump with a solid rocket driven ejector pump.

    I’m not really familiar with the physics here, and won’t dig into it. Efficiencies, pressures, ISP, cooling, etc. … figuring this out and make it all work seems to be a major development effort to me. So you save on developing a low pressure liquid rocket engine, but at the price of having to develop (or at least integrate and test) both liquid and solid plus inventing a rocket exhaust driven ejector pump.

    In some ways you combine the disadvantages of solid and liquid. Operationally you have to support both solid and liquid infrastructure. Throttling will be very limited, especially if you use the liquid fuel to cool the injector/ejector. Controlled turning off of the engine (e.g. for stage separation) is very hard, especially if you would like to have a reusable system. Restarting the engine in flight (e.g. for fly back or vertical landing) is even harder.

  17. Eric Collins says:

    Ok. I’ve probably spent way too much time pondering this concept, but it’s been difficult to put out of my mind. So, here is what I’ve come up with.

    If the solid exhaust is subsonic as it moves past the liquid fuel inlets, then the pressure in the exhaust gasses at that point must be less than that in the tanks/fuel lines. So, high pressure in the solid combustion chamber must be reduced before it encounters the fuel injectors. If the flow is constricted, like say through a venturi, then it will accelerate thus reducing the local pressure in accordance with Bernoulli’s principle. The liquid fuel then enters the exhaust stream and becomes entrained in the flow, though hopefully not combusting until they are carried downstream to the secondary chamber. Unfortunately, I cannot currently see anyway to take advantage of the secondary combustion event with this configuration. Any combustion will add energy to the fluid, which will increase the temperature and pressure of the flow as a result. Since the flow is subsonic, the pressure wave will be able to find its way back upstream to the fuel injectors, and possibly as far back as the feed lines and tanks. The gasses will always prefer to flow from higher pressure to lower, so the geometry of the secondary chamber will have to be designed to immediately relieve this pressure. In my mind this implies that the secondary chamber will have to be an expansion nozzle. Thus, leading to my previous comment about thrust augmented solids.

    Now, there are a couple of possibilities for getting around this limitation without resorting to pumps. The simplest conceptually, but possibly problematic in implementation, is to use the solid exhaust gases to drive a piston which increases the pressure in the liquid tanks. This could be a direct pneumatic arrangement, which I’m not sure could be made compact enough to be practical for a rocket. Or, the heat of the solid exhaust could be transferred through a heat exchanger into a secondary fluid, which then expands (possibly by vaporizing) to provide additional pressurization to the liquid fuel tanks.

    The other possibility is to use a nozzle on the solid exhaust to choke the flow and cause it to go sonic in the throat. In this arrangement, I’m imagining the liquid fuel inlets are like backward facing steps downstream of the sonic point in the nozzle. As the Mach 1+ fluid passes the step, the flow stream will attempt to turn the corner, but if the maximum turn angle, as given by the Prandtl-Meyer function can be kept below 90 degrees, then there should be a small vortex generated in the corner of the step which could be used to entrain the liquid fuel into the passing exhaust gasses. (See also Prandtl-Meyer expansion fan.)

    I am still not sure that the secondary combustion event can be sufficiently isolated from even this kind of injector design. Perhaps I’ll get my flow solver running on my laptop sometime in the near future and try to run a few tests.

  18. john hare says:

    Habitat Hermit
    I hadn’t seen that site before, seems interesting at first glance. I’ll check it out when I have more time. Actually I have mostly dismissed solids out of hand until the recent Griffenschaft thoughts.

    I think your objections are quite valid. They would be the basis of any reasonable investigation of the concept.

    Your third suggestion is the one I see. The venturi section was sort of supposed to represent sonic flow. Below the venturi, the flow slows down for pressure recovery. The slowing as the area widens is also the injection plane since there are no formal injectors. The flow through the lower chamber will be higher velocity than normal because of the high speed hot gas-gas injection. The lower chamber restriction will be neccesary to create the second sonic throat before expansion.

    I appreciate the patience with my incomplete (and sometimes wrong) descriptions. I see the flow as subsonic in the solid chamber. Sonic in the venturi with the pressure drop to the lower chamber. Subsonic in the expanded lower chamber. Sonic again through the true throat. Supersonic through the expansion nozzle.

Leave a Reply

Your email address will not be published.