Ejector/Injector TAN Modules

guest blogger john hare

The comment exchanges with Eric on crazy 319 triggered another thought. Jons’ post on Thrust Augmented Nozzles (TAN) mentions that the work on scramjets led to the technology. That part never made sense to me as I tend to think of scramjets as about the most useless thing ever conceived. The handling concepts of supersonic gas in the last post and scramjets does seem to merge  in the TAN technology now.

If you place a supersonic inlet inside a rocket nozzle at a specific point, it will always ‘see’ the same mach number. That inlet can be optimized for that one value for a high efficiency and low mass hardware. The leading edge of the inlet is in the lowest pressure region of the compression zone and is a good  place to inject propellants. The propellants injected are carried along with the compressing gas and pressurized in the process. They are also vaporized rapidly by both the shear and temperatures of the supersonic stream.

By injecting the fuel and oxidizer on opposite leading edges of the inlet, they don’t get a chance to mix until about the time they go through the final normal shock of the supersonic intake.  At the last inlet shock, the hot vaporized propellants are forced into contact just as they enter the subsonic combustion chamber. Given the favorable conditions for rapid combustion, L* will be very short. After combustion, a short convergent/divergent nozzle accelerates the TAN exhaust out into the main expansion nozzle. With no pumps or formal injectors, expansion nozzle stolen from the main engine, and a short L*, module thrust to weight ratios should be extremely high.

ejector injector

In this sketch I suggest several nozzle wall mounted TAN units. They could boost thrust off the pad by a factor of two or more and be dropped when no longer wanted, like the classic Atlas boosters, only much lighter. The direct thrust would be augmented by the elimination of over-expansion losses in the main engine. If a factor of three thrust increase is possible, with lower Isp penalty of course, then a core lower stage could add two strap on tanks of equal mass to the main vehicle. These tanks could be dropped with the tan units leaving the fully fueled original vehicle at high altitude and mach 6 or so, with a very large expansion ratio engine for high Isp. This vehicle could possibly reach orbit without an upper stage.

The applications for newspace are obvious. If a single workhorse engine can have its’ operating thrust range trippled with low development cost, then some intermediate to large engine sizes can be skipped. This would both accelerate the schedule and reduce development costs. Both of those impact the bottom line and the ability to get investors.

A good development team could probably figure out how to incorporate a boost unit like this into a nozzle without screwing up the flow for the later high expansion nozzle use. Then these TAN units would be reusable with hopefully low maintenance costs.

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.

14 Responses to Ejector/Injector TAN Modules

  1. Wesley Johnson says:

    What I can’t understand here is why you want to put the liquids in the solid engine at all. Solids are generally used for their high propellant density, so now you integrate a low density propellant and the required tankage/feedline/thermal control systems into what could have originally been a high mass ratio engine. Not to mention that the conventional solid won’t burn for all that long, so you will probably need a second stage anyways. Additionally, solids generally have higher g-loadings than their liquid counterparts, so adding thrust to a solid motor would only intensify the g-output (though you do add the weight associated with the liquid fuel/ox storage system). It also might be necessary to use a regenerative nozzle system since you will need cooling with the higher temperature gas in the nozzle. I may have completely missed the point though, naysayers impede progress.

  2. johnhare john hare says:

    This TAN concept is separate from the tribrid of the last post and could apply to any high pressure/high expansion ratio engine.

    The liquids in the tribrid would improve Isp from better gas qualities.
    Increase available thrust without the expense of developing a new solid rocket.
    Improve mass ratio with the light liquid tanks.
    Add throttling capability.
    Automatically add a flak jacket to the solid in the form of liquid propellant tanks.
    Dampen vibrations with the liquid mass.
    Control g-loading with additional propellant mass.
    Have the possibility of a regeneratively cooled nozzle lighter than the ablative or heat sinks of most solids.

    Solids are not that high a mass ratio system compared to well engineered liquids. The high density comes with a price, low Isp being one of them. This could well be a far higher mass ratio system with better Isp. The normal short burn duration and required upper stage are legitimate concerns. That would confine the tribrid concept to multistage systems.

    My communications skills being what they aren’t, several of the best commenters here have to get clarification on anything I write that they find interesting. Without their input and interaction, this would be another echo chamber concept idea series. As opposed to Jon Goffs’ posts, most everything I suggest has serious fundamental flaws. It is kind of a game to see who spots them.

    Naysayers come in several styles. Some, like Jim Davis, know more far about the subject than I do. Some, usually anons, throw mud at things they don’t understand. One type is valuable, one usually isn’t, and several can be either depending on the subject.

  3. jsuros says:

    Here’s a thought to extend the “one nozzle to orbit” version of this design. Could you operate the solid gas generator as a series of interconnected chambers with different burn times, dropping the faster burning chambers as they empty of fuel?

    You have already discussed drop tanks for the liquid fuels. Could you interconnect solid gas generators, isolated by some sort of one-way valve?

    You have already mentioned active cooling of the solid throat using the liquid propellants, maybe a similar technique could ease the engineering of solid chamber interconnections?

  4. johnhare john hare says:

    Interconnecting solids is a new thought to me. I’ll mull it over. No promises of course. A friend put me onto a really interesting application of the Coanda principle months ago and I still can’t quite pin down the app properly.

    This particular post was supposed to be a seperate concept from the 319 tribrid, and usefull for a liquid or hybrid also. It’s going to get worse as I stumbled on another varient of this one for the expander deflector nozzle that so far seems as if it would work better and weigh even less.

  5. Tim says:

    As I understand it you’d have TAN combustion inside the inlet/combustor, and hot gases from the main nozzle on the outside. How wold you keep the TAN modules cool? Really beefy ablation?
    Some time ago I came up with a pump (sort of) that works on a similar principle, you want me to email you some pictures?

  6. Habitat Hermit says:

    About cooling these maybe it would be easier to loop them outside of the nozzle having a minimal siphoning inlet as described by the post but immediately directing that flow into a similar system only located on the outside of the nozzle and then letting the combustion products from the additional fuel & oxidizer flow back into the nozzle further down?

    Since I don’t know much about fluid dynamics I’m getting a bit confused about the Venturi effect and supersonic flows, just how is it supposed to work in this case?

  7. I had to read the post in a hurry, so my comments may not be fair or accurate, but here goes:

    I thought the Aerojet TAN concept was straightforward and a model of relative simplicity. I also assumed that the apparent fact that the nozzle had to be oversized off the pad in order for the TAN concept to work was a stroke of serendipitous genius: because you have afterburning going on in the nozzle, this negates the downside of an over-expanded nozzle off the pad, and, because the TAN function is for boost-phase only, the already expanded nozzle is ideal for the rest of the launch sequence.

    Did I get that wrong on the Aerojet TAN? What I’m getting at is that the TAN concept here seems overly-elaborate. Tell me what I’m missing?

  8. johnhare john hare says:

    Work schedule went nuts. Drive by answers for the moment.

    Tim. Regen cool both sides was my first thought. There is a lot of fluid going in relative to the exposed ares. I’m interested, but response might take a while.

    Habitat. I think of that method as the Quasimoto option. Might work a little cleaner and give more volume if required. The ejection here would be a step in the inlet area creating a low pressure shadow similar to what Eric suggested in comments on the last post. The pressure recovery of the supersonic gas would include the propellants.

    Roderick. I didn’t see any actual details of Aerojets TAN concept, just the claims. Until this little series started, I couldn’t see how they could claim a large thrust from such a small volume. It seems possible that using the supersonic hot exhaust gas to vaporize, pressurize, and mix the propellants is on the road to their methods. I don’t see at the moment how else scramjet research would apply to rocket combustion. If you have details of their actual methods, I’m interested.

  9. Tim says:

    Not sure if this was mentioned in relation to an earlier to post, but it occurs to me that discrete TAN modules that can be individually throttled allow thrust vectoring without a gimbal; both from the differential thrust from the modules, and the directed flow from the main nozzle.

  10. Axel says:

    John, here is a question, which may just show that I didn’t understand your idea: what is the difference of the original TAN concept and the one you discuss here?

    Is your nozzle in your figure shaped differently on purpose? Or is it a standard expansion nozzle? http://aerorocket.com/MOC/MOC.html shows speed distribution in a normal nozzle.

    Wouldn’t your idea end up with pretty much the same physical configuration as shown in figure 3 in
    http://stinet.dtic.mil/cgi-bin/GetTRDoc?AD=ADA454615&Location=U2&doc=GetTRDoc.pdf description of TAN engine?

    Would be a pity if your idea is already covered by the Aerojet patent. Reading patents is a pain, but if you are serious about this idea, you should.

  11. johnhare john hare says:

    Differential throttling as you suggest might be a lighter method of TVC than normal. I don’t know though.

    The nozzle is shaped some differently to allow room for the TAN units, or it was a quick sketch to show the idea depending on your viewpoint.

    I went back and reread the patent to see if it showed me anything I missed last time. I still don’t see how they are going to get enough residence time in the chamber for a good burn based on the information available.

    It is my opinion that for sound business reasons they are holding back information critical to the technology. It is easier to protect an idea by the trade secret rout than with a patent. Information not known cannot be reverse engineered. Patents are better for investors than protection in most cases. IMO of course.

  12. “”” If you have details of their actual methods, I’m interested. “””


    This is the most comprehensive thing I can point you to:


  13. johnhare john hare says:

    I’ve read that and the patent and believe there are some serious proprietary details not available to us. Note about halfway through the mention of how difficult is was to get Kero/LOX to perform with such a short residence time.

    If it were a truly easy technology, paying them license fees would be a no brainer as opposed to trying my scheme. I believe they got the results they claim, and that they used advanced cleverness to do it. I don’t believe it would be that easy for a NewSpace start up to duplicate the results under the stated conditions.

  14. John,
    Sorry I’ve been so busy lately I haven’t even been reading my own blog! The tie-in between Scramjet technology and TAN is entirely in the injector design. It has nothing to do with scramjet inlets or other things like that. Basically for scramjets, they need to inject propellants into a supersonic stream in a way that doesn’t disrupt the stream badly (causing shocks and other efficiency losses), but in a way that they mix, burn, and expand, all within an extremely short amount of time.

    But yeah, I’ve spoken with the Aerojet guys. There are some definite trade secrets under the hood. Though the patent they did manage to get is pretty narrow, so there are probably some other approaches that wouldn’t violate their patent, but would achieve similar results…but it isn’t trivial by any stretch. Knowing a bit more about how they make it work, I’m not 100% sure I could get something like that to work efficiently myself. I’m hoping to try soon though, because I think there’s a couple of options that would simplify construction a bit, while still making the overall concept work…

    …ok, enough dreaming for me, back to LLC stuff (and preparing my SPACE 2009 prop depot presentation for next week).


Leave a Reply

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