Cagejet Turborocket

Air breathing engines are virtually useless for the acceleration  mission of a spacecraft. Fairly simple math demonstrates their lack of utility again and again. Any real air breathing engine (ABE) has a limited mach range and large weight compared to a rocket engine performing the same acceleration job. Turbojets work from mach 0 to 3 and ramjets from mach 2 to 6. Scramjets have never demonstrated useful thrust after half a century of off and on development. Rockets can cross the mach range of either real ABE with a mass ratio on the order of 1.5. The 0.5 is the propellant burned during the acceleration and weighs less than the ABE system it replaces. This is because ABEs are heavy and require very precise inlet air handling equipment that weighs even more. Plus the wings and other airframe compromises neccesary to make the system work. Trading a ton of propellant for a ton of engine, inlet, and airframe is a very bad economic move.

The problem gets worse after shutting down the ABE. The rockets now have to carry all that dead mass to whatever final velocity they are shooting for. That extra mass comes straight out of the payload or upper stage. When someone recommends ABEs for spaceflight, it is usually the case that they have not done a systems approach because they are so focused on a single component. Of the remainder that have run all the numbers, most of them are getting paid to produce studies and mock ups, not operational hardware. There are a very few that have done serious ABE work with integrity and they will mostly admit the limitations of their work. As Henry Spenser quotes from time to time, “wind is free, sails are not”. Or as I say it, If air is free, why does the convenience store charge more for it than gasoline?

There is really only one use for an ABE in spaceflight. If a cruise phase is useful to your desired launch system, ABEs are vastly superior in cruise. If you are flying out for an airlaunch or cruising back to base, it is hard to beat an ABE.

If you have to carry the ABE through an acceleration phase, then you look for a much lighter engine even at the expense of fuel economy. The air-turborocket invented in the 1950s is one such system. There are a number of variants proposed that double to triple the thrust to weight of a turbojet. This is at the cost of extreme fuel consumption compared to any jet, though much better than a rocket. Dense fuel Isp ranges from 400-700 during subsonic cruise with hydrogen about 50% better. It turns out that the standard turborocket burns too much fuel for cruising, and weighs too much to carry through serious acceleration.

I spent a good bit of time on the problem before realizing that modifying existing components just wasn’t going to work. Designing an effective ABE for a spaceplanes’ cruise phase really needs a new approach. You need better T/W and Isp than a normal turborocket. Designing around known systems just doesn’t get the job done. What is needed is a turbojet that is really really light for the power it produces. Or an engine that mimics a turbojet/turbofan and is really really light.

Some way must be found to eliminate as much mass and as many components as possible while retaining the desired capability. This usually involves finding ways to have one component do two jobs, and do them better than the original. LANL holds the rights on one patented method for increasing the performance of a turbine engine. The turbine blades are hollow and used as a centrifugal compressor. The air keeps the turbine blades cool enough to allow much higher turbine inlet temperatures than normal. The heat absorbed by the air is used for power (regenerative cooling)as it is mixed with fuel in the burner before driving the turbine. With one part being both compressor and turbine, weight is less. By using higher turbine inlet temperatures, the engine is more fuel efficient and gets more thrust out of each unit of air (specific thrust). Higher specific thrust means that less inlet and duct work is required per unit of thrust. This is a win for everybody if they get it into use.

I stumbled across another way of getting similar results. Some varieties of the squirrel cage fan have the characteristics of a short blade centrifugal compressor, and a short blade radial inflow turbine. By using one wheel as a compressor for 75% of the cycle, and as a turbine for 25% of the cycle, the blades are regenerative cooled, and it is possible to burn stoichiometric in front of the turbine. The T/W and specific thrust could be even better than the one owned by LANL. The catch is that the cycle doesn’t close. The drive gas pressure for the cagejet stage must be higher than the pressure it produces. Enter the turborocket. By using a fuel rich rocket to drive the turbine and burning the excess fuel in an afterburner with the compressed air, a very high T/W  can be achieved with a very simple system. The drive rocket should use about 40% oxygen and 60% kerosene.

The cagejet is capable of using multiple staged rotors in the same manner as a multi spool turbofan except that each stage has its’ own bearing and doesn’t have to share shaft space with other spools. The cagejet compressor can attain higher stage compression ratios than axial flow because the adverse pressure gradient on the blades is countered by the centrifugal force. The separate stages can run at different flow rates to match airflow conditions in ways that even sophisticated turbofans can’t do. Each stage is capable of a compression ratio of 2, so three stages can get a compression ratio of 8. A fourth stage can be driven by the rocket compressing 25% of the air to a final compression ratio of 16 for a fraction of the total flow.

The 16 atm air compressed by the  fourth stage burns stoichiometric with the fuel rich turbine exhaust in an afterburner and is used as drive gas to compress the total air flow in stage three to 8 atm. The exhaust from stage three can be used as is to drive the first two stages and give decent thrust with an Isp in the 1,500 range. Or the 8 atm afterburner can add kerosene for a stoichiometric burn. Then 25% of the flow could drive the first two stages and have an exit Isp in the 50 range including the air. With the mixture ratio of 16, net fuel Isp for 25% of the flow is 800. The other 75% of the flow is used as an 8 atm fuel/LOX/air rocket. Net Isp for this portion would be in the 1,200 range considering the LOX mass. Engine total net Isp at high thrust would be about 1,100. 

T/W for this system would be around 25 if the numbers work. T/W 25 and Isp of 1,000 is the lower end of acceptable performance for a HTHL vehicle. The engine is useless for VTVL, throttling problems and flight profile restrictions. One thing that helps this engine concept is that the cagejet layout naturally lends itself to very thin cages of large diameter. By turning this system 90 degrees from the normal jet engine placement, it is possible to mount these engines inside the wings or tail surfaces. The vertical tail in particular is a desirable location.

The uses I see for this are for vehicles that need medium duration climb/cruise with serious acceleration during portions of the flight. The Rocketplane proposals would benefit from a lighter engine with more thrust. The White Knight series could use something like this to supplement normal turbofans for accelerating off the runway and initial climb. Turn them off during the cruise then light them again for a pop up from 10,000 feet to 60,000 feet for a seriously enhanced release altitude and  attitude. The Air Launch group could enhance an airliner for their carrier.

Perhaps the most useful thing for it would be for testing spacecraft airframes. It would be nice to be able to wring out the airframes early on and have back up thrust during early rocket airframe trials.

This is the same concept I discussed at Space Access 2004 with a couple more rotors. It is a lot less complicated than it sounds. You can get an idea of it by driving a squirrel cage fan with shop air and feeling the wind it throws out the rest of the perimeter. Use a face shield.

The sketch is a simple two rotor system. A four rotor system is actually smaller because the burns are at 16 atm and 8 atm instead of 4 and 2 in the above. The boost rotor only has to handle 1/8 of the volume/second of the main rotor. It can be seen that the individual cages can operate at optimum rpm for the changing conditions of a flight without being slaved to the other stages. It is a simpler method of flow matching than the variable stators on modern turbojets and fans.

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.

11 Responses to Cagejet Turborocket

  1. Jeff Medcalf says:

    One thing I’ve long wondered is whether jets with forced oxygenation through part of the flight regime would be worthwhile. The idea would be to use Kerosene/LOx, but takeoff without the LOx. The jets in the wing roots get you up to altitude, where you tank up with LOx. (The idea is to save on structure weight, not having to take off and land with 90% of your mass or so being oxidizer.) Once you’ve done that, you could light up a conventional rocket in the tail, but to keep the jets from being dead weight, close off the air intakes and feed oxidizer into them so they are adding to the total thrust.

    Would this type of scheme provide a useful compromise, or would the additional weight of inlet flaps and controls and so forth be higher than the efficiency gained by weight savings and using the jets as sorta-rockets?

  2. Bob Steinke says:

    Can a squirrel cage fan really get a compression ratio of 2? I was under the impression that they were used for high volume, low delta-p applications. If it works it sounds like a great idea.

  3. john hare says:

    You can’t get around the fact that current jet engines are heavy. You could possibly get a 1,000 pound jet engine up to 10,000 pounds of thrust with a few tweeks. The extra gear would weigh more also. Or you can use a well developed rocket that weighs 100 pounds to deliver the same thrust with higher Isp. It is likely that you can get better performance and less weight with two specialized engines than if you try to make one do the job of both.

    Your idea has been discussed many times on usenet when the signal/noise ratio was better. You would have to sift a lot of junk to find the specifics though. The proposed concept here could possibly
    use your thought in a bit different profile. By using GOX to pressurize the LOX, the LOX tank is all gaseous oxygen at end of main rocket burn. It is possible that the cagejet could suck the remaining oxygen out of the tank for a little more V. Though even that is questionable because Air Launch? has already tested engines that used the pressurant gas for propulsion after all the liquid was used.

  4. john hare says:


    It wouldn’t be your normal shop fan at 1,400 fps tip speed. Squirrel cage refers to its’ nearest living ancestor.

  5. anon says:

    1. I have wondered about the idea of on board LOX generation for spaceplanes. The propellant transfer ideas have the flaw that they require oxidizer transfer which is problematic. If you could have an efficient on board LOX plant, you could simply transfer more RP-1 and use it to make LOX.

    One idea would even to have a two stage space plane – the other stage would carry the LOX generator. Both would fly to the tanker and fuel up, generate the LOX, and then the first stage would return to base, leaving the fuelled space place to go for orbit.

  6. john hare says:


    Getting the efficient on board LOX generation seems a bit iffy to me if mass is severely limited. While I haven’t given it that much thought, it is hard to see how developing functional aerial LOX propellant transfer would be more difficult than developing an entire LOX generation plant light and reliable enough to be useful in this application.

    If it’s just the cryo aspects that are the problem, H2O2 would seem to be a better choice for reloading oxydizer. Depending on the size craft you are suggesting, A bigger tanker would probably be cheaper than developing new LOX technologies from scratch. Besides, you could get a lot of the initial investigation done with a piece of paper that reads
    pay to the order of Mitch Burnside Clapp the sum of……………..

  7. jsuros says:

    Would this be usefull for aircraft that take off from water, like a flying boat or a wing in ground effect aircraft?

    Also, would you mind showing the four stage version of this engine. I tried to sketch it out but can’t quite see how it would work.

  8. john hare says:


    If it works to spec it is good for anything that needs very high(for air breathing engines) thrust to weight and is willing to sacrifice fuel economy to get it. VTOL military jets that use a dedicated vertical lift engine for instance.

    I will try to get a four stage sketch done in the next couple of days.

  9. jsuros says:

    Thanks for the 4 stage drawing. I had a couple questions, if you don’t mind.

    1) How hot is the 8 atm. flow coming out of stage 3 into the afterburner? Is it cool enough that a lightweight valve could direct it either backward to an accelleration afterburner or forward to a decelleration afterburner? I’m thinking of a very short or very heavy take off and landing application.

    2) Could you replace some of the mass of the air rocket’s nozzle with the 2 atm. flow you are dumping out of stage 1? Some sort of aerospike-like arrangement seems possible.

  10. john hare says:


    1. The 8 atm air has experienced compression heating in the intake if there is one operating at transonic or higher. It has been cooling three rotors from their turbine side heating. It has also experienced compression heating. Depending on assumptions, between 700 and 1,200 degrees f. The duct work will need some heat tolerance. There is no reason the 8 atm air couldn’t be directed to various throttleable thrusters. An 8 atm duct is considerably smaller than one at ambient.

    2. I don’t see why not. The flow from stage one is still delivering an exhaust velocity on the order of 500 m/s. There must be some scheme that would allow them to share a nozzle if desired.

  11. john hare says:


    If you really wanted more down thrust on landing or take off, you could have the 8 atm gas ducted to a burner driving a monster rotor specific to the purpose.

Leave a Reply

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