Cagejet Again

guest blogger john hare

I received a disturbing email today. John Bossard at critiqued my cagejet idea at my request. I have been down for various reasons for the past month or so and mentioned that the criticism had disturbed me until a few unrelated issues resolved. He apologised for torpedoing my idea. That’s disturbing, bad ideas and poorly explained ones need torpedoing to make room for something else.


This picture of my EZ Rocket model after I got it out of storage is symbolic the last month or so. It is the first or second scale model that XCOR had made as far as I know, and now it is filthy and has pieces missing and broken. When my attitude is bad, I either have few ideas, or stupid ones like the IRS/NASA joke. Now that a few issues are on the way to resolution, I’ll try blogging some more.

The idea that John B. critiqued for me was posted in October 2008. It is the cagejet variation on the turborocket. When a legitimate critique is received, the proper response is to address the issues, not the emotions. I have paid folding money for worse information on occasion.  He gave me permission to address the issues he raised in a public blog post.

First problem he points out is that the burn would be at a fairly low pressure with inherent poor efficiency because squirrel cage fans are high volume/low pressure ratio machines. The radial type compressors obtain their pressure rise in the diffuser downstream of the rotating machinery. I was seeing it as a uniform distribution sideways axial type flow with better boundary layer behavior. While I didn’t draw and label diffusers in the post, I don’t see them being left out in the first serious design pass. While he is correct that I didn’t address this point, I think any reasonable designer would correct it early.

Second major problem is that radial flow compressors, of which this is a subset, have large frontal areas compared to axial machines. My solution is to turn the ‘wheel’ sideways and mount it in the vertical tail. This has the multiple advantages of making the machine frontal area smaller than the inlet, making use of an area that is normally empty, providing a structural justification for beefing up the tail that is not ‘charged’ to the airframe, providing a reason for increasing the size of the vertical tail that is not ‘charged’ to the airframe, and placing the inlet drag above the center of lift to reduce supersonic trim drag.



 In the cartoon above I have the vertical tail section below the vehicle with air coming in left to right. There is a fixed supersonic inlet at the leading edge of the tail. A normal shock just before entering the center of the cage. The volute, which must be well designed as John B said, is inside the tail also. In the upper sketch you can see the afterburner chamber right at the root of the tail. I believe this layout would bring out the best in the concept.

The things I left out of the original post make John’s opinion correct. The things I add in here might make the difference.

He also mentioned that my Isp values for ATRs was much too low. The proper hydrocarbon bi-prop Isp should have been in the 1,200-1,500 second range.

I believe the proper application for this concept is as a flyback or air launched fly forward first stage. The vehicle climbs out of an airport to safe airspace on ATRs alone. In the clear, the ATRs are throttled up all the way for acceleration to mach 2-3 at 15-20km altitude. The rockets are then lit in addition to the ATRs for the acceleration to ~mach 6 at 40-60km altitude when staging takes place. The upper stage has about 6,000 m/s to go for reaching orbit.

Mass ratio of the first stage should be less than two. Torpedo’s away, if I can’t handle criticism, I have no business posting controversial ideas here.

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.

15 Responses to Cagejet Again

  1. Axel says:

    For everyone who missed the original posting (like I did):

  2. johnhare john hare says:

    Thanks Axel,

    That was rude of me, sorry guys.

  3. John Bossard says:

    John H,
    I am generally reluctant to criticize any novel or unique concepts, especially without a doing my level best to obtain as good an understanding as I can of the concept. I would not have provided a public comment at all, but you had requested it, as you mention in this posting.
    Hopefully, I can provide some useful, productive commentary, that may enlarge the dialogue, rather than shrink it. But mine is by no means the final word on the possible utility or workability of your concept.
    I’m sure you have other colleagues that you can, and should, consult with as well.
    I will be happy to continue to work with you on your concept. Maybe the best thing I could offer would be to suggest ways in which you could evaluate your concept, and what metrics might be meaningful to order to quantify performance. Just some thoughts.

  4. Mike Lorrey says:

    Okay, I’d like to criticize your original post a bit. Ramjets are not limited to mach 2-6, they can operate from any speed above 0 mph, and can go over mach 6 if you utilize certain technologies. Most ramjets used to date are designed for optimal performance at a given speed for whatever mission they are intended for. If you want a supersonic cruise missile, mach 1.5-3 is sufficient. If you want a bomber interception SAM, mach 3-4 is cool. If you need a small AA missile for dogfighting, mach 4-6 is needed.
    Ramjet Isp peaks at around mach 2 at about 1800 seconds for the most efficient kerosene designs, but it slopes down on either side, with 350 seconds at Mach 0.5 and 1000 seconds at peak transition to supersonic combustion.
    Back in the 40’s when ramjets were first being developed, the French developed the LeDuc 010 interceptor prototype (google it). It was a subsonic interceptor that was launched from the back of a DC-4 style transport/bomber at about 200 knots. A few later models were developed, and the Nord company got into the competition with their Griffon II interceptor in the 50’s which looks like a Large Mouthed Falcon (F-16) and featured a small turbojet to assist takeoff but was propelled mostly by the ramjet and could reach Mach 2 between takeoff and 50,000 ft.
    I would suggest that rather than a squirrel cage compressor that you use a two sided Francis Turbine, which accomplishes the same goal but is much more efficient and allows a higher compression ratio.
    However, IMHO you are not going to get away from drag losses just because you’ve turned things 90 degrees. The intake is still going to exhibit shock and compression drag, and you will still need sufficient inlet area to equal the same inlet cross section as a normal face on turbine compressor.

  5. A_M_Swallow says:

    At less than 100 mph you can ground tow a ramjet vehicle along the runway. There is no rule that says the first stage has to leave the ground.

  6. johnhare john hare says:

    A ramjet is the simplest engine known being basically a stovepipe with inlet and nozzle. It has been known for about as long as the modern rocket engine era. Don’t you think it odd that none of the New Space companies are even considering them?

  7. Mike Lorrey says:

    Yes it is odd, esp given their technological simplicity and low cost. The only reason I can think of is that its basically impossible to come up with new ramjet technologies to patent, investors want to see tangible value as a result of their investment, which patents can be, but building something thats public domain doesn’t have much potential for. Another thing is, you build a ramjet in your hangar, an investor comes by and looks down the intake and sees…. hangar through the tail pipe… and starts to wonder where his money went…
    I also think that most of the newspace people are dissilusioned rocket engineers leaving NASA or one of the major contractors, which are a different breed and community from the ramjet/hypersonics community. The Newspace guys should do some recruiting among the old Marquardt alumni.

  8. johnhare john hare says:

    Most of the stuff they are working on is not patented and can’t be. That answer doesn’t pass the first toll booth.

    Investors want something that turns profits and the simpler, the faster, which is good for ROI. Strike two.

    Very few of the newspace people I know have left major contractors, and none at all from NASA.

    Many of them though have run the numbers on ramjets just as I have though and found them not useful. It is almost the first alternative we all look at to get out from unther the rocket equation. The problem is that the engine sounds simple and cheap, but isn’t. New engine cost is largely in the testing and debugging. Ramjets are far more expensive to test than rockets. Ramjets are also very finicky about the incoming air, which makes the inlet a major engineering project. Ramjets give you five mach numbers if you really stretch them, and the integration issues will kill you. That ‘simple’ engine requires a sophisticated and relatively heavy airframe and constricted flightpath compared to a pure rocket vehicle.

    Just because ramjets aren’t on the list doesn’t mean it’s because we are all close minded. We’ve looked.

  9. Mike Lorrey says:

    Again, its about different communities. Rocket guys are basically plumbers. Ramjet guys are aeronautical types. Either way, I am of the opinion that most engineers today hamstring their companies with analysis paralysis, arguing that modeling something will save money, but in the end it really doesn’t and you wind up spending more money and have less or zero hardware to show for it.

    There’s plenty of ramjet production models out there that work just fine. The Russian GNOM missile, for instance, uses an air augmented rocket ejector to boost its range 80% for only a 20% weight penalty. The French have solid fuelled ramjets using boron fuels to propel their equivalent of the Pheonix missile (Russians do the same thing with their Adder missile, at over mach 5), and the USAF is looking to adopt the same propulsion module for an extended range AMRAAM.

    I agree compared with simple ammunition like ballistic rockets, ramjet propelled lifting vehicles can more complex, if you think too much, but they don’t have to be. Looking at the market, I see VG and XCOR as the leaders and both work with winged vehicles. If that pattern holds there has to be a place for ramjets to boost average trip Isp.

    But talking complexity, this cage jet concept is taking that to new heights.

    However, compared with multistaging of liquid fuelled systems, ramjets are a lot simpler, even with adaptive intakes and exhaust.

  10. johnhare john hare says:


    I think you are dead on about the analysis. The guys that are getting things done are getting their hands dirty.

    Have you gone to any conferences to discuss ramjets with the rocket guys? It seems to come up a time or two at every one I’ve been to and the conclusions are that it doesn’t close. You just might be bringing new insight to the problem. I have bought the papers on disk from the conferences of the ISOABE and several hardback references on inlets and ABE high speed engines. Many of the people that attend those conferences have more.

    I won’t fight over the cagejet or any other concept to the point of heated arguement. I throw ideas out there to see if there’s any meat on them. When something useful comes up, I look for ways to push it a bit further.

  11. MG says:


    The NewSpace industry needs more players. I look forward to seeing your ramjet fly.


  12. Mike Lorrey says:

    The missile guys tend to focus on weight performance. Ramjets get T/W of 30-40, while hydrolox averages about 50 and kerolox over 100. One thing to keep in mind is that turbine engines only get about 6-10 T/W. To compare them all you need to handicap each by the average Isp.
    Turbine: T/W: 6-10, Avg Isp: 3000 Sec, Handicap rating: 24,000
    Ramjet: T/W: 30-40, Avg Isp: 1500 Sec, Handicap Rating: 52,500
    Hydrolox: T/W: 50, Avg Isp: 450 sec, Handicap Rating: 22,500
    Kerolox: T/W: 100, Avg Isp: 300 sec, Handicap Rating: 30,000
    It is thus clear that looking at both T/W and Isp together, Ramjet performance is significantly ahead of Turbines and hydrolox engines. Kerolox is better than both as well but not quite up to par with ramjets.
    Thus, any argument by an engineer that the case for ramjets “doesn’t close” fails because if the case for ramjets fail for use in a zero stage/thrust augmenter, then the case for all the other propulsion options is much worse, and they shouldn’t close either.
    Most multistage rockets dump their first stages below the max altitude for ramjets, and below their max speed. Even Falcon 1, which is pretty high performance, dumps the first stage at 135,000 ft and mach 8.
    What I’d like to see for an experiment is to put an air augmentation ejector cowling around the base of the Falcon 1 first stage, and run the engine richer even than it normally runs. Kerolox engines run fuel rich as a matter of course usually because pure stoichiometric combustion is extremely hot and risks burning out the engine, so there is already surplus fuel in the exhaust stream for an air augmentation system to utilize. If you run the Merlin a bit richer than normal, with a slightly fuller than normal fuel tank and slightly less LOX than usual, I think SpaceX would see a significant boost in payload to orbit.

  13. Jonathan Goff Jonathan Goff says:

    The key problem as I see it is that you’re not doing a fair comparison. In order to stay in the atmosphere that long compared to a rocket, you need to have bigger wings, much better TPS (by the time a rocket is going even Mach 3, it’s typically out of most of the atmosphere), more drag losses, etc. Ramjets might make more sense if you were trying to do a suborbital point to point (or some other sort of cruise mission), but for an orbital mission, staying down in the atmosphere just makes everything more complicated. I’d much rather build a better rocket, than try to build a hypersonic airplane. Building a vehicle that can fly around at Mach 6, and is reusable is something that even the military (which cares about cruise missions) has had to spend a lot of money to come close to…it really doesn’t sound like a good way of reducing the cost of orbit.

    Now, there may be some niche orbital situations where small airbreathing engines for landing purposes or such might make sense, but I’m completely unconvinced that a ramjet is going to be of any real use for ascent. It’s more pain than it’s worth. Isp isn’t everything.


  14. Martijn Meijering says:

    In order to stay in the atmosphere that long compared to a rocket

    Would staying in the atmosphere for long enough really be a problem? If at some stage during ascent you use only ramjets and no rockets that sounds plausible. But what if you only used ramjets simultaneously with rockets? That wouldn’t give you a spectacular Isp improvement, but it might make cheaper, lower performance propellants and engines competitive and reduce mass fraction constraints on your orbiter stage.

  15. Mike Lorrey says:

    Well I understand the issues with staying in the atmosphere. There have been a number of TPS improvements since Shuttle, which I’m sure you are aware of: X-33’s metallic stuff, SHARP, etc. I do understand the air breathers problems of greater aerodynamic losses for a winged vehicle, etc.

    However, even if we stick to a two staged kerolox ballistic rocket, I fail to see how using an air augmented ejector cowling around the rocket engine would not significantly improve the rockets performance and orbital payload. The proof of concept at least suborbitally is the GNOM missile, the development program tested its PR-90 rocket:

    “The concept was evidently been proved on a subscale tactical missile, the PR-90, now on display at MAI’s museum at Orevo, north of Moscow. The PR-90, with a launch mass of only 1500 kg, of which 550 kg was payload, could reach an altitude of 40 km and a range of 100 km. The booster unit used 200 kg of RAM-10 ballistite with a specific impulse of 180 seconds to get the missile up to ram-air ignition speed. Then the air-augmented unit, with a specific impulse of 550 seconds, cut in and used 300 kg of propellant to boost the vehicle to its 1 km/s cut-off speed. An equivalent liquid propellant missile (such as the American Lance) weighed over twice as much. A solid propellant equivalent (such as the French Pluton) would weigh three times more. ”

    When it comes down to the nuts and bolts, other than Gnom and the French Onera program, as well as retirees from Marquardt, I don’t see anybody testing hardware today. The examples that have been flown IMHO disprove the unsupported claims of detractors.

Leave a Reply

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