Why the 'Autophage Engine' is nothing more than an interesting thought experiment

Ok, so after my last article about 'fantasy' space projects seemingly getting funding from both government sources and private investors, I faced some well deserved push-back on my opinions about why the example given, the autophage rocket engine, can never actually be anything more than an interesting thought experiment.

There are a number of these kind of wild ideas related to launch that seem to get re-visited at regular intervals as new generations come across the same basic seed idea that, on paper, looks like it could lead to significant advantages for launch technology. The experienced rocket engineers generally look at these ideas each time they are regurgitated and laugh and wink at each other and share the reasons WHY these ideas have never resulted in viable hardware or any actual useful products or services entering the space market. But the appeal of those 'unicorn' projects never ceases to attract those, often from a more academic background, who still believe that sheer determination and application of superior intellect can overcome ANY practical hurdles and that THEY will be the ones to finally realise those theoretical gains such ideas seem to offer. Sometimes, those enthusiasts are sufficiently charismatic and committed to the idea that they are able to persuade others to invest in it and, in the current 'gold-rush' age of space entrepreneurship, there are an awful lot of people who have found themselves in a position to give money to space projects, but who have relatively little practical engineering experience, who are willing to ride that wave of blind enthusiasm and put their trust in such projects in the hope that they are going to be the one that spotted the unicorn which everyone else thought was just a knackered horse with an ice cream cone on its head.

With all this in mind, and although there are plenty of other such ideas getting funding (like rockoons, centrifugal catapults etc, which are all equally guaranteed to never result in a launch to space, let alone a commercially viable service, for practical and legislative reasons), it was the autophage engine that I used as my example and so it is that which I feel obliged to 'debunk' rather than just leave it as my unqualified opinion that it's a dead end from a practical perspective.

There are in fact MANY reasons why the autophage idea can never work, but any one of them is usually enough by itself, so I'll only bother to cover the most important ones. To clarify: I am not saying it is impossible for an autophage to ever be made to run, or even to possibly fly, but I can state with absolute certainty that an autophage engine powered vehicle will never have any advantages over a conventional rocket of the same size/mass and, as such, any claims that it is a technology worth pursuing as anything other than an interesting thought experiment, is false.

0: The Comparison:

So, what's our yard-stick for this evaluation? The team working on the autophage stated that they genuinely believed the system had applications in small or 'micro' launchers, so we're talking launchers in the <20ton range with a payload mass to LEO of <200kg or so, or maybe <100kg in the 'micro' class. The yard-stick would be a 3 stage micro-launcher probably using conventional but very refined Lox/Kerosene engines with a sea-level ISP of around 300 seconds, possibly pressure fed and therefore needing fairly structurally sound tanks making up it's main airframe, probably made from thin stainless steel or filament wound carbon composite (class 5) tanks.

1: The Propellants:

Lets start from the main (only?) stated advantage of the autophage idea: That being that an autophage rocket can do away with the airframe/tanks containing the propellants (which in our conventional small launcher would account for, at most, maybe 15% to 20% of the first stage or second stage mass at launch) by using the fuel AS the tank/airframe. So, to be clear, that's what we're chasing here, a 15 - 20% gain in mass fraction on the first stage/s. That's the unicorn's horn in this whole debate. For an autophage to therefore work and achieve that goal, the system must be a kind of hybrid, having an (initially) solid fuel and a liquid oxidiser.

Having worked with hybrid development for 26 years or so myself, and having worked recently with some of the companies making the greatest progress with them globally for launcher use, I know that Lox/Paraffin is about the best choice of propellants for a large hybrid. This combination gives a theoretical ISP of around 250 - 260 seconds in a well designed engine at sea level. A little way off from the 300 seconds + that a well designed sea-level Lox/Kerosene liquid engine can produce, but still pretty reasonable. The trouble is, neither of these propellants can be used in an autophage.

Paraffin is not structurally very strong. Cast into the inside of a metal or carbon-composite combustion chamber it will hold up to flight, acceleration and combustion loads pretty well, but it won't work as a pressure vessel wall (as would be needed in an autophage) and wouldn't work as the outer airframe of a supersonic launch vehicle. Imagine casting a space capable rocket out of hollow candle wax, it's just not practical. What is needed is a structural thermoplastic, high density polyethylene is probably the only practical option that meets the density, structural and performance needs.

Lox is great but as a cryogenic liquid at -200c it will drastically alter the structural properties of any plastic it is in direct contact with. HDPE for example will become brittle and fracture under any amount of pressure if chilled to that temperature. As such, a cryogenic oxidiser is also out of the question in an autophage design where the oxidiser is stored directly inside the tubular fuel-grain/airframe.

The best option you are left with is probably HTP (90%+ purity Hydrogen Peroxide) and HDPE as your propellant combination for an operational autophage. If anyone has indeed come up with a better combination then I'd love to know, but certainly from an ISP perspective, that's probably as good as it could ever get, and that combination has a sea-level ISP of about 235 - 240 seconds.

So, comparing our 'ideal' autophage launcher with a conventional Lox/Kero small launcher we have already gone from a 300s ISP to a 235s ISP, a loss of over 20%. Remember, our maximum potential gain in this whole project if we created a 'perfect' autophage launcher was up to 20%. Any advantage is lost at the first hurdle before we've even begun to look at the engineering challenges of actually making such an engine work.

2: The combustion chamber:

So, the autophage idea, if ever turned into a practical launcher, has to be a hybrid, of sorts. In a conventional hybrid, the combustion chamber is a long cylinder containing a solid tube-like fuel grain which supports combustion along it's entire internal surface as the surface of the fuel vaporises and combusts with the liquid oxidiser. In an autophage, the fuel is a plastic tube which forms both the tank for the oxidiser and the airframe of the whole rocket. The solid, cold plastic is 'fed' into a heated ring/chamber which has to almost instantly change it from a solid to a liquid state and then inject the liquid into the combustion chamber along with the liquid oxidiser.

Instantly we have a few unusual problems. To melt a thick (probably at full scale, 80mm thick or more?) tube of HDPE to a fully liquid state, you need to transfer a LOT of heat to it. HDPE melts at about 130 - 140c . Once the engine is running it is of course hoped that the heat of combustion will do this job but to get the engine started, the heat has to come from somewhere else. This 'ring' has to be pre-heated to a high enough temperature to melt at least 50kg - 60kg per second of HDPE in order to even start the engine running. Realistically, feeding it into a volcano is unlikely to manage that rate of melting, and the point at which the state change occurs also has to be VERY tightly controlled and it has to happen over a very short distance. You need a massive power source on the launch pad to provide this initial heating, another massive cost/practical disadvantage.

For practical reasons, the plastic tube say, 300mm above this heating ring, has to be at 50c or lower in order to maintain it's structural properties to contain the oxidiser inside at a high enough feed pressure for the engine to run. Het transfer to an 80mm thick block of plastic that melts it to a liquid at that rate, whilst not allowing any heat transfer less than one diameter up the tube that could affect its structural integrity is a near impossible engineering feat.

Even if this were possible, you then have the problem that you are not dealing with two similar viscosity liquids as you are in a conventional liquid engine, you are dealing with a thick plastic 'sludge' trying to burn with a liquid oxidiser. Given the complexities of how the boundary layer interaction works in a conventional hybrid, I'd suggest that getting enough fuel to combust fast enough with the oxidiser, whist keeping the fuel grain thin enough walled in comparison with the volume of oxidiser inside it, and balancing all of that with the structural needs of the plastic fuel grain in it's role as the walls of a pressure vessel AND the airframe of a supersonic space launcher make this an equation with no viable solution.

Then you have the problem of contamination and sealing....

3: Preventing flash-back and explosion:

A rocket engine is just an explosion you are constantly trying to engineer out of happening. The usual way to do this is to only allow a controlled amount of fuel and oxidiser to meet at any given time and to keep the stored propellants separated by VERY non-combustible tanks/walls/plumbing at all times. In an autophage you are forced to break the cardinal rule of rocket engine design and put the oxidiser directly inside the fuel with no barrier at all to prevent them interacting. As your massive, invariably slightly uneven extruded plastic tube is forced down into the combustion chamber, what are the odds of it carrying traces of oxidiser on the inside surface? What are the odds of that oxidiser causing small flashes of ignition as the plastic is heated and what are the odds of you being able to seal the whole system up-stream to stop any of the heat or those small pockets of unwanted combustion back into the tank, causing the entire rocket to explode?

Odds are, any autophage is going to ultimately result in a flash-back explosion of the whole tank/vehicle within seconds. Certainly the odds of maintaining enough of a seal around that big an area to prevent it for the full 3 or 4 minute burn of an orbital first stage are miniscule at best. The odds of you persuading any commercial spaceport to let you fill a massive fuel grain with oxidiser on their launch pad and heat it up are low. You are literally making a bomb.

4: Propellant feed:

This is the real killer, and the one which I think (alongside the propellant choice constraints) really should make anyone sit up and say "Yeah, no amount of academic qualifications can solve that one": How is the fuel (and the oxidiser it contains) fed down into the combustion chamber?

I've genuinely heard people suggest 'acceleration' and/or 'gravity' as the answer to this but, I don't think that was ever a serious suggestion so I won't bother explaining why we don't have any rockets that use acceleration or gravity to feed their propellants yet :-)

For the autophage idea to work the whole plastic tube containing the oxidiser needs to be 'pushed' (or pulled) down into the combustion chamber generating enough pressure to keep the propellant flow at a rate to maintain efficient combustion AGAINST the pressure of the combustion itself. Pumps are pointless as that would only work on a low viscosity liquid, pressure feed (in the conventional sense) is impossible due to the fact the pressure inside the oxidiser tank would be trying to push the plastic fuel tank up and away from the chamber, not into it.

The only answer is that there needs to be some mechanism, attached to the top of the combustion chamber, that 'pulls' the whole fuel tube/oxidiser down towards the chamber with an enormous amount of force. I admit, I haven't had time to work out how much force is needed but, hopefully anyone with a little imaginations and a basic grasp of materials will be able to picture a 6m tall, 400mm diameter, 80mm wall thickness HDPE tube filled with Hydrogen Peroxide needing to be crushed at a rate of about 35mm/second into a heated ring against a few hundred psi back-pressure from the chamber. This process also has to happen completely evenly, the tube must stay absolutely straight as an offset of even a degree or so as the tube feeds into the chamber would drastically affect the flight of the rocket as that would equate to a constant 1 degree thrust offset.

Such a mechanism is going to be huge and need to be very powerful. It is going to need immense power which is going to have to come from somewhere. In the demonstration a hydraulic or pneumatic cylinder is used, a massive 6m long cylinder could possibly be used inside the tank, or a motor running a massive screw thread that pulls the top bulkhead down a 6m long threaded central rod could work, though that would leave the rod/s sticking out of the top of the booster so the length of the rocket wouldn't actually reduce.

No matter how you look at it, or what solution you propose, such a system, if it can have enough power to perform the task of crushing the propellant stack down into the chamber at that scale, is going to be extremely heavy. Far heavier than a thin-walled steel or composite shell over the outside of the rocket would be certainly. This alone means that it can never be possible to achieve any mass savings using such a system, because the mechanism that crushes the rocket down into the chamber will ALWAYS weigh more than the alternative tanks in a conventional rocket, or the system can't possibly have the power to achieve the task.

Ultimately, having flown over three thousand rocket with L/D's of between 5/1 and 45/1, I can assure you that the control methodology for rockets of drastically different lengths need to work drastically differently. Trying to achieve real time flight control of a rocket that is constantly getting shorter, and which therefore has a constantly changing CG/CP relationship, is just the final straw that breaks this particular unicorn's back.

So, an autophage rocket can never be built that actually works, doesn't explode, and has any advantage what so ever over a conventional launch vehicle of the same size. It's just very basic physics and engineering and that can be proven with maybe half an hour of top-level thought and calculation about how such a system could actually be built.

That in itself isn't the problem, nor is the fact that a sub-scale demonstrator of such an impossible system might actually be built and tested. As has been mentioned, the pursuit of impossible projects may well yield useful side discoveries which have applications in the real world, so I'm certainly not against such things being built. My problem is with anyone gaining funding for such projects by claiming that such a thing could yield a genuinely useful technology or could actually really 'work' in the long run.

Chasing unicorns yourself is fine, especially if you are actually hoping to find new ways to attach ice cream cones to horses as your actual goal, but trying to get money out of the government or investors by persuading them the unicorn is REALLY out there, is just not right in my opinion.