Aluminum/Magnesium alloys can burn quite nicely. The stoichiometry is favoriable. Instead of $CO_2$ which is mostly oxidizer by mass, you might end up with $Al_2O_3$ and $MgO$ as exhaust.

The question Using lower stage as reaction mass and answer got me thinking - what if reaction mass referred to a chemical reaction and not just a kinematic reaction?

However I can't imagine any possible way to actually cannibalize the structural material of a spent stage and then burn it. One problem is that the metal oxides of the exhaust are basically "ceramics" and even at very high temperatures they might agglomerate into nanoparticles, basically "glass soot". They'd be hot, but there may be little gas or volume and therefore little pressure and therefore little thrust.

Has this every been considered & written about in any semi-scientific way?

  • 2
    $\begingroup$ Unless you're going with NTR, whatever savings you make in fuel, you'll likely lose in extra oxidizer and worse ISp of non-optimal fuel. Which would have better ISp: 1 ton of hydrogen + 6 tons of oxygen (+1 ton of scrap you dispose of in the beginning) at 450s of ISp or 1 ton of scrap-based fuel + 7 tons of oxygen at 350s of ISp? Personally, I believe it would be (eventually) feasible to bring the launch stages to an orbital refining station and turn them into fuel for interplanetary missions. $\endgroup$
    – SF.
    May 31, 2016 at 7:20
  • $\begingroup$ @SF. I always like quantitative comments - thanks! Aluminum/Magnesium came to mind because they are both useful structurally and release a huge amount of heat when oxidized (compared to anything other structural material I could thing of). If you were willing to take a little more space and flesh out the ISp discussion as an answer, that would be great - I really just don't "get" ISp, and working an example I'm interested in would be great! Also, I like your idea of doing the consumption/conversion at a "facility" in orbit much better than trying to make a rocket do it in-flight! $\endgroup$
    – uhoh
    May 31, 2016 at 8:48
  • 2
    $\begingroup$ In this equation, $I_{Sp} = { v_0 \over g_0 }$ units: $[s] = {[m/s] \over [m/s^2]}$ - with $g_0$ being an unchanging, dumb adjustment constant (gravitational acceleration on Earth, regardless of where the ISp measurement occurs) ISp depends directly, and only(!!) on exit velocity of reaction mass. Now, with $E_k = 0.5 mv^2$, burn fuel to produce X Joules of energy. If you raise the mass of the accelerated particles (use metals instead of hydrogen) you reduce achieved velocity. Even very high energy reactions don't help squat if they eject a lot of reaction mass, more slowly. $\endgroup$
    – SF.
    May 31, 2016 at 9:03
  • 1
    $\begingroup$ (and this comes from conservation of momentum: $mv_{rocket} = mv_{propellant}$. If you eject the same mass of propellant slower, you're not gaining as much momentum for the rocket. And after the initial atmospheric flight, density of the fuel doesn't mean nearly as much as its mass. So 100 tons of hydrogen to orbit vs 100 tons of zinc isn't that much of a difference in cost despite considerable difference in size. $\endgroup$
    – SF.
    May 31, 2016 at 9:09
  • 2
    $\begingroup$ ...especially that with larger exhaust mass comes higher thrust = higher speed gain early on, which immensely affects travel time. $\endgroup$
    – SF.
    May 31, 2016 at 9:17

3 Answers 3


This Scottish/Ukranian design doesn't use a metal body, opting for polypropylene instead, but it consumes that as it goes. The body contains a mix of the oxidisers ammonium perchlorate and ammonium nitrate.

They've published a paper "Autophage Engines: Toward a Throttleable Solid Motor" in the Journal of Spacecraft and Rockets, July, Vol. 55, No. 4 : pp. 984-992 (preprint)

As first stages (from earth at least) don't tend to get into orbit, using them after they've expended their fuel would be a tricky proposition - but build an annular version of this engine, add a second plastic liner and some not-even-balloon tanks inside that....

The references section of the paper contains a number of interesting looking entries:
Damblane, L. "Self-propelling projectile." U.S. Patent No. 2,114,214. 12 Apr. 1938
Typaldos, Z. A. "Autophage rocket." U.S. Patent No. 3,250,216. 10 May 1966
Corbett, M. J., and Belisle, J. A. "Single stage autophage rocket." U.S. Patent No. 4,703,694. 3 Nov. 1987
Yemets, V. V., Dron’, M. M., and Yemets, T. V. “An extremely small autophage rocket for orbiting a pico satellite.” Вісник двигунобудування, Vol. 1, 2015 and Yemets, V. V., Prince, S., and Wilkinson, R. “Investigation of a Combustible Inertial Launch Vehicle Design.” Journal of the British Interplanetary Society, Vol. 68, 2015


The main problem is mechanical in nature. Getting the structural into a form that can be burned and getting that material funneled through a nozzle while burning.

Any thrust that can't be directed ranges from useless to harmful to a spaceship. Since when you burn the structural material, you are currently burning up what could direct the gas flow, you have a slight problem at your hands...

Other than that, Aluminum and Magnesium ARE used already as fuel. Solid booster rockets usually use Ammonimum perchlorate composite propellant. One portion of this composite is a high energy fuel like Aluminum, Magnesium or Zinc see Wikipedia

  • $\begingroup$ They are used as additives or supplements only. They don't make much gas by themselves, but they can heat the exhaust produced from the combustion of the main components of the SRB fuel, as far as I know. And you are right - you shouldn't burn your nozzle first! $\endgroup$
    – uhoh
    May 31, 2016 at 13:35
  • $\begingroup$ They supply the energy to propell the gas out of the nozzle at as high speed as possible. Higher temperature within a gas means higher pressure. Funneling higher pressure through a nozzle means more impulse. $\endgroup$
    – Adwaenyth
    May 31, 2016 at 13:38
  • $\begingroup$ Yep - what gas? $\endgroup$
    – uhoh
    May 31, 2016 at 13:39
  • $\begingroup$ The low energy fuels mentioned in the wikipedia article. The organic composites burn to gases that are heated further by the metal components - which is exactly the problem. If you burn the outer shell of the spaceship, how would you transport that energy to anywhere it could be useful. $\endgroup$
    – Adwaenyth
    May 31, 2016 at 13:41

This would not work for the first stage, but the idea is so interesting I'm compelled to add this as a supplementary answer.

A future electric propulsion system with high Isp using a variety of solid propellants is being developed by Newmann Space:

According to their science page:

What fuels does it use?

We’ve tried out all kinds of materials as fuel. We’ll be posting a showcase about several of them over the coming few weeks, and I’ll add links to these posts as they appear. The list below contains a non-exhaustive list of the things that we’ve tried:

  • Molybdenum – our fastest fuel – best for sending people to Mars
  • Magnesium – our most efficient fuel – best for sending equipment on long missions
  • Aluminium – best for recycled space junk
  • Carbon – our most interesting fuel – reusing * ahem * waste-products from astronauts to get them to where they’re going
  • Titanium
  • Vanadium
  • Tin – a pretty poor fuel that we learned a lot from
  • Bismuth – our most useless fuel… just don’t bother

The potential to use things as diverse as aluminum and poop as propellants is compelling! But this would not work on ascent, the total thrust is very low, and would need to be used in orbit or in deep space.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.