As I was watching Tom Mueller describe the Merlin engine, he said that the RP-1 propellant is pumped through ribs in the nozzle to regeneratively cool the engine. Also, a video about Copenhagen suborbital's TM65 rocket engine shows how the alcohol fuel is used to cool the engine. Since both of these engines use liquid oxygen as their oxidizer, why would they use the fuel to cool the engine. Why not use the very cold liquified oxygen instead of the room temperature hydrocarbon fuels?

  • 12
    $\begingroup$ In short "If stuff goes wrong, the fuel would burn, where the oxidiser explodes" $\endgroup$
    – Criggie
    May 4, 2017 at 6:43
  • 1
    $\begingroup$ Does raising the temperature of the fuel increase the efficiency in the combustion chamber? If so, this could be a case of killing two birds with one stone. $\endgroup$ May 4, 2017 at 17:17
  • $\begingroup$ In normal internal combustion engines, you use extra fuel to cool the engine. Allowing it to run lean (more oxygen, less fuel) will cause it to run hotter. $\endgroup$
    – rpmerf
    May 4, 2017 at 20:54
  • 6
    $\begingroup$ Fuel doesn't burn without an oxidiser. The rocket isn't made of oxidisers, so it's perfectly safe to use fuel as coolant. Oxidisers burn fuels. The thing is, at high temperatures and with strong oxidisers like liquid oxygen, almost everything is a fuel. Remember rust? That's oxygen oxidising iron/steel (and it does the same thing with pretty much any metal). Now imagine a much more powerful oxidiser than atmospheric oxygen - that's how you get things like thermite. Fuels are dangerous in air, since there's plenty of oxidiser around to let them burn or explode; oxidisers are always dangerous. $\endgroup$
    – Luaan
    May 5, 2017 at 8:22
  • 2
    $\begingroup$ Saying that LOX is colder is a fallacy, since at the temperature of the combustion cooling zone, the temperature will be hundreds of degrees, at which point oxygen is a gas under any pressure. You are not using the intrinsic temperature of the propellants as a cooling factor, but the heat transport capability, which is higher in the liquid. $\endgroup$
    – MikeW
    May 9, 2017 at 11:25

3 Answers 3


Whatever you use as a coolant will become hot. Hot oxygen will (a) vaporize, making the plumbing somewhat more difficult, and (b) react with and erode (or maybe even ignite) the cooling channels, unless they're made of special materials.

  • 4
    $\begingroup$ Not sure about the vaporize part - the h2 coolant in the SSME certainly did - but the reactive oxygen part is spot on. $\endgroup$ May 4, 2017 at 4:02
  • 48
    $\begingroup$ To rephrase b): "Because oxidizer tends to oxidize stuff". $\endgroup$ May 4, 2017 at 7:52
  • 3
    $\begingroup$ Most engines operate at pressures well over the critical pressure therefore neither fuel nor oxidizer will vaporate. E.g. Merlin operates at 6,14 MPa or 9,72 MPa chamber pressure while oxygens critical point is 5,04 MPa. An example for fuel would be hexane with a critical pressure of 2,99 MPa. $\endgroup$
    – Christoph
    May 4, 2017 at 8:45
  • 33
    $\begingroup$ Put another way, without careful metallurgy and design, your propellant mixture may end up swinging toward engine-rich combustion... $\endgroup$
    – Tristan
    May 4, 2017 at 14:04
  • 4
    $\begingroup$ Don't worry, @Tristan, that will correct itself once you run out of engine. $\endgroup$
    – user
    May 5, 2017 at 16:54

The specific heat capacity of the fuel is higher than that of the liquid oxygen at the operating temperature/pressure, hence for the same rate of flow, using the fuel cools better ... and apart from any of the above-mentioned liquid/vapour/pressure issues.

The operating temperature range when using kerosene / RP-1 is also much greater than if using LOX.

As per my comment above, any "coolant" is not used because it's intrinsically cold, but because it can transport heat effectively in the required region at the required temperature range.

  • 1
    $\begingroup$ I wondered about that, but that's actually not true at least for RP-1/LOx, as far as I can tell - LOx's specific heat capacity at its boiling point is something like 20% higher than RP-1 at 273 K. Those were at 1 atm though, I know the pressure in the cooling channels is much higher - were you able to find data suggesting they switch around at ~10 MPa? $\endgroup$ May 4, 2017 at 17:55
  • 1
    $\begingroup$ @pericynthion: However, Cp for liquid O2 is the incorrect figure, since at the temperature encountered, the oxygen would be a supercritical fluid. $\endgroup$
    – MikeW
    May 10, 2017 at 9:56

Late entry but: turns out that whilst there are some arguments as to why you might not want to boil oxygen inside a metal chamber next to a multi-thousand-K flame, it turns out that not only is it entirely possible to use oxygen as coolant, but it has in fact been done on test firings already, with a view to doing so on a real flight in the future.

Some test firing was done a few decades ago: Cooling of rocket thrust chambers with liquid oxygen in 1990, where they took the step of deliberately cutting holes in the coolant channels to simulate cracking and leaking of LOX into the combustion chamber. "The tests were conducted safely without any damage to the facility or attached hardware" they say, possibly with some relief. Clearly it was possible and indeed practical back then, but presukably not worth the effort and risk compared to conventional designs and already available engines.

More recently though, a company called Launcherspace have built an tested a (partially) 3d-printed engine which uses both fuel and oxidiser for cooling purposes, and decided to call it Engine-2. Their naming skills aside, it seems to work.

Launcherspace Engine-2 test fire

(click image for a youtube video of the firing)

There has been some recent research on the matter too, including this one: Improving the performance of LOX/kerosene upper stage rocket engines, saying interesting things like

Analysis of existing oxygen/kerosene engines showed that replacing of kerosene regenerative cooling with oxygen allows a significant increase of achievable specific impulse, via optimization of mixture ratio.

There are other things that a search engine might throw up at you in response to a suitable query, but I couldn't separate the good stuff from the bad so I won't regurgitate any more of it here.

  • 1
    $\begingroup$ Also Reaction Engines/EADS-Astrium tested an oxidiser cooled combustion chamber in 2010, That being the planned cooling of the SABRE engine $\endgroup$
    – user20636
    Dec 12, 2019 at 22:36
  • $\begingroup$ Launcherspace really needs to hire me as a Chief Namingthings Officer. Other than that they’ve got a sweet little engine there. $\endgroup$ Dec 13, 2019 at 6:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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