Why does liquid space propellant make a good engine coolant? I was researching the regeneratively cooled nozzle and combustion chamber on the Merlin 1D when I found this out. I do know that the pipes are built around the combustion chamber and nozzle, but don’t know how it gets cooled by the propellant.

  • 3
    $\begingroup$ Is it relevant to observe that, unlike, say, a car engine, the coolant is open cycle. It is used once to cool the engine and then burnt and discarded. It is not a closed cooling loop, for which other fluids would surely be better. $\endgroup$ – Steve Linton Nov 26 '20 at 8:32
  • 3
    $\begingroup$ It has the double positive effect. 1/ The propelent used as coolant helps to cool the engines. 2/ The propelent is preheated by engines, what increases engine performance, as higher temperature and leaving speed of propelling gases are reached, compared to cold propelent. $\endgroup$ – Poutnik Nov 27 '20 at 10:38

Really, there is nothing particularly special about it.

Any liquid flowing through pipes in the engine wall will carry heat away from the engine as it heats up. Obviously, some liquids will be more effective than others. (Also, some propellants won't work because they'll either clog the pipes or explode, but RP-1 is especially formulated to avoid this).

With rockets, however, you can then burn the heated-up propellant in the engine, which saves energy since the heat lost to cooling isn't wasted. This also (more importantly) saves you from needing a separate supply of coolant and a huge radiator to get rid of the heat.

  • 16
    $\begingroup$ It's especially convenient that the rocket engine requires a large volumetric flow of propellant, which means a lot of heat can be transferred away from the combustion chamber walls. $\endgroup$ – Russell Borogove Nov 25 '20 at 21:33
  • $\begingroup$ You would essentially need to carry an amount of cooling that equals the amount of fuel, which is entirely prohibitive! With a better performing coolant you may optimize the amount to 50% or 20%. 20% would still be around 100t, much more than the payload. (This assumes that you cannot radiate away the heat fast enough, so cooling would still have to be open-ciruit.) $\endgroup$ – Peter - Reinstate Monica Nov 26 '20 at 13:53
  • 1
    $\begingroup$ @Peter just because you have that much cooling capacity available doesn't mean you need it. However it synergizes nicely with the need to vaporize the fuel to combust it. You have to inject enough heat into the fuel mass to overcome the enthalpy of vaporization. If you don't do that in the engine wall, you have to do it in the combustion chamber, where it'll leach energy from thrust. But then you'd still have to cool the jacket... $\endgroup$ – Harper - Reinstate Monica Nov 27 '20 at 20:35

The volumetric heat capacity of liquids is much higher than that of gases. You need much more energy to heat a certain volume of water than the same volume of air. The density of liquids is much bigger than that of gases, that is why they can transport much more heat energy.

Therefore a liquid flowing through a pipe takes away much more heat than a gas flowing through the same pipe at the same speed.

To cool the combustion chamber, the liquid fuel flowing through the pipes takes away the heat and gets warmer. It should get not too warm so it remains a liquid without gas bubbles.


Because it's 'free' energy for the taking!

In a rocket, you have liquid fuels that need to be vaporized to burn correctly. To vaporize a liquid, you need to add heat energy to it: specifically its latent heat/enthalpy of vaporization, which is quite considerable. The enthalpy causes zero practical temperature increase: the substance is still the same temperature; it's just now vapor instead of liquid.

If you do nothing else (say you have external cooling of the nozzle), this enthalpy must come from the fire in the combustion chamber. That is, the enthalpy of vaporization is "stealing" energy that would otherwise be used for thrust. Stolen energy is a big deal; the Rocket Equation is harsh.

Think of the heat in an automobile. Sure, they could have the engine spin a generator and make resistive electric heat, but that would burden the engine. The engine's radiator is already throwing away waste energy from engine cooling... that's free. No impact on fuel economy or performance.

Therefore, even if the nozzle was made of indestructium... the opportunity to get "free energy" by cooling the engine nozzle is a win for efficiency. By adding heat to the fuel or oxidizer there, it will require less enthalpy of vaporization when it gets to the combustion chamber, and that means more thrust. For free.

How to capture that energy, without boiling the "coolant"

Mind you, having it turn to vapor in the nozzle cooling jacket would be a huge problem.

But we have something working in our favor. There are high-pressure pumps which raise the fuel and oxidizer from the native pressure in the rocket's tanks to a much higher pressure for injection into the combustion chamber. Raising a liquid's pressure also raises its boiling point. So this pressurization lets you add energy to the fuel/coolant without it boiling in the pipes.

So the pressurization helps you "get ahead" of enthalpy, by letting you preload some of the enthalpy into the pressurized fuel.

  • $\begingroup$ But the fuel used for engine cooling should not vaporize within the cooling pipes or channels. Gas bubbles within the liquid would reduce the heat transport. So the temperature should be lower than the boiling point. Another problem would be the extreme increase of volume and pressure if a substantial part of the liquid is vaporized. $\endgroup$ – Uwe Nov 27 '20 at 22:10
  • $\begingroup$ @Uwe Good point, I'll address that with another paragraph. $\endgroup$ – Harper - Reinstate Monica Nov 27 '20 at 22:31
  • $\begingroup$ The heat removed from the chamber and the nozzle by liquid cooling is reused in the chamber, so that was free. But the energy needed to pump the fuel through the cooling channels is lost, at least the energy lost in the turbine to drive the pump. $\endgroup$ – Uwe Nov 28 '20 at 10:59
  • $\begingroup$ @Uwe Yes, but if you had an alternative way to cool the nozzle, wouldn't you still need those high-pressure fuel pumps? Would natural tank pressure suffice for feeding into the rocket's combustion chamber? $\endgroup$ – Harper - Reinstate Monica Nov 28 '20 at 17:20
  • $\begingroup$ You need those high-pressure fuel pumps to maintain a high combustion chamber pressure. If natural tank pressure is sufficient for feeding into the rocket's combustion chamber, chamber pressure would be too low to get a fast exhaust speed. To increase natural tank pressure you need much heavier tanks to resist the pressure. Of course you need more pressurizing gas which increases dead weight. $\endgroup$ – Uwe Nov 28 '20 at 18:53

Just two short comments on the other excellent answers:

Wall cooling is not free: What happens is that you (can) have a significant reduction in chamber pressure (depending on chamber size), reducing your thrust and efficiency. There may also be some second order effects on combustion efficiency when heat is removed from the process. So if we could run an engine with adiabatic walls, it could absolutely be advantageous.

Supercritical fluids don't boil: Furthermore, all current rocket engines operate at chamber pressures exceeding their propellant critical pressures. As the pressurization occurs before entering the cooling channels, strictly 'vaporization' cannot occur at supercritical pressures, as there is no boiling point anymore. However, at near-critical pressures, something similar may occur, 'pseudo boiling', with similar outcome.

[edit: address some comments from organic marble]

  • $\begingroup$ Can you back up this statement with some references or calculations: "What happens is that you have a significant reduction in chamber pressure, reducing your thrust and efficiency. " $\endgroup$ – Organic Marble Dec 1 '20 at 18:16
  • 1
    $\begingroup$ I frankly don't have any publications off the top of my head, so this is more based on my personal experience in running these simulations. If you look at model combustion chambers, such as Mascotte, DLR Lampoldshausen, PennState burner, you would see a ~10% reduction in chamber pressure in simulations if you accounted for heat transfer compared to adiabatic boundary conditions. If you consider the chamber gas and you reduce its temperature at the wall, you decrease its volume significantly, but its more complicated that that with combustion and nozzle flow. $\endgroup$ – bernulli Dec 1 '20 at 18:51
  • $\begingroup$ That's great info. I take it these are pretty small chambers, though? My experience with larger engines doesn't jive with a large reduction. $\endgroup$ – Organic Marble Dec 1 '20 at 18:52
  • 1
    $\begingroup$ Could be, good point! The surface/volume ratio is certainly larger in small chambers, so you'll extract a lot more energy relative to what you have in the chamber volume compared to full scale engines. $\endgroup$ – bernulli Dec 1 '20 at 18:55

The speed at which heat is transferred from a surface depends on the transfer capacity of whatever is touching its surface and the speed at which it is taking that heat away from the surface thus making a heat gradient that pulls more heat out of the surface faster.(that's as simple as I could put it without fancy technical words)

Rocket fuels that are commonly used have good heat transfer properties and they are liquid so they can be pumped very quickly off of a hot surface thus taking the heat away with them, this serves to pre-heat the fuel fur combustion lowering the fuel wastage as well.


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.