From NPrize

This page is an introduction to rocket engines, explaining and comparing some of its subsystems. Our rocket will be based on regular rocket technology, as explained on this page. However some choices need to be made in order to gauge the feasibility of the project as a whole, in particular to have ideas of the possible dimensions of the rocket. These choices are presented on the page of the first approximations for the rocket. Other information and pages about the rocket and its flight can be found in the Rocket category.


Rocket Engine

The general principle may be simple, but there are numerous ways of achieving it. Different features and properties differ between existing rocket engines, and they all have consequences on complexity of manufacturing, complexity of operation, cost and weight for example.

We gather in this table the main properties of some of the existing rocket engines, mostly innovative designs.

Rocket engines features
Company Rocketdyne NPO Energomash XCOR XCOR Armadillo
Model SSME RD-107 series (Soyuz) XR-4A3 (EZ-rocket) XR-5K18 (Lynx) LOX/methane (no name)
Propellants LOX & LH2 LOX & Kerosene LOX & Alcohol LOX & Kerosene LOX & LCH4
Tank pressurization Yes, with O2 and H2 gases Yes, with Nitrogen (same pump than propellants) No No Yes, with Helium
Fuel pump Turbopump Turbopump driven by gaz generator using hydrogen peroxide decomposition (8300rpm) Piston pump Piston pump No
Cooling Regenerative w/ LH2 in three stages Regenerative w/ kerosene (5 mm deep channels milled in the inner wall) and film of kerosene Regenerative (w/ Alcohol?) Regenerative w/ Kerosene ?
Injector ? 337 swirling/mixing injectors, ring of kerosene only for film cooling - view cut ? ? ?
Chamber metal Copper or iron? 6 mm thick chromium bronze alloy inner wall, steel outer wall Copper Copper ?
Ignition system ? Pyrotechnic, soon hypergolic ? ? ?
Energy Hydraulic Electric
Provided by Engine's turbopumps ?
Actuator Six hydraulic servoactuators Static engine, control by vernier engines None None Servo-motor
Valves Hydraulically or pneumatically (helium) actuated ? ? ? ?

Pumps and tank pressurization

In order to get fuel from the tanks into the combustion chamber, the tanks must be either pressurized or the fuels pumped. In some cases, both techniques are used. The choice for this concern has a large impact on the design of the engine's hardware, and the complexity of manufacturing and operations.

Historically, only turbo pumps have been able to feed the engine at a large enough rate (high pressure chamber). Reciprocating pumps have been used in the past, but provided lower pressure and probably more weight. Innovative solutions appeared in research projects or private space projects, like the use of piston pumps for LOX or simple pressurization using liquid helium.

Several possibilities exist for tank pressurization:

  • vaporization of liquid propellants back into their own tanks
  • external vaporization of inert gas like Helium (can Nitrogen be used for that?)
  • smoke generator, that basically react fuel and oxidizer and use the resulting smoke for pressurization.

The tank design is by itself complicated and now has a specific page.


There are four known ways to cool a rocket engine:

  • Film cooling (aka the cooling curtain) takes place inside the chamber, generally using a ring fuel injector at the periphery of the injector plate, and acts both by cooling the chamber walls by contact and by isolating the walls from the combustion
  • Regenerative cooling is most widely used in rocket engines, since it is the most efficient way to have the chamber not being destroyed by heat. The general principle is to use the fuel, or sometimes the oxidizer, to cool the chamber walls before injecting those propellants into the chamber. The coolant flows into a series of pipes or milling into the external or intermediate walls of the engine, either around the nozzle, the chamber or both of them.
  • Ablative cooling is based on materials that provide cooling by being gently destroyed, like the heat-shield of spaceships, or the carbon fiber composite nozzle of SpaceX Merlin 1A engine.
  • Radiative cooling uses the natural capacity of materials to radiate (in infrared light for example) when they are hot. Doing this, they lose energy, and thus cool. This is efficient in the void of space, and is used as the nozzle cooling method for the SpaceX's Merlin Vacuum nozzle (with regenerative cooling for the chamber).

Cooling for a LOX/E85 engine

For our rocket engine, based on LOX and a cheap fuel like E85 or JP-A, we will consider the use of LOX as the coolant, instead of fuel, since cheap fuel polymerizes into cooling pipes, resulting in obstruction and engine failure. LOX as coolant already has been studied by NASA:

LOX cooling at chamber pressures to 1500 psia was demonstrated by in-house testing at the NASA Lewis Research Center in the late 1980s. Chambers were fired with cracks to demonstrate wall integrity at elevated LOX mixture ratios. See AIAA paper 89-2739 or NASA TM 10211 3.

and by Rotary Rocket and seems feasible as stated here by Doug Jones (Rotary Rocket):

"Jet A is a lousy coolant, we have 2.9x the mass of LOX as of fuel available for cooling, and (most important), the LOX has more pressure available for cooling. Bear in mind that flowing through the coolant passages requires a substantial pressure drop, and since the LOX is denser than the fuel, it reaches higher pressure in the centrifugal pumping of the wheel. Thus it is the logical choice for coolant- and it does not foul, no how no way."

Using LOX for film cooling has also been demonstrated, by Armadillo Aerospace.


Injector role is to mix propellants in the combustion chamber in a way that will produce the most efficient possible combustion. It faces several challenges, such as flow variations, pressure variations in the chamber leading to POGO, film cooling of the chamber walls. It determines the precise start sequence that will not explode the chamber, a process amusingly also called spontaneous disassembly. The temperature of combustion, the combustion ratio, and chamber pressure directly depend on the injector's design.

Injectors are most often composed, nowadays and in expensive engines, by hundreds of coaxial fuel/oxidizer injector elements. They assure a combustion efficiency over 99%, so many injector elements mixing very nicely the propellants together.

An alternative design comes from the research of TRW in the sixties, and is called the pintle injector design, or pintle engine. In this recent paper, TRW summarizes all achievements and the numerous benefits of such engines, which are very interesting for our goal here. Pintle engines only have one injector element, and are thus much less expensive to produce than traditional hundred-elements injectors. They however provide a perfectly stable combustion, with efficiency over 96%, for engines of any scale, with any propellants, and are able to deep throttle up to 1:35. The propellants enter in collision at the exit of the pintle, mixing them efficiently, but requiring more space than in traditional injector design. The Lunar Module Descent Engine is probably the most famous pintle engine, but SpaceX is using them too now.

It has never been seen (by TRW at least) that a pintle engine failed or had combustion instability. Bomb tests have always been successful, for any engine size. There may be only three drawbacks to these injector designs:

  • combustion efficiency is a bit lower than highly complicated injector designs but still good,
  • combustion chamber requires to be longer than in multiple-injector elements since the combustion is not made parallel to the injector's head but in a torus/cone a bit more distant,
  • film cooling may be more complicated to design, because there is no specific elements for this purpose. In the paper, it is said that the film is obtained by pintle tuning. I think it must be difficult to have both a good film and a good combustion efficiency. From SpaceX's experience, it seems that regenerative cooling is sufficient to sustain the combustion temperature hitting the walls.

Pneumatic and hydraulic pressure for actuators and valves

As we can see in the table at the top, different possibilities exist for actuating. The SSME uses hydraulic in nominal mode and pneumatics using He for backup. In satellites, lots of valves are pyrotechnically actuated.

Obtaining the pressurization in the system is not easy and is generally done by the fuel pump. SpaceX provided an elegant solution to hydraulic pressure by using the fuel (RP-1) as hydraulic fluid for the launcher, fuel pressurized by the main fuel turbopump.

Personal tools