Categories
BW Member Blog

The Rutherford Engine


Ernest Rutherford, considered the father of nuclear physics, was born in Brightwater, New Zealand. It is fitting then that the radically new rocket engine from the small aerospace company founded by New Zealander Peter Beck, which powers the Electron rocket that launches from the Mahia Peninsula in New Zealand, bears his name.


Peter Beck, founder of Rocket Lab, and his engine

Two aspects of this engine set it apart from all others currently in existence. First, except for some piping and pumps, it is 100% 3D printed, formed of powder melted by a high-power electron beam gun in a high-vacuum chamber. If that weren’t impressive enough, it is also powered by high-pressure pumps driven by brushless DC motors.

To gain an appreciation of the simplicity of the Rutherford engine, a quick dive into other types of rocket engines is in order.  A rocket works by throwing mass away from itself at high velocity. In a Chinese bottle rocket, a powder mixture of something that burns (fuel) and something else that creates oxygen that allows it to burn (oxidizer)  is squeezed into a tube and lit off. The heat of the reaction expands the gases that result and the increased pressure shoves them out the end of the tube. The solid rocket boosters that were used on the Space Shuttle and are seen strapped to the sides of cylindrical launchers like the US Titan IV, or most Chinese rockets, are the same technology.

When a solid rocket is ignited, it doesn’t shut off until it’s out of fuel. A more controllable technology is the liquid fueled rocket engine, which mixes a propellant (such as kerosene, methane, or propane) together with an oxidizer (almost always liquid oxygen) and lights it, potentially shutting down and relighting it several times during a flight. Controllable, yes. Efficient, yes. Simple and cheap, no.


SpaceX Raptor, a full-flow staged combustion cycle engine

At its least complex, a bi-propellant liquid-fueled rocket engine is just a combustion chamber with a nozzle sticking out of it, into which fuel and oxidizer are pumped at high pressure and ignited. One can be built in your garage out of some plumbing parts fed with pressurized gases, such as propane or acetylene and oxygen. However, to develop a useful amount of thrust (at least enough to lift the engine’s own weight)  a rather huge amount of propellant mass has to be expelled through the nozzle, and practically speaking to accomplish this the propellants have to be in a liquid rather than a gaseous state. Thus the need for high-speed, high-pressure turbopumps.

This isn’t the place to get into all the details of how this has been accomplished over the years, what with gas generator (open) cycles, staged combustion (closed) cycles, full-flow staged combustion, and so forth, but suffice to say the result is a nightmare of plumbing. Essentially, these are all different ways to use a second smaller rocket engine (called a preburner) that makes hot exhaust gases that drive a turbopump – in exactly the same way a turbocharger on an automotive engine uses hot exhaust gases to spin a turbine that pumps air into the intake manifold at higher than atmospheric pressure.

Like automotive turbochargers, rocket turbopumps operate hot, but at very much more extremely hot temperatures than the one on your neighbor’s diesel-powered Dodge pickup truck. To make their lives harder, they are also physically near to the extremely cold  propellants they are pumping. The temperature difference over as little as a few inches can be several thousands of degrees Fahrenheit. This puts difficult demands on the metallurgy and mechanical design and may limit the time the engine can fire.


The Rutherford 3D-printed, electrically-pumped rocket engine

R utherford disposes of all this complexity by tossing out the preburner altogether and driving the turbopumps with electric motors. Compare the appearance of the Raptor engine, above left, with the Rutherford engine to the right (the blue component is a gimbal actuator, and the red one is one of the electric motors; the other actuator and pump are on the other side). Fewer parts, higher reliability. Which would you rather depend on?

The numbers are simply amazing. Each engine has two motor-turbopump assemblies which together use 50 horsepower to drive impeller shafts to 40,000 RPM, pushing kerosene and liquid oxygen into the combustion chamber through the 3D printed injector plate. According to the company, the pumps achieve a 95% efficiency, far surpassing the typical 60% of rocket-turbine-driven pumps. Given that there are nine engines in the first stage of the Electron rocket, we can calculate that the pumps use a minimum of 745 watts/hp x 50 hp x 9 ea / 0.95 eff  = 352,894 watts of power – enough to power an entire community of houses running their air conditioners full blast.

[Note: at this writing I have been unable to determine if these motors are 50hp each, or together. References are ambiguous. If each one is 50hp, then double these numbers.]

The electrical juice comes from a lithium polymer battery pack the size of a large shoebox that is nonetheless capable of providing a full megawatt of peak power.

“It’s really only the advancement in battery technology that has enabled us to go to electric turbopumps,” [Rocket Lab CEO] Beck says. “Even three or four years ago, the technology wouldn’t have been sufficient. But there have been enormous advances in a short time period, and now the electric motor is about 95 percent efficient, versus the 60 percent efficiency of the gas motor.”

T he other major advance that is embodied in the Rutherford engine design is that it is largely comprised of 3D printed structure. It is doubtful that the company could have achieved its goal of rapid manufacturing using ordinary machining techniques, and the cost would certainly have been significantly higher. From a Rocket Lab press release:

“The Rutherford engine was designed from the beginning to be both high performing and fast to manufacture on a mass scale,” said Lachlan Matchett, Vice President of Propulsion. “By enabling faster, scalable engine production we speed up production of the whole vehicle. We can print an entire engine in as little as 24 hours. This allows us to build and launch at unprecedented frequencies to democratize access to space, enabling the creation of crucial orbital infrastructure.”


ignition!

9 tiny nozzles

Rocket Lab has produced a total of 40 flight-ready engines to date, and aims to produce another 100 engines by the end of this year. The Rutherford engine’s production scalability is facilitated by additive manufactured (3D printed) primary components. With a 3D printed combustion chamber, injectors, pumps, and main propellant valves, Rutherford has the most 3D printed components of any rocket engine in the world.

The company has succeeded in coaxing a significant amount of thrust out of the tiny, 70-pound engines (nine of them fit into the four-foot diameter of the Electron rocket). From another news release:

Rutherford is used as both a first stage and as a second stage engine, with sea level versions on Electron’s first stage producing 24 kN (5,500 lbf) of thrust and has a specific impulse of 311 s (3.05 km/s), while the vacuum optimized version operating on Electron’s second stage produces a max thrust of 24 kN (5,500 lbf) of thrust and has a specific impulse of 343 s (3.36 km/s).

In January of 2018, Rocket Lab’s Electron launch vehicle succeeded in putting three Cubesats in orbit. The payload rode to space first on nine of these little engines in the first stage, and then on the single one that drove the second stage to complete the journey, ending in an orbit 300 miles above the earth. By anyone’s reckoning, this is a significant achievement for a company started largely by one man in 2007. A man who started out as an apprentice tool and diemaker at Fisher & Paykel and had dreams of touching space.


For those who like to watch videos of things that make loud noises, here is a short program on qualifying the Rutherford engine for flight, and another of the successful orbital launch:


One last thing: despite being started by a Kiwi, and having a launch facility in New Zealand, Rocket Labs is incorporated as an American company and has American headquarters. Because it’s easier to do business in America.


8 replies on “The Rutherford Engine”

You’re welcome!

I didn’t get into the issue of the need to carry the dead weight of the battery (dead in the sense that it isn’t burnt as propellant), which impacts the overall gains of using electric motors. If you watched the launch, you noticed that they dropped the battery somewhat into the 2nd stage burn, where its mass makes it a significant part of the payload. That was the “hot swap” comment on the radio — they swap to a smaller battery for the rest of the burn.

There’s a lot of math involved in the best time to do this, so I thought it was better just to stick to the overview and not get into the nitty gritty engineering design tradeoffs.

But in my book, simpler is always better in the long run.

I probably shouldn’t be (I spend too much of my time in software and ought to keep up with hardware better), but I was surprised that the energy density of li-polymer batteries enables a “shoebox”-sized pack to suffice for running the pumps the whole way! Makes sense that at some point they ought to jettison, though I guess you have to weigh the cost of replacement with the cost of additional fuel (if there is capacity) to keep the battery on board. (Then again, I’ve grown so used to thinking about self-landing boosters thanks to SpaceX. If they aren’t recovering any of the booster intact in this case, that calculation may not apply.)

I actually haven’t been able to find any references to the actual dimensions of the primary battery. The “shoebox” was from somebody or another’s article, without links.

As a bit of a systems engineer, I look at it the extra weight argument from an overall cost standpoint, not from the (typically NASA) point of view that they have to get the last ounce of payload to orbit “X”. We are no longer struggling just to get a monkey into LEO, and we can afford a little more fuel or structure. I mean, hell, if a rocket this size built by a company this young can put 330 pounds 300 miles up, we can afford to make things simpler.

330 pounds is sufficient for a small nuke, just saying. It no longer takes a government to make an ICBM.

What a beautiful blog, both visual and literary. I have come to eagerly anticipate a Steve Walton blog as much as a new Right Angle or Firewall.

Thanks for all you do, Steve.

Thank YOU, Cheryl, for reading. I have to admit, it takes awhile to put together one of the long ones (4-8 hours), so they won’t happen every day. I suppose I might get better at it. I have to go beyond the editor that we have available here to do it, so it takes some fiddling….

Leave a Reply