I've never seen a rocket engine and thought, "Well it's dull," but this design makes everyone else look positively pedestrian. This is because this was designed in 2 weeks, rather than the traditional process of countless engineers spending months or manually designing models in programs like CAD.Mostly thanks to AI.

Dubai-based AI engineering company Leap71 says it created the design using its large computational engineering model, Noyron. The company is said to be autonomously designed "without human intervention" and post-processed at the University of Sheffield, before being 3D printed in copper by the German metal 3D printing company AMCM, and the new engine iterations generated by the Tess

AI model will take several minutes, compared to months of work put into traditional rocket engine designs . The engine uses cryogenic liquid oxygen (LOX) and kerosene as propellants, and the injector head features a "state-of-the-art coaxial swarler" for mixing them.

Yeah, it's kind of Wallace's description of Wallace and Gromit's fame and used to explain his latest rocket design. "Coaxial volute". What time is it alive? The engine was designed to generate 5kN of thrust (equivalent to a lift mass of 500kg/1120lbs or 20,000 hp), making the Leap71 suitable for the final "kick-stage" of the orbital rocket.

The test launch took place at Airborne Engineering's facility in Wescott, UK.If you're someone who likes to watch rockets make you laugh, do I have a video for you:

The air-punching one, it initially looks like a strange choice for a rocket engine given its low melting point, but when actively cooled it clearly allows for a "compact, high-performance engine." For reasons that remain unclear, to my admittedly dull mind. Still, the more you know.

That said, there is a severe warning here. Leap71 states that if cooling fails, it will melt quickly. Then a bit like my AMD Ryzen7 7700X. I am a child, I am a child.

Its oddly patterned design facilitates, in rocket science terminology, a thin cooling channel that wraps around a chamber jacket where more kerosene is pumped to cool things down, or relatively cool them down. Not only that, there are things that can be immediately recognized as AI-generated in swirling fractal-like patterns.

As you can see, the test fire was successful and the engine ran a full 12-second long burn. This means that the Leap71 is enough to confidently declare that steady state is possible, and it can essentially operate as much as it needs with its ability as a powerful booster.

But that's not to say it's perfect. Some analysis done on the engine after the test determined that the resistance of the cooling channel was higher than expected, and since this is a phenomenon due to the surface roughness of the 3D printing, it seems that there is still some work to optimize the design before it potentially looks at the boundaries of our planet.

And that's the overall goal here. Josefine Lissner, Managing Director of Leap71, said: "We can now automatically create functional rocket thrusters and move directly to practical verification."

"Space propulsion innovation is difficult and costly. With our approach, we want to make the space more accessible for everyone.

So we go. Aesthetically interesting and functional, and much faster to implement than traditional designs. Perhaps the AI generation may be the key to future rocket engine development. 

Now I fancy a spot in Wencelidale. More cheese, gromit?