Tom Stanton wanted to push a projectile past the speed of sound using a machine powered only by a falling counterweight. In his latest video, his final measurement comes out at 346.4 m/s, or 776 mph.
The number is entertaining. The path to it is better.
Stanton says he started calculating this project almost six years ago. His Patreon archive also shows years of related trebuchet builds, launch mechanisms and experiments before the final video. This was not a weekend print built around a single clever idea.
Use cheap material to find expensive mistakes
To chase that speed, Stanton moves away from a conventional trebuchet layout. Gravity still provides the energy, but a transmission spins up a relatively short arm before the sling does the final acceleration.
The useful part is not the construction recipe. It is how he decides what deserves to be manufactured.
Before cutting the final arm from carbon fiber, he runs stress simulations and prints a plastic test version. That cheap prototype immediately exposes a buckling problem he does not want to gamble on at full speed. He changes the geometry, tests again, and only then commits to the expensive material.
The plastic prototype is not trying to imitate the finished machine. Its job is to make a mistake visible while the mistake is still cheap.
Measure the miss, then change the problem
Early high-speed launches expose another issue: energy is disappearing before it reaches the projectile. Stanton compares measurements, reduces aerodynamic losses, removes mass where it hurts rotational speed most and uses slow-motion footage to inspect the release timing.
A test with a 40 kg counterweight reaches 716 mph by his measurement, still short of the target. He then pushes the machine beyond the load it was designed around and the sling fails.
Instead of simply adding more input energy, the next iteration reduces projectile mass and changes the release mechanism. The engineering question shifts from “how do I add force?” to “how do I reach the same target while asking less of the machine?”
On the final run, Stanton reports an arm speed of 2,342 RPM. He measures the projectile travelling 1.94 metres in 5.6 milliseconds and calculates 346.4 m/s.
That measurement has not been independently replicated. Stanton also describes the machine as the first gravity-powered supersonic trebuchet, a priority claim IRZ has not independently established. What matters here is the documented experiment, not a world-record label.
Desktop fabrication changes which questions are practical
The last minute of the video is almost more revealing than the sonic crack. Stanton credits CAD, a homemade CNC mill and 3D printing with making the project possible in his own workshop.
Printing makes temporary geometry cheap. Simulation removes some bad paths before machining. CNC gives a small workshop access to precise parts that would once have required more specialised equipment. High-speed video becomes another measurement tool.
None of those technologies solves the project on its own. Together they make each experimental loop cheaper.
That is what makes the build feel modern despite the medieval mechanism. The remarkable part is not merely that a trebuchet can go very fast. It is that one maker can now move repeatedly from calculation to prototype to measurement to failure to a new physical part, in a home workshop, until a six-year mechanical question finally produces an answer.