SpaceX's Raptor engine has undergone a striking visual transformation from the Raptor 1 to Raptor 3, evolving from a messy tangle of pipes and wires into a clean, streamlined design. The Raptor 3 first flew in May 2024 and delivers about 35% more thrust than the original version. The transformation was so dramatic that Tory Bruno, then-CEO of United Launch Alliance, initially dismissed it as exaggeration until SpaceX's Gwynne Shotwell posted video proof of the engine firing successfully. SpaceX keeps its engine designs classified and hasn't published official schematics or allowed teardowns, so the specific engineering changes remain partially obscured—though Elon Musk's occasional comments and speculation from SpaceX enthusiasts provide some clues about what actually changed. To understand the Raptor's design, you need to know how it works: it's a full-flow staged combustion (FFSC) engine, which is far more complex than simpler alternatives. Rocket engines produce thrust by expelling propellant at high speed. The simplest approach—cold gas thrusters—just vents pressurized gas, but that's weak. Better engines burn fuel and oxidizer together, but then you hit a problem: the combustion chamber pressure must stay lower than the tank pressure, or flames will backflow and choke the engine. The solution is turbopumps that pressurize the propellants before they enter the chamber. These pumps demand enormous power—the Saturn V's F-1 engine turbopump required 41 megawatts, roughly equivalent to what an Ohio-class submarine's nuclear reactor delivers. Most rockets solve this by burning some propellant to drive a turbine that powers the pump itself. The Raptor uses full-flow staged combustion, meaning it runs all its propellant through two separate preburners instead of just part of it. One preburner mixes a small amount of oxidizer with excess fuel; the other mixes a small amount of fuel with excess oxidizer. Both exhaust streams then enter the main combustion chamber together. This is brutally complex—before the Raptor, only two FFSC engines had ever been built, and neither successfully flew. The Russians developed the RD-270 in the 1960s without flying it, and the US built a partial demonstrator in the 1990s-2000s. SpaceX chose this approach because the high mass flow through the turbines theoretically lets them run cooler and at lower pressure, improving reliability—critical for reusable rockets that need to fly repeatedly.