It sounded incomprehensible to me too but as I did some more learning about the process of building rocket engines I learned some interesting details. First, remember that people have been making high strength metals through careful processing for thousands of years. Second, the parts of engines are not made as part of large-scale industrialized manufacturing. Almost all the parts are made as few-offs, with far more energy, time, and effort put into making sure that a single instance of something is extremely reliable. Third, we got damn good at materials science in the past 100 year, and metals can be absurdly resistant to deformation under heat.
The alloy on the oxygen side pump in the Raptor engine is a work of magic and or art. An oxygen rich turbopump runs so hot it was thought impossible to create because the turbine would fail before the burn was over. SpaceX had to invent the alloy before the first Raptor full duration fire. That's a very low level, never done before, breakthrough required just to get to the real hard stuff.
edit: if you watch gas generator tests on youtube the gas coming out is dark because it's very fuel rich which keeps it cool (in a relative sense)
> SpaceX had to invent the alloy before the first Raptor full duration fire. That's a very low level, never done before, breakthrough required just to get to the real hard stuff.
Russians had oxydizer-rich gas generators in large engines since e.g. early 1960-s (see Proton 1st stage engines). Oxygen-rich gas generators are at least since mid-1980-s (RD-170). So not exactly never done before.
Russians sold RD-180 with technology to reproduce them in USA. It was just too expensive to make them in America, so eventually Energomash got all orders to make them for Atlases. And now Falcons are pretty good too, so less need in RD-180 - Orbital still buys RD-181, single chamber engine though. And Raptors could be in use soon. And BE-4 from Blue Origin hopefully will power new ULA Vulcan rockets soon too... and that oxygen-resisting metallurgy is likely used at least in some of them (RD-181 for sure).
A jet engine can't have such power density because it uses gaseous air which is about 1000 times less dense than liquid oxygen and only contains 20% oxygen.
Everything follows from that.
The pumps are not challenging temperature wise since they pump cold liquids.
The turbine is challenging, but the temperature can be limited by varying the ratio of propellants in the preburner (very lean or very rich means lower temperature). If you use lean, then it's a very oxidizing environment. If you go very rich, there's soot (if you use fuels with carbon).
And the chamber is not so challenging because there is so much cool liquid available for cooling.
You can boil water with a candle and a paper cup.
A high performance jet engine is a harder problem than a medium performance rocket engine.
Designing for thermal cycles and serviceability[0] is at least as difficult a problem as running a hypothetical rocket engine an equal amount of time in one longer, hotter burn.
(Such a design isn't needed and wouldn't be practical, but then again multiple aspects of SSMEs being reusable turned out not very practical either, depending on what version of design criterea you evaluate and how the expected vs actual usage changed over the lifetime of the program.)
[0]In both the engineering sense, as durability of the various loading cycles (ie lifetime turbine rotations or number of thermal cycles before eol or failure), and as being constructed as able to undergo maintenance and refurbishment between launches.
Not as much as RL-10s. SSME you can disassemble - because you should do that, as thermal stresses on turbine blades are too dangerous, so you have to periodically replace the parts which are nearing the fault.
Engines in dragsters are also know to have very high power comparing to engine size, but they also make less than 10k rotations at full power before they fail. That is enough to last one drag race which is several seconds.
