The SR-71 was very complex. For example it had multiple heat exchangers to get rid of heat, to transfer it to the fuel before it was combusted. The titanium structure was complex to maintain. Welding needed to be done in "bubble" work stations under protective gas.
> The titanium structure was complex to maintain. Welding needed to be done in "bubble" work stations under protective gas
Operational complexity is orthogonal, and tends to be inversely proportional to simplicity in implementation...
For instance they needed to refuel the thing immediately upon reaching operating temperature in flight, since it leaked like a sieve on the runway until everything expanded. Rather than try fix that somehow with more engineering, they shifted the complexity into operations.
IIRC it couldn't even start its own engines cold, relying instead on hot-rod v8s setup on the runway to bootstrap the thing. More operational complexity in favor of leaving an entire subsystem out of the plane.
Don't get me wrong, I'm a huge fan of the SR-71, but so much of that appreciation stems from its ruthless pursuit of its narrow operational goals, for as much what it isn't as it is.
It's the polar opposite of stuff like the space shuttle or F-35 where the aggregate complexity is through the roof to accommodate a kitchen sink.
Automatic transmission in an automobile, operationally simpler than the manual constant-mesh trans it broadly replaced: Driver (AKA the Operator) selects drive or reverse when stationary, no need to participate in gear changing while in motion. Substantially more complex in implementation in exchange; go look at the hydraulic valve body of your average slushbox.
Manual trans: Simple implementation, single clutch assembly, sets of gears selected mechanically by axially sliding dogs on splined shafts. Operationally complex since the driver must continuously participate in the selecting of gears and clutch operation while in motion.
The arguably most complex technology of Space Shuttle was SSME. Also, arguably the most important part of LEO rocket technology is engine - as soon as engines matured enough, space era started. Yes, there are gyros and superlight tanks, but Lambda-4S still does illustrate the importance. And hypersonic winged flight problems were greatly decreased in the Space Shuttle case by a "brute force" approach with ceramic tiles.
SSME was more or less repeated in results by RD-0120, and rather soon. I'd argue RD-0410 was more complex. SR-71 engine works in two different modes, and even today is not really repeated elsewhere. I think that shows how awesome the SR-71 design was - especially, but not only, for its time.
I'd appreciate a good analysis of design problems and solutions of SR-71 to decide which project was technically more complex.
