This is way more fun than just "upside-down" — it's all different orientations!
More seriously, it looks like an excellent testing and optimization program that cannot be done as well in simulation (and in any case would be needed to validate simulations). There's always something that pops up in reality that had not (yet) been accounted for in the simulation, and when that much is on the line, this kind of testing is critical. Great to see a glimpse into how it's done!
> (and in any case would be needed to validate simulations)
No matter how high fidelity a simulation, at the end of the day it is still a simulation. Correlation issues plague almost all rapid computer design processes (looking at you F1)
You had to read pretty far to get to the reason for mounting upside-down:
"The way to solve that problem was totally placing the aircraft upside-down. With antennae on the belly of the plane and the belly facing the sky, we could rotate, tip and spin the plane any way we wanted and the pedestal would be safely below the aircraft and out of the way."
I see that this started back in the day before computers and simulations, but it all comes down to integration testing. I've seen plenty of instances where all the parts worked in testing but in the real world it failed miserably.
It is ingenious that they figured out to just mount the planes upside-down on a short test stand as a way to do the testing so there wouldn't be the interference of the pole in the way.
I am willing to bet they had a time where they thought about hanging the planes from a giant crane or maybe actually fly the plane over a huge track where the test equipment was on a train following the plane.
I'm surprised doing this more or less full scale was worth it compared to testing individual components. As a software engineer it's always eye opening that processes like integration testing are just as important in physical domains.
The number of flight hours goes way up once you have to factor in all the combinations of pylon-mounted ordnance on something like an A-10 or F-16. Your jamming system might work fine with a missile mounted, but not with a JDAM at the same location. And you want your pilots to be confident that the aircraft systems work every single time, no matter what.
I would have liked to be in the meeting where this was proposed to a General. "You want to WHAT??"
Individual component tests (which absolutely do happen, but can be done by the contractor themselves, prior to installation and as part of any normal engineering process) aren't enough. All EM systems influence one another in some way and only integration testing allows you to verify the correct operation of these systems together.
More seriously, it looks like an excellent testing and optimization program that cannot be done as well in simulation (and in any case would be needed to validate simulations). There's always something that pops up in reality that had not (yet) been accounted for in the simulation, and when that much is on the line, this kind of testing is critical. Great to see a glimpse into how it's done!