I have a blurry memory of Tom Danley chiming in on a thread on the old LAB talking about his work with NASA back in the day. The thread was originally about what was the most powerful amplifier ever. There was a lot of talk about Crest 10001s and the like and then Tom talked about the amp they built back in the early days of the US space programme.
There had been a lot of launch failures in the early days, so a team was put together to reproduce the SPL at launch and see if that would shed any light on the problem.
The team did manage to build a transducer which achieved the target and they discovered why the launches were failing.
IIRC the spectrum had, as well as a heap of low end, a spiky HF element at 8k(?). When these higher frequencies were introduced to the test signal at the appropriate SPLs, a bunch of machine screws unscrewed themselves. Problem identified.
It was a very interesting read. I've tried a search of the Archives but couldn't find the original thread.
Resonant frequencies can be quite interesting.
It is not just low freq that cause issues, sometimes it is much higher.
And you have to be "dead monkey nuts on" in order for it to take effect.
Just like a buzz in a speaker cabinet. Some freq don't do anything-but when you sweep it and stay on the problem, it will "take off" and be really bad,
But a couple of Hz higher or lower and there is not an issue.