Re: Just a test
"Hmm, seems a bit silly for just a test. If they want to see how a bomb works they could do that on the ground."
It's hard to simulate the entire combination of aerodynamic and engine variables on the ground. There aren't a lot of wind tunnels able to accommodate 3 hot firing large rocket motors and a shaking table.
It's worth reviewing the Apollo 6 mission to see its failures. Apollo 6 experienced severe pogo oscillations that engineers had thought eliminated - which says something about "engineered envelopes." Then the upper stage spacecraft adapter suffered structural damage as air and water in it expanded as the rocket entered vacuum, something not spotted in ground tests. Ground conditions masked another design flaw, this one in a hydrogen line to the second stage J-2 motors: liquid air built up around the hydrogen line on the ground and damped vibration, while in flight vibrations tore the line free, leading to 2 engine failures.
And all these problems happened after the successful Apollo 4 test flight.
"All their components will have an engineered envelope in which to work,"
Oh, yes, there are envelopes for aerospace components. Components get initially spec'd out based on guesses that include healthy safety margins; refined in simulations (which are built on estimates and approximations); and then chiseled away to get to flight weight targets. The resulting component is thus enveloped in *some* sort of engineering numbers.
Whether or not those numbers are representative of the component's real world performance is another matter. I've been through the process of spec'ing out a solid rocket motor. It wasn't a novel design. It was built by a veteran solid motor subcontractor; used a proven alloy for the casing; was welded in a standard fashion; and there were generous margins and knockdowns in all design elements.
It still split its seams in hot fire testing because the "engineered envelope" didn't accommodate all real world problems. Problem 1: incomplete fusion of the welded joint, an expected issue, hence the post-weld inspections. Problem 2: x-ray and fluorescent dye inspections didn't spot the bad welds, an unexpected issue.
You can get past all the lab tests, ground tests, and even successfully launch related hardware and still find new and surprising problems that no "engineered envelope" will encompass.
Hence: test flights.