This shows the “golden age” during and just after the war. Even large things with a lot of design input could not be designed to be close to optimal right from the start. The B-36 experienced failures in structural testing and required modifications. Though it was outside known things because of its size. It just wasn’t a thing that popped out of the master designer’s imagination and flew perfectly when metal was bent according to the napkin conception.
I love how these old films assume the person who is watching them to understand something. There really needs to be some tv channel for this stuff (and it’s not Discovery!).
Of course, nowadays computers can be used for a lot in design and optimization (anchored into reality, so they can interpolate and extrapolate), but I’m interested that structural testing isn’t done more for example in ultralight aircraft design, building and modifications. You wouldn’t need to test the whole thing, just for example a wing spar, rear fuselage cone or engine mount.
The issue is that the plane might be overbuilt and have general margin in most of the places, but then fails because of some unexpected part being too thin, unsupported etc.. Like in the B-36, just adding some little support there adds less than a percent more weight but makes the whole structure 30% stronger. With structural testing you could even get a stronger AND lighter design. Also ease of manufacture, which in turn should increase quality, could be optimized, for example by reducing parts count or the number of joints.
In home building, a lot of the time it’s not about what is the theoretical strength of some material, but what kind of quality from manufacture can be expected. For example glue starvation in sandwich structures, resin voids in composites, material quality in wood, welds in steel tubes or aluminium. They are relatively simple to test in bench before even building anything bigger, to anchor the design methods and approximations to some semblance of reality. I loved Ian Lea’s tests of homebuilt wing trailing edge rib construction methods. You see that kind of stuff very rarely, instead people just copy other designs and use old rules of thumbs from some books, coming from through who knows how many broken telephones.
Whole size testing of large things is very expensive. They had to build whole new facilities for the B-36, yet they chose to do it, and it provided very valuable information. Rather a breakdown in controlled testing than in flight. You might not even know the cause of inflight testing failures. Nowadays the design and manufacturing space is more explored and new designs of most things should have much higher chances of working. Yet there are unknown territories. Look at large composite structures. The 787 wingtips move quite a lot here.