Originally Posted by
tdracer
To be fair, while CFD has much improved over the years, it's still far short of computational structural analysis - which is very highly evolved.
Sure the timing is pretty lousy, but structural testing to destruction of a completed structure is a bit archaic in this day and age. 300 meter tall skyscrapers are becoming the norm, no testing to destruction of completed structures needed. If one of those ever structurally fails, the death toll would be massive.
The materials that they are composed of are indeed tested to destruction, and the designs are (generally) tested the same way. Whenever there is a hurricane or earthquake, engineers look at the buildings and make recommendations for any needed changes to building codes. If I were to propose some super-duper polymer glue that replaces foundation bolts , you better believe there would be some pretty significant testing to destruction before that technique was accepted by national and local building code societies, and it probably would not get widely accepted until an actual earthquake happened and it was proven.
Breaking stuff is the fun part! The process that engineers use to validate building materials is really worth a TV show, especially for earthquakes and hurricanes. I would have loved to have been part of a wing breaking team. Seriously, though, that is one of our few defenses against Murphy; when something fails in a different way than you expected it can lead to a whole bunch of the original assumptions being challenged, which ultimately improves the final design. By definition, computer modeling can't really challenge assumptions since what is in the computer are the assumptions.
Not to say that building science is anywhere near perfect, the Grenfell Tower fire is a pretty good example of what happens when you
don't test the materials that you apply on a large scale.
Edit: I think that is where the engineers and the finance guys differ. Engineers love surprises (during the design phase), the stock market hates them. If you test the plane for real, you might find something wrong and that would impact the schedule, which could really move the stock price. If you test the plane using the same program that you used to design the plane, nothing unpleasant will have to be reported to the shareholders.