I work for an aerospace firm. While I mostly do computing systems, I've learned some by osmosis from working with engineers, mechanics, and pilots. Also, as I work with a new group, I study their field so I can talk their language.
The physics are well understood: Navier-Stokes equations. But they are so hard to solve that we simplify where we can: Assume away viscosity, or compressibility, or heat/chemical effects.
Even then real problems are too hard to solve symbolically. All real work is done with models. Hardware mockups and scale models in wind tunnels have a long history, and a long history of complexities to overcome (e.g., scaling effects). So much of the work has turned to computational simulations -- which themselves have problems.
Theory tells what the shape must be. Structural engineering is how we obtain and maintain that shape using materials and designs that are still light enough to fly. Systems engineering is how we tell the control surfaces to change, so we can maneuver.
Takeoff is optional; landing is mandatory. Structural engineering makes them possible. Systems engineering makes them predictable and safe.
Structural engineering is always a tradeoff of weight and strength. A key insight is to put strength only were you need it -- and that requires lab and simulation studies of stresses. Aircraft designers are always looking for better (lighter) materials. Alimunim was a big breakthrough, and composites are now taking the lead.
See Niu (niu2006) for traditional airframe structural design. He also has one on Composites.
System engineering for aircraft is largely the same as anywhere, just the problems are specific. We need
Humans can't fly by themselves, and they can't even float safely down to earth if something goes wrong (remember Icarus?). So you'd better do everything right, first time, every time. Even a perfectly designed and built aircraft can fall out of the sky if you do something wrong.
Further, you don't kill just yourself. An aircraft falling from the sky can kill innocent bystanders. Therefore the regulatory bodies have a need to ensure safe designs, and safe pilots. Pilots need highly repeatable habits for prefight, in-flight, landing, responses to emergencies, etc. All this calls for extensive training.
The regulations and the handbooks reflect a century of hard-earned experience on what can go wrong and how to prevent it or cope with it. See FAA handbooks for airplanes (faa_airplane2007), rotocraft (faa_rotor2007), and gliders (faa_glider2007)
See also Illman (illman2000) for more general/related information.
There are a lot of military aircraft, and some private/business craft, but the bulk of airtravel is on commercial airlines. It is a business, with a business's needs to plan and execute.
Creator: Harry George