Just to explain how things really work in an OEM: I had no text books at my desk, and didn't own a calculator. I just used the one built into Windows the few times it was needed. Most engineering work is in fact done out in the field, or on a test bench, or in a plant. The guy who signed off on the handling characteristics of a truck, a requirement for GVW, did not calculate it. He bolted a set of outriggers on, went out to the track , and tried to flip it. When the Ford managers were whining about how the F-150 doesn't stop as fast as the Tundra, the brake engineer took the tires off the Tundra, put them on the F-150, and tested it, then said "There, see you idiots, it's the tires, not the brakes!"
This is very much practical stuff. I determined the pressure capability of an oil system not with a finite element analysis. I take a bunch of prototype parts, assemble them, use a pressure gauge, and crank up the pressure until they blow up.
Now, there *are* guys who do a lot of advanced computer stuff, but that is usually done way before the grunts like me got to it. In your axle example, the Design and Release engineer (that's me), would take the axle he was told to use, and drive it around Bovingdon until it breaks, then make sure that met the requirements.
Now, when mechanics get mad at engineers for making things unservicable... that's not our fault. See... nobody cares about service. The service engineer is one voice, in the back of the room filled with guys trying to get things out the door as fast and cheaply as possible, and nobody listens to him. Not my rules, that's just the way it works.
