I toured McDonald Douglas when I was in college. One interesting fact they mentioned is that auto makers will spend ~$10 to lose a pound, aircraft makers will spend around $1000 to lose a pound and aerospace would spend around $10,000 to lose a pound.
My company once had to lose 50 grams due to a mass budgeting error on a Japanese satellite. We were given $1000/gram to carve out some mass. We eventually did only to discover that the savings were wiped out by a 50 g error in the original estimate.
I suppose most are composite and held together with adhesives now. They had one clean room where they assembled some subs for the shuttle program. Very cool stuff.
Not all S/C are composites. If you watch the Canadarm II on the Space Station doing its thing, I worked on the Force and Moment Sensors at each end of the arm (at the wrist locations). It is made of Aluminum 7075-T6. Yet to be launched is the hand for the arm, most of that is titanium 6Al-4V.
With respect to design for vibration, I was mentioning that the ratio of stiffness/mass (density) is the driver for what frequencies a structure will vibrate at. Divide the stiffness modulus of steel, titanium and aluminum by their respective densities you get roughly the same answer for each. With magnesium you get a slightly higher number. If you made identical trusses from each material, each would quiver at nearly the same frequencies. However the absolute stiffnesses of the trusses differ, if aluminum has a stiffness of 1.0 the titanium would be 1.6 while the steel would be 2.9. Despite being dense, steel competes against the others quite while since it brings stiffness along with it mass. When you add in non-structural mass like people, interiors, etc. the answer is much muddier. This is why you can choose any of the three and make a car body perform almost identically. Then when you consider strength (particularly fatigue strength) steel still dominates. Titanium is nice but expensive, and aluminum has poor fatigue life.