Theoretical question about frequency duration

Sorry for the abstract question, but for obscure reasons I've been thinking more and more about PianoTeq and modeling in general today. Could anyone with a few minutes of free time look over this?

On an acoustic instrument, what contributes to the duration of a single frequency? Just the following, not considering the materials from which the oscillator is made, the  proximity of the listener to the sound source, or reverberation (all of which are big factors)?

1. The force with which the oscillator is struck or plucked or dropped from the third floor. 
2. The frequency itself. (High frequencies expend the energy from the strike faster.)
3. The materials near it or striking it, which can absorb\diffuse, reflect it, or vibrate from it and do all three in varying ways and for varying lengths of time.
4. Other frequencies struck at the same time.  (The force with which they are struck? Their exact frequency? Their own duration?)

About #2: Is the ratio of frequency to duration linear--is there a strict ratio of frequency  to duration, so that a freq of 220 hz has a duration X times that of 440 hz, or does duration change at another predictable but nonlinear rate?   

About #4--I remember reading, but can't now find, a discussion of the way different frequencies affect each other, and how frequency ranges that are multiples or integers of each other do affect each other. Anyone know a good source for more information that doesn't put me chin-deep in math too fast?

Are there other factors that contribute to the duration of a single frequency?