Algodoo

you wouldn't be able to hear much from making that csv audio, it would be just some noise for a fraction of a second. Audible sound would be alternating currents of 10 Hz to around 18 KHz

i guess you could somehow make a better air that has a kind of pressure and tries to keep the (very very light) molecules separate, obviously not at 60 Hz but much more, and then try to make it vibrate in a sine fashion or something like that

The Butterfly effect would make your simulation produce different noises each time, a more accurate simulation may help, but I don't think this would be viable. It's pretty cool though!

Thanks!

mike