Make a walkie-talkie with two empty soup cans and some string. Hammer a hole
into the closed ends of the cans. Pull a string or wire through the opening and tie
washers to the ends so they cannot come out through the holes. Place a tape along the
edges of the open end of the can for safety. Talk into one can while your friend listens
in the other. Pull the string or wire taut. Why can he or she hear you? Place one can
under water and listen to the other while your friend splashes. Do you hear anything?
Why?
Answers & Comments
Answer:
I think that they could go quite far through a very taut, smooth, uniform steel wire due to the 4:1 ratio of the speed of sound in steel to that in water. Nylon should work though not as well. I can find no data on the speed of longitudial waves in string (hardly surprising) but I would expect it to do poorly because of its low density, nonuniformity, and fuzziness.
Note that I am writing about longitudinal waves. Transverse waves on a fiber aren't going far under water. Your cans will need to be carefully designed to couple only to the longitudinal mode.
It's the ratio of speeds of sound that matters most. Think total internal reflection (though it's not really the same thing).
Internal frictional losses will certainly be less in steel than in string but the medium the fiber is immersed in shouldn't matter. I think that those depend on hysteresis in the stress-strain curve at small amplitudes. I seem to recall from reading about LIGO that quartz fiber is the champion there.