1. Not finding help here? Sign up for a free 30min tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Stationary Wave Lab

  1. Nov 9, 2005 #1
    The lab experiment involved stationary sound waves within an open tube produced by a speaker placed near one of the ends of the tube.
    There are three questions I am unsure of.
    a) The antinodes near the ends of the tube are not located exactly where the tube ends. Why is this?
    b) Estimate from the data the effective length of the tube.
    c) Does the effective length depend on the frequency?
    Question a doesn't bother me too much actually. Haven't covered waves (in a more advanced sense, I suppose) yet this year, but we did study waves a bit last year in the intro physics course. If I remember correctly, the professor explained that this occurs because the antinode isn't fixed to where the tube ends, or something like that.
    For part b (and c also) I've proceeded in the following manner, but I'm not sure if I'm correct.
    The first ten harmonic modes were located and the corresponding frequency measured. Through a least squares regression (f(n)=v*n/(2L)) the speed of sound was calculated. (This was required for the lab report.)
    Then using the same equation and the speed of sound I obtained I calculated L for each mode. Indeed, the values of effective L differ from L0, within about 1 cm of it. Actually, they're still technically within the error I calculated for them. L0 is approximately 60 cm, just to give a sense of proportion.
    The values of the effective L don't seem to be linear with mode either. Perhaps amplitude affects it, since we did change the amplitude between frequencies to keep it within the range of the oscilloscope (and make it bearable on the hearing)
    Frankly, I don't think this is the correct way to measure the effective length. Most likely I simply calculated the original length for subsets of the data, which is why they're slightly different yet still mostly within the bounds of the error.
  2. jcsd
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Can you help with the solution or looking for help too?