1. Limited time only! Sign up for a free 30min personal 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!

Measuring torque

  1. Nov 25, 2014 #1

    MMS

    User Avatar

    Hello everyone.
    I'm working on a physics lab project and I've reached a barrier.
    I need to measure the torque on a closed loop in a magetic field and I can't reach a way to do so. Of course, excluding all the torque measuring devices and such which I believe we don't have at the lab.

    Any ideas on how to measure it?


    Thank you.
     
  2. jcsd
  3. Nov 25, 2014 #2

    ShayanJ

    User Avatar
    Gold Member

    I think you can use a spring with a known spring constant(which should be chosen in accordance with the range of torque being applied to the loop, you should have at least a very rough estimate of the amount of torque). Then you attach the loop to one end of the spring and fix the other end. Then you measure the elongation of the spring. Now you calculate [itex] \frac 1 2 k \delta x^2 [/itex] which should be approximately equal to [itex] \int_{\theta_i}^{\theta_f} \tau d\theta [/itex]. If the rotation angle is small, I think you can take the torque to be constant and that gives you the amount of torque.
     
    Last edited: Nov 25, 2014
  4. Nov 25, 2014 #3

    MMS

    User Avatar

    Interesting.
    I can estimate what the torque should be by calculations so I believe that shouldn't be a problem. Just a question: how can I deal with the integral on the torque that you've given with large angles? Why can't I treat tge torque as a constant in both cases (small and large)?
     
  5. Nov 25, 2014 #4

    ShayanJ

    User Avatar
    Gold Member

    Take the interval [itex] (a,b) [/itex] where [itex] b=a+H [/itex]. Now I want to calculate the integral [itex] \int_a^b f(x) dx [/itex](f being any function). Now I substitute [itex] x=a+\frac H 2 +h \Rightarrow dx=dh [/itex].
    [itex]
    \int_{-\frac H 2}^{\frac H 2} f(a+\frac H 2+h) dh=\int_{-\frac H 2}^{\frac H 2} [\sum_{n=0}^\infty \frac{ f^{(n)}(a+\frac H 2)}{n!} h^n] dh=\int_{-\frac H 2}^{\frac H 2} [f(a+\frac H 2)+O(h)] dh=f(a+\frac H 2)H+O(H^3) \\ \Rightarrow f(a+\frac H 2)=\frac{1}{H} \int_a^b f(x)dx+O(H^2)
    [/itex]
    So by taking [itex] \frac{1}{\theta_f-\theta_i}\int_{\theta_i}^{\theta_f} \tau d\theta[/itex], to be the torque at [itex] \theta=\frac{\theta_f+\theta_i}{2}[/itex], you have an error of the order [itex](\theta_f-\theta_i)^2[/itex].
    All of this is because the torque may not be constant during the rotation and how it changes is unknown to us. You may think its actually constant from some calculations but that only makes the approximation better because in reality, things aren't as clean as on the paper!
     
    Last edited: Nov 25, 2014
  6. Nov 25, 2014 #5

    MMS

    User Avatar

    Thanks. I'll make sure to go over it as when I get home (I'm on my phone).
    Just a couple more questions if I may.
    1. Did you mean to connect the spring directly to the loop or by a string or something?
    2. Aren't there other things that contribute to the work in the system (except for the spring)?
     
  7. Nov 25, 2014 #6

    ShayanJ

    User Avatar
    Gold Member

    Connect them directly! Otherwise those things in between, will get some energy and that worsens the approximation.

    I think the spring's potential energy is, to a good approximation, the only thing that gets the loop's rotational energy.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Measuring torque
  1. Measuring Torque (Replies: 1)

  2. Torques and Normals (Replies: 2)

Loading...