Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Length of f()

  1. Jan 11, 2005 #1


    User Avatar
    Science Advisor
    Homework Helper

    Are there any formulas for finding the length of the line traced out by some function f()?

    IE: If f(x) = cx + b where c and b are constants
    The length from 0 to x is:
    l(x) = sqr(x^2 + (cx)^2)

    But I don't know what to do for any polynomial above a line.
    I can make a summation for it, but don't know how to simplify.

    l(x) = lim[t->infinity](sum[n = 0 to t](sqr( 1/x^2 + (f(x*n/t)-f(x*(n+1)/t)))^2)))

    Basically the sum of arbitrarily small linear approximations.

    I figure the length of sin() and cos() are related to pi somehow...
  2. jcsd
  3. Jan 11, 2005 #2


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    As written in my calculus textbook:

    Theorem: Let f be afunction s.t. f ' is continuous on [a,b]. The lenght L of the curve joining the points R(a,f(a)) and S(b,f(b)) is given by

    [tex]L=\int_a^b\sqrt{1+(f '(x))^2}dx[/tex]
  4. Jan 11, 2005 #3


    User Avatar
    Science Advisor
    Homework Helper

    For a polynomial y=y(x) of degree larger or equal to three,the integral cannot be computed exactly.U'd be dealing with so-called "LEGENDRE ELLIPTIC INTEGRALS".
    Only numerical methids would work.

  5. Jan 11, 2005 #4


    User Avatar
    Science Advisor
    Homework Helper

    This is exactly how the length is defined if you take the limiting process.
    I'll use slightly different notation then you did.

    A small line segment of width [itex]\Delta x[/itex] can be approximated by:
    [tex]\sqrt{(\Delta x)^2+(\Delta y)}[/tex]

    by cutting up the interval [a,b] into n subintervals of width [itex]\Delta x[/itex], you can approximate the length by:

    [tex]L \approx \sum_{i=1}^n\sqrt{(\Delta x_i)^2+(\Delta y_i)}=\sum_{i=1}^n\sqrt{1+(\frac{\Delta y_i}{\Delta x_i})^2}\Delta x_i[/tex]

    The approximation gets better of n gets larger.
    The length of the curve is defined by:

    [tex]L=\lim_{n \to \infty}\sum_{i=1}^n\sqrt{1+(\frac{\Delta y_i}{\Delta x_i})^2}\Delta x_i=\int_a^b \sqrt{1+y'(x)^2}dx[/tex]

    The square root often makes it very difficult or impossible to evaluate explicitly. We'll have to resort to approximating the length. With Simpson's rule for example.
    Last edited: Jan 11, 2005
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook