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

Green's Theorem Applied

  1. Nov 4, 2014 #1


    User Avatar
    Gold Member

    Hi PF!

    So let's say I have some vector field, call it ##\vec{F}## and let ##\vec{F}## be 2-Dimensional and suppose I wanted to compute ##\iint_D \nabla \cdot \vec{F} dD##. Using green's theorem we could write ##\iint_D \nabla \cdot \vec{F} dD = - \int_{\partial D} \vec{F} \cdot \hat n dS## where ##\hat n## is the outward oriented surface normal. Now if ##D = [0,1] \times [0,1]## I may use the formula to find the solution. Can you tell me what I'm doing wrong? $$\int_{\partial D} \vec{F} \cdot \hat n dS = \int_0^1 \vec{F} \cdot (- \hat j)dx + \int_0^1 \vec{F} \cdot ( \hat i)dy+\int_1^0 \vec{F} \cdot (\hat j)dx+\int_1^0 \vec{F} \cdot (\hat i)dx$$

    I know there is another way to write Green's Theorem but how to do it this way?

    Thanks a ton!

  2. jcsd
  3. Nov 5, 2014 #2


    User Avatar
    Gold Member

    Please help me out! I'm not looking for an answer, just looking for what I am doing wrong.
  4. Nov 5, 2014 #3


    User Avatar
    Gold Member

    For the record what I have on the R.H.S is what is wrong.
  5. Nov 6, 2014 #4
    You've almost got it, except for the errant - sign in the red equation, and the latter two integrals, which are not proper line integrals, as the upper limit must be greater than the lower limit (or you must used a signed differential to represent that they are parameterized in the opposite orientation to the prior two integrals. Either way is fine.) Using a standard positive differential, we could use the following explicit parameterizations to traverse the unit square in the counterclockwise orientation:
    [tex]r_1(t) = (t, 0), 0 \leq t \leq 1[/tex]
    [tex]r_2(t) = (1, t), 0 \leq t \leq 1[/tex]
    [tex]r_3(t) = (1 - t, 1), 0 \leq t \leq 1[/tex]
    [tex]r_4(t) = (0, 1 - t), 0\leq t \leq 1[/tex]
    And your integrals would then be:
    [tex]\int_0^1 \vec{F}(t, 0) \cdot (- \hat j)\, dt + \int_0^1 \vec{F}(1, t) \cdot ( \hat i)\, dt+\int_0^1 \vec{F}(1 - t, 1) \cdot (\hat j)\, dt+\int_0^1 \vec{F}(0, 1 - t) \cdot (-\hat i)\, dt[/tex]
    When it comes to vector calculus, you just have to be very careful about orientations. They are as sneaky as sign errors. :-)
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Discussions: Green's Theorem Applied
  1. Green's theorem. (Replies: 4)

  2. Green's Theorem (Replies: 4)

  3. Green Theorem (Replies: 1)

  4. Green's Theorem? (Replies: 2)