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!

Homework Help: Related to stokes theorem

  1. Apr 5, 2010 #1
    1. The problem statement, all variables and given/known data

    [tex] \nabla \times f \vec{v} = f (\nabla \times \vec{v}) + ( \nabla f) \times \vec{v} [/tex]

    Use with Stoke's theorem

    [tex] \oint _C \vec{A} . \vec{dr} = \int \int _S (\nabla \times \vec{A}) . \vec{dS} [/tex]

    to show that

    [tex] \oint _c f \vec{dr} = \int \int _S \vec{dS} \times \nabla f[/tex]

    2. Relevant equations

    I think the only that you really need to know is the the scalar triple product.

    3. The attempt at a solution

    Ok, I allowed

    [tex] \vec{A} = f \vec{v} [/tex]

    And subsituted into Stokes' theorem

    [tex] \oint _C f\vec{v} . \vec{dr} = \int \int _S (\nabla \times f \vec{v}) . \vec{dS} [/tex]

    which gives

    [tex] \oint _C f\vec{v} . \vec{dr} = \int \int _S f ((\nabla \times \vec{v}) + ( \nabla f) \times \vec{v} ). \vec{dS} [/tex]

    My problem is that I got to this stage, and thought that this would work really great if

    [tex] \nabla \times \vec{v} = 0 [/tex].

    So I got stuck on this for a while, looked at my textbook, and they said that we let

    [tex] \vec{A} = f \vec{v} [/tex] where [tex] \vec{v} [/tex] is a CONSTANT VECTOR.

    Hence, the curl of a constant vector is zero, and the RHS becomes

    [tex] \int \int _S f ( \nabla f) \times \vec{v} ). \vec{dS} [/tex]

    which we rearrange with the triple scalar product identity to give

    [tex] \int \int _S f ( \vec{dS} \times \nabla f) ) .\vec{v} [/tex]

    So now the thing looks like this:

    [tex] \oint _C f\vec{v} . \vec{dr} = \int \int _S f ( \vec{dS} \times \nabla f) ) .\vec{v}

    and cancelling [tex] \vec{v} [/tex] from both sides as it's constant, gives us:

    [tex] \oint _c f \vec{dr} = \int \int _S \vec{dS} \times \nabla f[/tex]

    The problem

    My problem is that this restricts the kind of vector field [tex] \vec{A} [/tex] can be, as it has to be a product of some function times a constant vector.

    Does this make the derived identity non-universal, or is this a necessary step in order to arrive at the identity?
    Last edited: Apr 5, 2010
  2. jcsd
  3. Apr 5, 2010 #2


    User Avatar
    Homework Helper
    Gold Member

    The vector identity you proved involves only [itex]f[/itex]. So, who cares what you restricted [itex]\textbf{A}[/itex] to be? The identity is valid for any scalar function [itex]f[/itex], and that's all you were asked to prove.
  4. Apr 5, 2010 #3
    Yeah you're right. I don't know why I got hung up on this, I guess I was reading too much into it.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook