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

I Divergence Theorem and Gauss Law

  1. Mar 24, 2016 #1
    Divergence theorem states that
    $\int \int\vec{E}\cdot\vec{ds}=\int\int\int div(\vec{E})dV$
    And Gauss law states that

    $\int \int\vec{E}\cdot\vec{ds}=\int\int\int \rho(x,y,z)dV$
    If $\vec{E}$ to be electric field vector then i could say that
    $div(\vec{E})=\rho(x,y,z)$
    However i cant see any reason for that since $\rho(x,y,z)$ to be unit charge
     
  2. jcsd
  3. Mar 25, 2016 #2

    BvU

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    Hi. (Use double $ for displayed, double # for in-line ##LaTeX## ).

    Nevertheless, ## \operatorname {div} \vec E = { \rho\over \epsilon_0 }## is one of the Maxwell equations. So ## \left | \vec E \right | \propto \rho## as you would expect.
     
  4. Mar 25, 2016 #3

    vanhees71

    User Avatar
    Science Advisor
    2016 Award

    The point is that your integral equations are valid for any volume ##V## and its boundary ##\partial V## (which of course is a closed surface):
    $$\int_{\partial V} \mathrm{d}^2 \vec{F} \cdot \vec{E}=\int_V \mathrm{d}^3 x \rho(\vec{x}).$$
    This is not very useful in practice, because integrals are not as easy to use as derivatives. The idea thus is to make the volume very small around some specific point ##\vec{x}##. On the right-hand side of Gauss's Law you can make ##V## so small that ##\rho## doesn't vary too much over its extension, i.e., you can write
    $$\int_V \mathrm{d}^3 x' \rho(\vec{x}') \simeq \rho(\vec{x}) V.$$
    Now to the left-hand side. Think of the volume as a little cube with edges parallel to a Cartesian coordinate system. It's best to draw the meaning of the integral to see that the surface integral can be approximated by
    $$V (\partial_x E_x(\vec{x})+\partial_y E_y(\vec{x}) + \partial_z E_z(\vec{x}))=V \mathrm{div} \vec{E}(\vec{x}).$$
    Now you can cancel the little volume ##V## from both sides of the equation to get the local form of Gauss's Law, which is one of Maxwell's equations (here written in Heaviside-Lorentz units, where you don't have the confusing conversion factors as in the SI units):
    $$\mathrm{div} \vec{E}(\vec{x})=\rho(\vec{x}).$$
     
  5. Mar 25, 2016 #4
    Hi
    Why did you put $$d^2$$ and $$ \vec{F}$$? I saw it on wikipedia too but did not get the idea.
     
  6. Mar 25, 2016 #5
    Well after your reply, i went to wiki and i saw the derivations of those equetion. Also thanks for $$ thing :) i did not know.
     
  7. Mar 25, 2016 #6

    vanhees71

    User Avatar
    Science Advisor
    2016 Award

    ##\mathbb{d}^2 \vec{F}## is my convention for the surface-element vector. If your surface is parametrized with generalized coordinates ##q^k## (##k \in \{1,2 \}##) then
    $$\mathbb{d}^2 \vec{F}=\mathrm{d} q^1 \, \mathrm{d} q^2 \; \frac{\partial \vec{x}}{\partial q^1} \times \frac{\partial \vec{x}}{\partial q^2}.$$
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted



Similar Discussions: Divergence Theorem and Gauss Law
  1. Gauss's Theorem (Replies: 7)

Loading...