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: Electric field of an infinite sheet of charge.

  1. Jul 15, 2012 #1
    1. The problem statement, all variables and given/known data
    Using direct integration show that the electric field due to an infinite sheet with charge density σ is independent from the distance from the sheet and equals [itex]\frac{σ}{2\epsilon_{0}}[/itex]

    2. Relevant equations

    [itex]\int\int k \frac{Q}{r^{3}}dxdy[/itex]

    Which should lead directly to the equation:


    3. The attempt at a solution

    We had to do the same thing with a line of constant charge and he told us we could use that result in this problem, which is basically the first half of the problem and gives you the result [itex]E=\frac{\lambda}{2\pi\epsilon_{0}z}[/itex]

    I don't know if my problem is with the surface integral or the physics but this is just baffling me. I can't get to the "middle" with what I start with and if I start with what he gave us, I can't get to the solution.

    Edit: I got the answer the first time by getting rid of the double integral and thinking of it as an infinite number of concentric rings of charge and integrating that but he didn't want it that way. He wants it with the double integral.
  2. jcsd
  3. Jul 15, 2012 #2
    [itex]kQ/r^3[/itex]? Who told you that? It should be more like, say

    [tex]\int \int k \frac{\sigma}{r^2} \; dx \; dy[/tex]
  4. Jul 15, 2012 #3
    You need to account for the direction of the field so on the end of yours should be [itex]\frac{\hat{r}}{\left|\hat{r}\right|}[/itex]. The magnitude of [itex]\hat{r}[/itex] is gong to be r again, which gives you the [itex]r^{3}[/itex] on the bottom.
  5. Jul 15, 2012 #4
    It is possible to do it his way by using a sequence of rectangular straight line elements of width dy oriented in the x direction, and integrating over these elements.
  6. Jul 15, 2012 #5


    User Avatar
    Homework Helper
    Gold Member
    2017 Award

    Also, make sure that when you use E = λ/(2[itex]\pi[/itex][itex]\epsilon[/itex]oz) for an infinitely long thin strip of the plane that you realize z here is not the perpendicular distance from the plane to the field point. It's the perpendicular distance from the strip to the field point. It might be better to write it as r rather than z.
  7. Jul 15, 2012 #6
    If I did it this way it'd reduce it down to a single integral again, and he specifically wants the double.

    I think this is were I'm hanging up. So when I do that, am I going to take it times the unit vector again, or is that all taken care of in the beginning? Maybe because it seems so simple I'm just psyching myself out because all I need to get rid of from that mid-step is a pi.
  8. Jul 15, 2012 #7
    I think you should consider converting from rectangular to cylindrical coordinates. It makes the problem much easier, as one of the two integrals is trivial.
  9. Jul 15, 2012 #8
    I'm not entirely sure how to plug the cylindrical coordinates into my equation.
  10. Jul 15, 2012 #9


    User Avatar
    Homework Helper
    Gold Member
    2017 Award

    Yes, you will need to take into account that the field from one of the strips will point directly away from the strip. Since you know the total field from all the strips will be in the z direction, you will need to take the z-component of the field of each strip and integrate this z-component for all the strips. This will be your y-integral from -∞ to ∞. This is what Chestermiller was hinting at.

    When you start with the double integral you can first fix y and integrate over x from -∞ to ∞. But that will clearly just produce the field of a thin strip of thickness dy. So, rather than actually doing the integral over x, you can just use the given result for an infinite line of charge. You need to convince yourself that λ = σdy for the thin strip.
  11. Jul 15, 2012 #10
    What are x, y, z in terms of cylindrical coordinates?
  12. Jul 15, 2012 #11
    Thank you, the light bulb just clicked on. I've been slaving over this thing for a week and I just couldn't get it to work out.

    Thanks a lot!

    Well, in the cylindrical coordinates [itex]p=\sqrt{x^{2}+y^{2}}[/itex] and [itex]\psi=arcsin(\frac{y}{p})[/itex] and z stays the same. I'm just not sure how to represent r in terms of p [itex]\psi[/itex] and z.
  13. Jul 15, 2012 #12
    I mean, you want to say x is some function of the cylindrical coordinates--e.g. [itex]x = \rho \cos \psi[/itex]. Same process for y and z.

    The point is your integral turns into this:

    [tex]\frac{\sigma}{4\pi \epsilon_0} \int_0^{2\pi} \int_0^\infty \frac{z}{(\rho^2 + z^2)^{3/2}} \rho \; d\rho \; d\psi[/tex]

    And it should be fairly obvious by this point that you can attack this with a variable substitution (see that [itex]d(\rho^2) = 2 \rho \; d\rho[/itex]). Though I see through your addendum to the first post that you already did it through concentric rings and this form should be familiar anyway.
  14. Jul 16, 2012 #13
    That's kinda what I was afraid of when I started thinking about it like you were saying. I'll probably work it out like this too to see if I can get some extra credit or something. Thanks for all the help!
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook