1. Not finding help here? Sign up for a free 30min 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!

Boundary Value Problem

  1. Oct 16, 2004 #1
    I need some help starting off on this question.

    Electrostatic potential [tex]V (x,y)[/tex] in the channel [tex]- \infty < x < \infty, 0 \leq y \leq a[/tex] satisfies the Laplace Equation

    [tex]\frac{\partial^2 V}{\partial x^2} + \frac{\partial^2 V}{\partial y^2}= 0[/tex]

    the wall [tex]y = 0[/tex] is earthed so that

    [tex]V (x,0) = 0[/tex]

    while the potential on the wall [tex]y = a[/tex]

    [tex]V (x,a) = V_0 \cos{kx}[/tex] where [tex]V_0 , k[/tex] are positive constants.

    By seeking a soln of an appropriate form, find [tex]V (x,y)[/tex] in the channel.
     
  2. jcsd
  3. Oct 16, 2004 #2

    arildno

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    Dearly Missed

    When in doubt, try separation of variables!
    V(x,y)=F(x)G(y).
    Inserting this into Laplace, yields, by rearrangement:
    [tex]\frac{G''(y)}{G(y)}=-\frac{F''(x)}{F(x)}[/tex]
     
  4. Oct 16, 2004 #3
    Thanks, I was trying to get it in the form [tex]V (x,y) = F (a) V_0 \cos{kx}[/tex] then sub the values into the Laplace eqn. Is that going about it the wrong way?
     
  5. Oct 16, 2004 #4

    arildno

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    Dearly Missed

    You DO mean (in my notation!)
    [tex]V(x,y)=G(y)V_{0}\cos(kx)[/tex]???
    If so, then it will work.
    Note that the separation of variables method in this case implies:
    [tex]G''(y)=k^{2}G(y)[/tex]
     
  6. Oct 16, 2004 #5
    Hmm im not really keeping up with you here sorry. How can I find things out about F(x) and G(y) (in your notation) when we have just introduced them?

    --------edit----------

    oh yes sorry im with you, then differentiate V (x,y) = G(y) V_0 cos (kx) to get the different bits to go in the Laplace eqn? ahh!
     
    Last edited: Oct 16, 2004
  7. Oct 16, 2004 #6

    arildno

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    Dearly Missed

    Your lightbulb switched on in your edit, I see..:wink:
    Hint:
    You should be able to see that G(y)=ASinh(ky), where A is some constant, and Sinh() the hyperbolic sine function.
     
  8. Oct 16, 2004 #7
    Now I've got G''(y) - k^2 G (y) = 0, still on the right track yeh?
     
  9. Oct 16, 2004 #8

    arildno

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    Dearly Missed

    Yes, you are.
     
  10. Oct 17, 2004 #9
    From this how do I get to G(y)=ASinh(ky)?
     
  11. Oct 17, 2004 #10

    arildno

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    Dearly Missed

    Note that:
    [tex]\frac{d}{dy}ASinh(ky)=kACosh(ky),\frac{d^{2}}{dy^{2}}ASinh(ky)=k^{2}ASinh(ky)[/tex]
    This shows that ASinh(ky) is a solution for G(y).
    Similarly, you may show that BCosh(ky) is another solution for G(y).
    Your general solution is therefore:
    G(y)=ASinh(ky)+BCosh(ky)
    Apply the boundary condition at y=0 to prove that B=0
     
  12. Oct 17, 2004 #11
    Thanks for being so patient. How can I show that ASinh(ky) is a soln for G(y)? Im used to looking at the differential eqn and substituting

    G(y) = e^{ry} so G'(y) = re^{ry} and G''(y) = r^2. e^{ry}

    Is there no such substitution to show that ASinh(ky) & BCosh(ky) are solns?
     
  13. Oct 17, 2004 #12

    arildno

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    Dearly Missed

    Your approach is ABSOLUTELY correct!
    Do you agree with the following result, using your method:
    [tex]r=\pm{k}[/tex]??
     
  14. Oct 17, 2004 #13

    Yeh which gives gen soln G (y) = Ae^{ky} + Be^{-ky}

    but the boundary condition at y=0 doesnt allow for B so we are left with

    G(y) = Ae^{ky} ?
     
  15. Oct 17, 2004 #14

    arildno

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    Dearly Missed

    A bit too fast there..
    Let's write the general solution for G(y) as follows:
    [tex]G(y)=K_{+}e^{ky}+K_{-}e^{-ky}[/tex]
    where the K's are constants to be determined by boundary conditions.
    Prior to that step, however, let's rewrite the general solution as:
    [tex]G(y)=(K_{+}+K_{-})\frac{e^{ky}+e^{-ky}}{2}+(K_{+}-K_{-})\frac{e^{ky}-e^{-ky}}{2}[/tex]
    1. We now set [tex]B=K_{+}+K_{-},A=K_{+}-K_{-}[/tex]
    (Clearly, A and B are as arbitrary as the K's!)
    2) We now recognize:
    [tex]Cosh(ky)=\frac{e^{ky}+e^{-ky}}{2}[/tex]
    [tex]Sinh(ky)=\frac{e^{ky}-e^{-ky}}{2}[/tex]
    Or, we may rewrite G(y) as:
    [tex]G(y)=ASinh(ky)+BCosh(ky)[/tex]
     
  16. Oct 17, 2004 #15
    right now I need to find

    B: iv done this and basically its because cosh can never be zero which implies B must be zero.

    A: at y=a V(x,a) = V_0 cos{kx}

    therefore for y=a, G(y)=1

    ASinh(ka) = 1 is that right?

    hmm im not sure about this.
     
  17. Oct 17, 2004 #16

    arildno

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    Dearly Missed

    Yep, so your solution is:
    [tex]V(x,y)=V_{0}\frac{Sinh(ky)\cos(kx)}{Sinh(ka)}[/tex]
     
  18. Oct 17, 2004 #17
    Thank you so much, I understand much more now than I did even yesterday!
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Boundary Value Problem
  1. Boundary value problem (Replies: 3)

  2. Boundary value problem (Replies: 1)

  3. Boundary Value Problem (Replies: 9)

Loading...