Green Function for Second Order Differential Equation with Initial Value Problem

  • Thread starter Thread starter mwg
  • Start date Start date
  • Tags Tags
    Green
Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
2 replies · 2K views
mwg
Messages
2
Reaction score
0

Homework Statement


need to find the green function for

d^2X / dt^2 + 2 dX/dt +(1+k^2)X= f(t)

with IV X(0)=dX/dt (=0) = 0

Homework Equations



eq. attached

The Attempt at a Solution


attached

whenever I try to solve the homogonous part, I get 0 due to the IV.
guessing my strategy is wrong.

what should it be??
 

Attachments

  • equation.JPG
    equation.JPG
    4.5 KB · Views: 472
  • iv.JPG
    iv.JPG
    3.1 KB · Views: 520
  • sol.JPG
    sol.JPG
    26.8 KB · Views: 465
Last edited:
Physics news on Phys.org
A "Green's function", [itex]G(t, \tau)[/itex], for a problem like this (the coefficient of the second derivative is 1) must satisfy several properties:
1) It must satisfy the homogenous d.e. in t for every [itex]\tau[/itex].
2) It must satisfy the boundary conditions.
3) It must be continuous.
4) The "jump" in the first derivative at [itex]t= \tau[/itex] must be one.

Here, the general solution to the homogeneous equation is
[tex]e^t(C_1cos(kt)+ C_2sin(kt))[/tex]
which means that the Green's function must be of the form
[tex]G(t, \tau)= \begin{pmatrix}e^t(Acos(kt)+ Bsin(kt) & 0\le t\le \tau \\ e^t(Ccos(kt)+ Dsin(kt) & \tau\le t\le 1\end{pmatrix}[/tex]

To satisfy the first boundary condition, since [itex]0\le\tau[/itex], we must have
[tex]G(0,\tau)= A= 0[/tex]
To satisfy the second boundary conditon, we must have
[tex]G_t(0,\tau)= A+ B= 0[/tex]
which, since A= 0, give B= 0. That is, [itex]G(t, \tau)[/itex] is identically 0 for all [itex]0\le t\le \tau[/itex].

But that does NOT say that [itex]G(t,\tau)[/itex] must be 0 for [itex]\tau\le t\le 1[/itex].

The continuity condition gives that
[tex]\lim_{t\to\tau^+}G(t,\tau)= e^{\tau}(C cos(k\tau)+ Dsin(k\tau))= \lim_{t\to\tau^-} G(t,\tau)= 0[/tex]
so we must have
[tex]e^{\tau}(C cos(k\tau)+ D sin(k\tau))= 0[/tex]
which leads to
[tex]D= -\frac{cos(k\tau)}{sin(k\tau)}C[/tex]

The "jump" condition on the derivative says that
[tex]\lim_{t\to\tau^-}G_t(t,\tau)- \lim_{t\to\tau^+}G_t(t,\tau)= 1[/tex]
which gives
[tex]e^\tau(Ck cos(kt)+ Dk sin(kt))+ e^\tau(-Ck sin(kt)+ Dk cos(kt))= 1[/tex]

those two equations give non-zero values for C and D.
 
Thank you so much!

one more question though, do I set sum lambda G = f(x) and solve? (sorry, couldn't figure the Latex here)