Page 500 - Electromagnetics
P. 500

requires a continuous superposition of eigenfunctions to satisfy the boundary conditions.
                        Let us try

                                    V (x, y) = [A x sinh k y x + B x cosh k y x][A y sin k y y + B y cos k y y].

                        The conditions at x = 0 and y = 0 require that B x = B y = 0.Thus
                                                    (x, y) = A sinh k y x sin k y y.
                                                  V k y
                        A single function of this form cannot satisfy the remaining condition at x = L 1 .So we
                        form a continuous superposition

                                                        ∞

                                             V (x, y) =   A(k y ) sinh k y x sin k y ydk y .  (A.113)
                                                       0
                        By the condition at x = L 1
                                              ∞

                                                A(k y ) sinh(k y L 1 ) sin k y ydk y = V 0 e −ay .  (A.114)
                                             0
                        We can find the amplitude function A(k y ) by using the orthogonality property
                                                          2     ∞

                                               δ(y − y ) =     sin xy sin xy dx.              (A.115)
                                                          π  0
                        Multiplying both sides of (A.114) by sin k y and integrating, we have

                                                            y

                                 ∞                  ∞                        ∞

                                   A(k y ) sinh(k y L 1 )  sin k y y sin k ydy dk y =  V 0 e −ay  sin k ydy.

                                                               y                        y
                                0                  0                        0
                        We can evaluate the term in brackets using (A.115) to obtain
                                      ∞                                 ∞
                                                      π

                                                                             −ay
                                        A(k y ) sinh(k y L 1 ) δ(k y − k ) dk y =  V 0 e  sin k ydy,
                                                              y
                                                                                   y
                                     0                2                0
                        hence
                                            π                      ∞
                                                                     −ay
                                              A(k ) sinh(k L 1 ) = V 0  e  sin k ydy.
                                                                            y
                                                        y
                                                 y
                                            2                    0
                        We then evaluate the integral on the right, solve for A(k y ), and substitute into (A.113)
                        to obtain
                                                        ∞

                                                  2V 0      k y  sinh(k y x)
                                         V (x, y) =                       sin k y ydk y .
                                                               2
                                                           2
                                                   π   0  a + k sinh(k y L 1 )
                                                               y
                        Note that our application of the orthogonality property is merely a calculation of the
                        inverse Fourier sine transform. Thus we could have found the amplitude coefficient by
                        reference to a table of transforms.
                          We can use the Fourier transform solution even when the domain is infinite in more
                        than one dimension. Suppose we solve (A.111) in the region
                                           0 ≤ x < ∞,   0 ≤ y < ∞,  −∞ < z < ∞,
                        subject to
                                                V (0, y) = V 0 e −ay ,  V (x, 0) = 0.

                        © 2001 by CRC Press LLC
   495   496   497   498   499   500   501   502   503   504   505