Page 99 - Essentials of applied mathematics for scientists and engineers
P. 99

book   Mobk070    March 22, 2007  11:7








                                                 SOLUTIONS USING FOURIER SERIES AND INTEGRALS        89
                      3. Solve the steady-state conduction

                                                   u xx + u yy = 0

                                                   u x (0, y) = 0

                                                   u(a, y) = u 0
                                                   u(x, 0) = u 1
                                                   u y (x, b) =−h[u(x, b) − u 1 ]

                          Note that one could choose a length scale either a or b. Choose a. Note that if you
                          choose

                                                               u − u 1
                                                          U =
                                                               u 0 − u 1
                          there is only one nonhomogeneous boundary condition and it is normalized. Solve by
                          separation of variables.


                   5.3    FOURIER INTEGRALS
                   We consider now problems in which one dimension of the domain is infinite in extent. Recall
                   that a function defined on an interval (−c , c ) can be represented as a Fourier series
                                         c                   c
                                     1               1                   nπς          nπx
                                                        ∞
                             f (x) =        f (ς)dς +          f (ς)cos       dς cos
                                    2c               c                    c            c
                                      ς=−c             n=1 ς=−c
                                              c
                                       1                  nπς          nπx
                                         ∞
                                    +            f (ς)sin      dς sin                           (5.88)
                                       c                   c            c
                                        n=1 ς=−c
                   which can be expressed using trigonometric identities as
                                            c                   c
                                        1               1                   nπ
                                                           ∞
                                f (x) =        f (ς)dς +          f (ς)cos    (ς − x) dς        (5.89)
                                       2c               c                   c
                                         ς=−c             n=1 ς=−c
                   We now formally let c approach infinity. If     ∞  f (ς)dς exists, the first term vanishes. Let
                                                           ς=−c
                    α = π/c .Then

                                                      c
                                               2
                                                  ∞
                                        f (x) =         f (ς)cos[n α(ς − x)dς α                 (5.90)
                                               π
                                                 n=1
                                                    ς=0
   94   95   96   97   98   99   100   101   102   103   104