Page 160 - Essentials of applied mathematics for scientists and engineers
P. 160

book   Mobk070    March 22, 2007  11:7








                     150  ESSENTIALS OF APPLIED MATHEMATICS FOR SCIENTISTS AND ENGINEERS

                            Let the sine transform of u be denoted by S n u(r,θ, z) = U n (r, n, z) with respect to θ
                       on the interval (0, π). Then

                                            2               2      2
                                          ∂ U n   1 ∂U n   n      ∂ U n
                                                +       −    U n +      + q(r)S n (1) = 0
                                           ∂r  2  r ∂r     r 2     ∂z 2
                       where S n (1) is the sine transform of 1, and the boundary conditions for u(r,θ, z)on θ have
                       been used.
                            Note that the operator on   in the r coordinate direction is

                                                         1 d     d      n 2
                                                     
                             2
                                              (r,µ j ) =       r      −      =−µ
                                                                                   j
                                                         r dr    dr     r 2
                       With the boundary condition at r = 1 chosen as  (1,µ j ) = 0 this gives the kernel function as
                         = rJ n (r,µ j ) with eigenvalues determined by J n (1,µ j ) = 0
                            We now apply the finite Hankel transform to the above partial differential equation and
                       denote the Hankel transform of U n by U jn .
                            After applying the boundary condition on r we find, after noting that

                                                   N β [U n (z, 1)] = S n (1)

                                                    N [ (1, z)] =−µ j J n+1 (µ j )

                                                     β
                                                    2
                          2                        d U jn
                       −µ U jn + µ j J n+1 (µ j )S n (1) +  2 + Q j (µ j )S n (1) = 0. Here Q j (µ j ) is the Hankel trans-
                          j                         dz
                       form of q(r).
                            Solving the resulting ordinary differential equation and applying the boundary condition
                       at z = 0,

                                                         Q j (µ j ) + µ j J n+1 (µ j )
                                     U jn (µ j , n, z) = S n (1)            [1 − exp(−µ j z)]
                                                                 µ 2 j

                            We now invert the transform for the sine and Hankel transforms according to Eq. (9.19)
                       and find that

                                                           ∞
                                                       ∞
                                                    4        U jn (µ j , n, z)
                                         u(r,θ, z) =                      J n (µ j r)sin(nθ)
                                                    π         [J n+1 (µ j )] 2
                                                      n=1 j=1
                            Note that

                                                                      n
                                                      S n (1) = [1 − (−1) ]/n
   155   156   157   158   159   160   161   162   163   164   165