Page 504 - Electromagnetics
P. 504

H  (1) (z) and H  (2) (z) are the Hankel functions of the first and second kind of order ν, and
                         ν          ν
                        are related to the Bessel functions by
                                                   H (1) (z) = J ν (z) + jN ν (z),
                                                    ν
                                                   H (2) (z) = J ν (z) − jN ν (z).
                                                    ν
                        The argument z can be complex (as can ν, but this shall not concern us). When z is
                        imaginary we introduce two new functions I ν (z) and K ν (z), defined for integer order by
                                                            −n
                                                     I n (z) = j  J n ( jz),
                                                           π  n+1  (1)
                                                    K n (z) =  j  H n  ( jz).
                                                            2
                        Expressions for noninteger order are given in Appendix E.1.
                          Bessel functions cannot be expressed in terms of simple, standard functions. However,
                        a series solution to (A.124) produces many useful relationships between Bessel functions
                        of differing order and argument. The recursion relations for Bessel functions serve to
                        connect functions of various orders and their derivatives. See Appendix E.1.
                          Of the six possible solutions to (A.124),

                                                    
                                                    A ρ J ν (k c ρ) + B ρ N ν (k c ρ),
                                                    
                                                                   (1)
                                                    A ρ J ν (k c ρ) + B ρ H ν  (k c ρ),
                                                    
                                                    
                                                    
                                                     A ρ J ν (k c ρ) + B ρ H ν  (k c ρ),
                                                                   (2)
                                                    
                                             R(ρ) =                  (1)
                                                    A ρ N ν (k c ρ) + B ρ H  (k c ρ),
                                                                    ν
                                                    
                                                    A ρ N ν (k c ρ) + B ρ H  (2) (k c ρ),
                                                    
                                                    
                                                                     ν
                                                    
                                                    
                                                     A ρ H  (k c ρ) + B ρ H  (k c ρ),
                                                        (1)          (2)
                                                         ν            ν
                        which do we choose? Again, we are motivated by convenience and the physical nature
                        of the problem. If the argument is real or imaginary, we often consider large or small
                        argument behavior. For x real and large,
                                                        )
                                                          2          π    π
                                                J ν (x) →    cos x −   − ν   ,
                                                          πx         4    2
                                                        )
                                                          2         π     π
                                                N ν (x) →    sin x −  − ν    ,
                                                          πx        4     2
                                                        )
                                                          2   j(x− −ν )
                                                                    π
                                                                 π
                                                (1)
                                              H   (x) →     e    4  2 ,
                                                ν
                                                          πx
                                                        )
                                                          2  − j(x− −ν )
                                                                  π
                                                                     π
                                                (2)
                                              H ν  (x) →    e     4  2 ,
                                                          πx
                                                        )
                                                           1  x
                                                I ν (x) →    e ,
                                                          2πx
                                                        )
                                                          π  −x
                                               K ν (x) →    e  ,
                                                          2x
                        while for x real and small,
                                              J 0 (x) → 1,
                                                       2
                                              N 0 (x) →  (ln x + 0.5772157 − ln 2) ,
                                                      π
                        © 2001 by CRC Press LLC
   499   500   501   502   503   504   505   506   507   508   509