Page 268 - Excel for Scientists and Engineers: Numerical Methods
P. 268

Chapter 11



                              Numerical Integration of


                Ordinary DHerential Equations

                     Part 11: Boundary Conditions






                   In the preceding chapter, we saw that a differential equation of order n could
               be converted into a set of n first-order differential equations.  For example, if the
               problem to be solved  is a second-order differential equation, it is converted into
               two first-order differential equations; two "known" values of the function or its
               derivative  will  be  needed  in  order  to solve  the  problem.  In  the  second-order
               differential  equation  example  illustrated  in  Figure  10-16,  the  value  of  the
               function and its first derivative were both known at x  = 0.  The problem was then
               solved using the standard methods described in Chapter 10.
                   If information  about a second-order differential equation  is known  at two or
               more different values of the independent variable, then the problem is known as a
                boundary-value problem  (BVP).  The points  where  the  function  is known  are
                usually (but not always) the limits of the domain of interest - hence the term
               boundary-value  problem.  Problems  of  this  type  must  be  solved  by  different
                methods than those we applied to initial-value problems.
                   Two approaches are commonly used to solve boundary-value  problems: the
                "shooting'' method and the finite-difference method.  This chapter shows how to
                apply  these  methods  to  differential  equations  of  order  two;  fortunately,  most
                important physical  systems are described by  differential equations of  order  no
                higher than two.


                The Shooting Method

                   The shooting method  is a trial-and-error  method.  To solve a problem where
                the values of y  are known at xo and x,,  the boundaries of the interval of interest,
                we  set  up  the  problem  as though  it  were  an  initial-value  problem,  with  two
                llknowns" given at the same boundary - for example, at xo.  (See Figure  10-17
                for an example of an initial-value problem of this type: the two knowns, shown in
                bold, are the value of y at xo and a trial value of y' at xo.) Using the trial value of




                                                  245
   263   264   265   266   267   268   269   270   271   272   273