Page 197 - Computational Fluid Dynamics for Engineers
P. 197

6.3  Finite-Difference  Method                                        185




             I
          77
            \
              A                             D
           1











             B                             C
          0                                      ^
            0               n/2             n

         Fig.  6.5.  Computational  plane  for  flow  over  a  circular  cylinder.


         and  with
                                                                          (6.3.9b)

         perform  numerical  calculations  in  the  computational  plane  (£, rj)  (Fig.  6.5).
         Here the  upper  boundary  line,  r\ =  1, represents  the  body  surface  and  the  lower
         boundary  line,  r\ =  0,  represents  the  freestream  at  infinity.  The  left  and  right
         symmetry  lines  at  £(=  6)  =  0  and  £  =  TT become  boundary  conditions  repre-
         sented  by  AB  and  DC,  respectively.
            To express  Eq.  (6.3.6)  and  its boundary  conditions  given  by Eqs.  (6.3.7)  and
         (6.3.8)  in the computational  plane,  we use the  transformation  in Eq.  (6.3.9)  and
         write  the  Laplace  equation  as  (see  Problem  2.20)

                                           d±
                                  2&±  +             - 0                  (6.3.10)
                                    dr\ 2  drj + &e
         and  its  boundary  conditions  as


                                           along  AD                     (6.3.11a)
                              Of]
                                   c o s
                              1      £     along  BC                     (6.3.11b)
                              0 =
                                    V
                              I-           along  AB  and  DC             (6.3.11c)

            We  now  use  the  procedure  in  Section  4.5  to  write  the  finite-difference  ap-
         proximations  to  Eqs.  (6.3.10)  and  (6.3.11).  For  convenience,  we  drop  the  tilde
         on  (j). Replacing  each  second  derivative  in  Eq.  (6.3.10)  by  Eq.  (4.3.10)  and  the
                                             j
         first  derivative  by  Eq.  (4.3.7),  all  at  (i, ),  we  obtain
   192   193   194   195   196   197   198   199   200   201   202