Page 237 - Computational Fluid Dynamics for Engineers
P. 237

226                                             7.  Boundary-Layer  Equations


                                                                                n
         This  test  is  done  in  MAIN.  If  this  test  is  satisfied,  we  set  r/ e (x n+1 )  =  rj e(x ).
         Otherwise,  we  call  GROWTH  and  set  J new  =  J Q\d  +  £, where  t  is  a  number  of
                                                                              r
                                                             (
         points,  say  t  =  1.  In  this  case  we  also  specify  values  of /f  > ^?>^?>&?)  f°  the
                                                                        n
                                                           1
         new  rij  points. We take the  values  of u 3; =  1, v^  =  0, j  =  (r/j — r]^)u j + j ,  and
                                                                           /
                                                         /
                                                          1
                                                  1 1
                                                                    1
         Vf  =  V}.  This  is  also  done  for  the  values  of /J " ,  v?" ,  and  6?" .
         7.4.5  Subroutine  COEF3
         This  is one  of the  most  important  subroutines  of BLP. It  defines  the  coefficients
         of the  linearized  momentum  equation  given  by  Eqs.  (7.3.23)  and  (7.3.26).
         7.4.6  Subroutine  SOLV3

         The  solution  of  Eq.  (4.4.29)  by  the  block-elimination  method  discussed  in  sub-
         section  4.4.3  can  be  obtained  by  using  the  recursion  formulas  given  by  Eqs.
         (4.4.32)  and  (4.4.34),  and  determining  the  expressions  such  as  Aj,  Fj,  Wj  and
         6j.  To  describe  the  procedure  let  us  first  consider  Eq.  (4.4.32).  Noting  that  the
                                                                        7
         i~j  matrix  has the  same structure  as  Bj  and  denoting  the  elements  of j,  by  7 ^
         (i,k  =  1, 2, 3),  we  can  write  Tj  as
                                      (711)  (7l2)j  (7is)j
                                r,=   (721)  (722 )j  (723)j             (7.4.11a)
                                        0      0     0

           Similarly,  if the  elements  of  Aj  are  denoted  by  o ^  we can  write  Aj  as  [note
        that  the  third  row  of  Aj  follows  from  the  third  row  of  Aj  according  to  Eq.
         (4.4.32c)]

                         (an)j  (an)j   («i3)j
                   ^     (a 2i)j  (o 2 2 )j  (a 23)j  0  <  j  <  J  -  1  (7.4.11b)
                            0    - 1     -hj+i/2

        and  for  j  =  J,  the  first  two  rows  are  the  same  as  the  first  two  rows  in  Eq.
         (4.4.32b),  but  the  elements  of the  third  row,  which  correspond  to the  boundary
        conditions  at  j  =  J,  are  (0,1,0).
           For  j  — 0,  AQ =  AQ; therefore  the  values  of  (0^)0  are
                            (an)o  =  1  (012)0  =  0  (013)0  =  0
                                                                         (7.4.12a)
                            (021)0  =  0  (0^2)0  =  1  (023)0  =  0

        and  the  values  of  (7^)1  are

            (711)1  =  - 1  (712)1  =  --hi      (713)1  =  0
                                                                         (7.4.12b)
            (721)1  =  (54)1  (723)1  =  - 2  ——  (722)l  =  (56)l  +  (723)l
   232   233   234   235   236   237   238   239   240   241   242