Page 375 - Computational Fluid Dynamics for Engineers
P. 375

366                                   12.  Compressible  Navier-Stokes  Equations



         and  approximated  by
                                    =  T^i  -  (Ti£  -  Ti ti)/2           (12.6.8)
                                 T w
         with  the  known  interior  values.  The  density  g w  and  (Et) w  can  be  calculated
         from  Eq.  (12.6.4).
                                    -WwML               Pw
                                       lv L
                                     J
                                      W - oo
                              Qw  =        ,   (Et t  w                    (12.6.9)
                                      T w   '  —        7 - 1
         The  wall  temperature  T w  can  also  be  computed  from
                                               V?
                                CpT w =  CpTi +  —±-  =  const.          (12.6.10)
                                              2T™
         where
                                       _jR_
                                               p =  QRT                 (12.6.11a)
                                            !
                                       7 - l
                                 2
                      V 2  =  u 2  + v ,  M  2  =  - ^ ,  a 2  =  jRT   (12.6.11b)
                                             2
                                            a
         Since  Eq.  (12.6.9)  is  in  dimensional  form,  we express  it  in  dimensionless  form,

                   •Lw  _  rp*  li.  _.*£.               V?
                          J        +
                   rp     U>
                   -* oo       J- oo  ^-LooCp   oo  \  ^J-lCp
                       = II    1  +  (7                 i  +  { ^=AM 2
                          T                a i /

         or
                                              ( 7 - 1 )
                                 r w  =  Ti 1  +     Mi                  (12.6.12)

         Our  studies  for  this  particular  problem  show  that  both  choices  (Eqs.  (12.6.8)
         and  (12.6.12))  of  computing  T w  are  the  same.
            On  the  line  of  symmetry,  y  =  H,  we  set

                                     9(gu)            dE t
                                           =  v  =  0,    =  0,          (12.6.13)
                            dy         dy              dy
         The  outflow  boundary  conditions  at  i  =  I,  0  <  j  <  J  are  again  obtained  by
         using  second-order  accurate  extrapolation,

                                       2
                                 Qij  = Qi-i,j  ~  QI-2J
                                                                         (12.6.14)
                              (QV)I,J  =  2(gv) I- 1j  -  (gv)i-2j

                              (Et)ij  =  2(E t)i-ij  -  {Et) 1-2,3
   370   371   372   373   374   375   376   377   378   379   380