Page 112 - Aerodynamics for Engineering Students
P. 112

Governing equations of fluid mechanics  95

         If we wish  the model tests to produce useful information about general characteristics of the
         prototype’s flow field, in particular estimates for its aerodynamic drag, it is necessary for the
         model and prototype to be dynamically similar, i.e. for the forces to be scale invariant. It can be
         seen from Eqn (2.103) that this can only be achieved provided

                                         Re,  = Re,,                       (2.104)

         where suffices m and p  denote model and prototype respectively.
           It is not usually practicable to use any other fluid but air for the model tests. For standard
         wind-tunnels the air properties in the wind-tunnel are not greatly different from those experi-
         enced by the prototype. Accordingly, Eqn (2.104) implies that

                                        u, =-up                            (2.105)
                                             LP
                                             L,
         Thus, if we use a 1/5-scale model, Eqn (2.105) implies that U,  = 5UP. So a prototype speed of
         100 km/hr (c. 30 m/s) implies a model speed of 500 km/hr (c. 150 m/s). At such a model speed
         compressibility effects are no  longer  negligible.  This  illustrative  example  suggests that,  in
         practice, it is rarely possible to achieve dynamic similarity in aerodynamic model tests using
         standard wind-tunnels.  In  fact, dynamic  similarity can usually  only be achieved in aerody-
         namics by using very large and expensive facilities where the dynamic similarity is achieved by
         compressing the air (thereby increasing its density) and using large models.
           In this example we have briefly revisited the material covered in Section 1.4. The objective
         was to show how the dimensional analysis of the Navier-Stokes  equations (effectively the exact
         governing equations of the flow field) could establish more rigorously the concepts introduced
         in Section 1.4.


           2.10  Exact solutions of the Navier-Stokes
                   equations


         Few physically realizable exact solutions of the Navier-Stokes  equations exist. Even
         fewer are of  much  interest  in  Engineering. Here we  will  present  the  two  simplest
         solutions, namely Couette flow (simple shear flow) and plane Poiseuille flow (channel
         flow). These are useful for engineering applications,  although  not  for the  aerody-
         namics  of  wings  and  bodies.  The  third  exact  solution  represents  the  flow  in  the
         vicinity of  a  stagnation  point.  This  is  important  for  calculating  the  flow  around
         wings and bodies. It also illustrates a common and, at first sight, puzzling feature.
         Namely, that  if  the  dimensionless Navier-Stokes  equations can  be  reduced to an
         ordinary differential equation, this is regarded as tantamount to an exact solution.
         This is because the essentials of the flow field can be represented in terms of one or
         two curves plotted  on a single graph. Also numerical solutions to ordinary differ-
         ential equations can be obtained to any desired accuracy.


         2.10.1  Couette flow - simple shear flow

         This is the simplest exact solution. It corresponds to the flow field created between
         two infinite, plane, parallel surfaces; the upper one moving tangentially at speed UT,
         the  lower  one  being  stationary  (see  Fig.  2.30).  Since the  flow is  steady  and  two-
         dimensional, derivatives with respect to z and t are zero, and w = 0. The streamlines
   107   108   109   110   111   112   113   114   115   116   117