3 Blasius solution

The Blasius problem deals with flow in the boundary layer around a stationary plate. The setup is shown in figure 2. At a large distance the fluid has a uniform velocity U oo . It interacts with a plate whose edge is at x = 0 and which extends to the right from there. As before, we need to think about the physical situation that we expect to develop before tackling the mathematics. Thus, at x = 0 fluid right at the surface of the plate must be brought to rest (ux = 0). As before, the effect of the presence of the plate propagates outwards into the fluid, ie the boundary layer becomes broader. However at the same time the fluid is flowing downstream along the plate. So instead of the boundary layer getting thicker as time progresses, it instead gets thicker as you move further along the plate.


PIC
Figure 2: Boundary layer around a stationary plate - Blasius problem

Now fairly obviously the presence of this boundary layer will disrupt the uniform flow of the fluid, and so in the boundary layer the flow will have components ux and uy. We will assume that the growth of the boundary layer is quite slow, i.e. its thickness is small compared to x at any point x down the plate. In this case uy < ux, and the Navier-Stokes equations simplify to

ux@ux-+ uy@ux- =   -@p-+ n@2ux-                        (3)
  @x      @y        @x     @y2
    @ux-  @uy-
    @x  + @y   =   0                                   (4)
In addition we will consider the case for which there is no pressure gradient along the plate in the direction of the flow, i.e.
@p
@x-= 0
This will be the case for a flat plate : curved surfaces may differ however, a feature which will be considered later in the course. At any point in the boundary layer the flow velocity at y may depend on U oo [LT-1], kinematic viscosity n[L2T-1] and distance from the leading edge of the plate x[L]. From these the we can form the quantity
  ----
 V~  nx-
   U oo    which has dimensions [L]
and so we are going to want to write the solution as a function of the single variable
z =  V~ -y-----
      nx/U oo

In fact it is best to write ux and uy in terms of a stream function

     @Y           @Y
ux = @y-   uy = - @x-
In turn, the stream function
Y =  V~ nxU--f(z)
        o o
The stream function will automatically satisfy equation 4 whilst equation 3
  @ux     @ux     @2ux
ux-@x-+ uy-@y-= n @y2--
reduces to an O.D.E for f
 d2f    d3f
fdz2 + 2dz3 = 0
(5)
We note in passing
 ux            uy    1 V~ --n--
U-- = f'(z),    U-- = 2  U--x-(zf'- f)
  oo               oo         oo

Thus, if we can solve equation 5 to find f(z), we can find the flow throughout the boundary layer. In particular, the friction coefficient

       t0
Cf = 1---2-
     2rU oo
where t0 is the surface stress (force per unit area of the surface). But
      (@u  )
t0 = m  --x-      is the surface stress
        @y  y=0
We can evaluate (@ux)
  @y y=0 from our solution as,
(@u  )        V ~  U--(d2f )
  --x-    = U oo   -o o  --2
   @y  y=0        nx  dz   z=0
which is how the equation
Cf = 0 V~ .664-
       Rex
is derived.