2.

INCOMPRESSIBLE FLOW OVER A FLATPLATE:

THE BLASIUS SOLUTION

Consider the incompressible, two-dimensional flow over a flat plate at 0◦ angle of attack, such as sketched in
Figure 17.7. For such a flow, ρ = constant, μ = constant, and dp e/d x = 0 (because the inviscid flow over a flat
plate at α = 0 yields a constant pressure over the surface). Moreover, recall that the energy

equation is not needed to calculate the velocity field for an incompressible flow. Hence, the boundary-layer
equations, Equations (17.28) to (17.31), reduce to
∂u ∂v
+ =0 (18.1)
∂x ∂y
∂u ∂u ∂ 2u
u +v =ν 2 (18.2)
∂x ∂y ∂y
∂p
=0 (18.3)
∂y
where ν is the kinematic viscosity, defined as ν ≡ μ/ρ.
We now embark on a procedure that is common to many boundary-layer solutions. Let us transform the
independent variables (x, y) to (ξ, η), where

V∞
ξ = x and η = y (18.4)
νx
Using the chain rule, we obtain the derivatives
∂ ∂ ∂ξ ∂ ∂η
= + (18.5)
∂x ∂ξ ∂ x ∂η ∂ x
∂ ∂ ∂ξ ∂ ∂η
= + (18.6)
∂y ∂ξ ∂ y ∂η ∂ y
However, from Equations (18.4) we have

∂ξ ∂ξ ∂η V∞
=1 =0 = (18.7)
∂x ∂y ∂y νx
(We do not have to explicitly obtain ∂η/∂ x because these terms will eventually
cancel from our equations.) Substituting Equations (18.7) into (18.5) and (18.6),
we have
∂ ∂ ∂η ∂
= + (18.8)
∂x ∂ξ ∂ x ∂η

∂ V∞ ∂
= (18.9)
∂y νx ∂η
∂2 V∞ ∂ 2
= (18.10)
∂ y2 νx ∂η2
Also, let us define a stream function ψ such that

ψ = νx V∞ f (η) (18.11)
where f (η) is strictly a function of η only. This expression for ψ identically satisfies the continuity equation,
Equation (18.1); therefore, it is a physically possible stream function. [Show yourself that ψ satisfies Equation
(18.1); to do this, you will have to carry out many of the same manipulations described below.]
From the definition of the stream function, and using Equations (18.8), (18.9),
and (18.11), we have 
∂ψ V∞ ∂ψ
u= = = V∞ f  (η) (18.12)
∂y νx ∂η
  
∂ψ ∂ψ ∂η ∂ψ 1 νV∞  ∂η 
v=− =− + =− f − νx V∞ f (18.13)
∂x ∂ξ ∂ x ∂η 2 x ∂x
Equation (18.12) is of particular note. The function f (η) defined in Equa-
tion (18.11) has the property that its derivative f  gives the x component of
velocity as
u
f  (η) =
V∞
Substitute Equations (18.8) to (18.10), (18.12), and (18.13) into the momentum
equation, Equation (18.2). Writing each term explicitly so that you can see what
is happening, we have
⎛  ⎞ 
  
∂η 1 νV ∞ ∂η V∞  V∞ 
V∞ f  V∞ f  −⎝ f + νx V∞ f  ⎠V∞ f = νV∞ f
∂x 2 x ∂x νx νx

Simplifying, we obtain
 
2 ∂η 1 V∞2
∂η V2
V∞ 
f f − 
f f − V∞ 2
f  f  = ∞ f  (18.14)
∂x 2 x ∂x x
The first and third terms cancel, and Equation (18.14) becomes

2 f  + f f  = 0 (18.15)

Equation (18.15) is important; it is called Blasius’ equation, after H. Blasius,

who obtained it in his Ph.D. dissertation in 1908. Blasius was a student of
Prandtl, and his flat-plate solution using Equation (18.15) was the first practical
application of Prandtl’s boundary-layer hypothesis since its announcement in
1904. Examine Equation (18.15) closely. Amazingly enough it is an ordinary
differential equa-tion. Look what has happened! Starting with the partial
differential equations for a flat-plate boundary layer given by Equations (18.1)
to (18.3), and transform-ing both the independent and dependent variables
through Equations (18.4) and (18.11), we obtain an ordinary differential
equation for f (η). In the same breath, we can say that Equation (18.15) is also
an equation for the velocity u because u = V∞ f (η). Because Equation (18.15)
is a single ordinary differential equa-tion, it is simpler to solve than the original
boundary-layer equations. However, it is still a nonlinear equation and must be
solved numerically, subject to the transformed boundary conditions,
At η = 0: f = 0, f  = 0
At η → ∞: f = 1
[Note that at the wall where η = 0, f  = 0 because u = 0, and therefore f = 0 from Equation
(18.13) evaluated at the wall.]
Equation (18.15) is a third-order, nonlinear, ordinary differential equation; it can be
solved numerically by means of standard techniques, such as the Runge-Kutta method
(such as that described in Reference 52). The integration begins at the wall and is carried
out in small increments y in the direction of increasing y away from the wall. However,
since Equation (18.15) is third order, three bound-ary conditions must be known at η = 0;
from the above, only two are specified. A third boundary condition, namely, some value
for f (0), must be assumed; Equation (18.15) is then integrated across the boundary layer
to a large value of η. The value of f  at large eta is then examined. Does it match the
boundary condition at the edge of the boundary layer, namely, is f  = 1 satisfied at the edge
of the boundary layer? If not, assume a different value of f (0) and integrate again. Repeat
this process until convergence is obtained. This numerical approach is called the “shooting
technique”; it is a classical approach, and its basic phi-losophy and details are discussed at
great length in Section 16.4. Its application to Equation (18.15) is more straightforward
than the discussion in Section 16.4, because here we are dealing with an incompressible
flow and only one equation, namely, the momentum equation as embodied in Equation
(18.15).

The solution of Equation (18.15) is plotted in Figure 18.2 in the form of f (η) = u/V∞
as a function of η. Note that this curve is the velocity profile and that it is a function of η
only. Think about this for a moment. Consider two different x stations along the plate, as
shown in Figure 18.3. In general, u = u(x, y), and the velocity profiles in terms of u = u(y)
at given x stations will be different. Clearly, the variation of u normal to the wall will
change as the flow progresses downstream. However, when plotted versus η, we see that
the profile, u = u(η), i s

Figure 18.2 Incompressible velocity profile for a flat plate; solution

of the Blasius equation.