CHAPTER 4: VISCOUS BOUNDARY LAYERS

Continuity equation;

(4.1)

x component of momentum
(4.2)

y component of momentum

(4.3)

If the boundary layer is thin and the streamlines are not highly curved, then we will assume that
u >> v, which will typically be true as long as we are not too close to the stagnation point.

we can conclude that: This is true whether the boundary layer1

2
BOUNDARY CONDITIONS

3
INCOMPRESSIBLE, LAMINAR BOUNDARY LAYER

4
Numerical Solutions for the Falkner-Skan Problem
There are no analytical solutions to this third-order ordinary differential equation, which is known
as the Falkner-Skan equation . Nevertheless, there are a variety of well-known numerical
techniques available to solve it.

5
For the accelerating flows (i.e., positive β), the velocity increases rapidly with distance from the
wall, so ∂u/ ∂y at the wall is relatively large. We would expect that the shear force at the wall
would be relatively large.

6
7
Although the transverse component of velocity at the wall is zero, it is not zero at the edge of the
boundary layer.

8
9
Another convenient formulation for skin-friction on a flat plate is found by integrating the “local”
skin-friction coefficient, C f , found in equation (4.21) to obtain a “total” or “average” skinfriction
drag coefficient on the flat plate.

10
EXAMPLE 4.1
A rectangular plate, whose streamwise dimension (or chord c ) is 0.2 m and whose width (or
span b) is 1.8 m, is mounted in a wind tunnel. The free-stream velocity is 40 m/s. The density of
the air is 1.2250 kg/m3, and the absolute viscosity is 1.7894x10- 5 kg/m.s. Graph the velocity
profiles at x = 0.0 m, x = 0.05 m, x = 0.10 m, and x = 0.20 m. Calculate the chordwise
distribution of the skin-friction coefficient and the displacement thickness. What is the drag
coefficient for the plate?

11
12
EXAMPLE 4.2
The streamwise velocity component for a laminar boundary layer is sometimes assumed to be
roughly approximated by the linear relation

where δ = 1.25 x 10-2 √x. Assume that we are trying to approximate the flow of air at standard
sea-level conditions past a flat plate where ue = 2.337 m/s. Calculate the streamwise distribution
of the displacement thickness (δ*), and the skin-friction coefficient (Cf). Compare the values
obtained assuming a linear velocity profile with the more exact solutions presented in this chapter.

13
14
BOUNDARY-LAYER TRANSITION
Factors effect transition;
1. Pressure gradient
2. Surface rougness
3. Compressibilty effects (Mach number)
4. Surface temperature
5. Suction or blowing at the surface
6. Freestream turbulence

15
Once the critical Reynolds number is exceeded, the flat-plate boundary layer would contain
regions with the following characteristics as it transitioned from the laminar state to a fully
turbulent flow:
●
Stable, laminar flow near the leading edge
●
Unstable flow containing two-dimensional Tollmien-Schlichting (T-S) waves
●
A region where three-dimensional unstable waves and hairpin eddies develop
●
A region where vortex breakdown produces locally high shear
●
Fluctuating, three-dimensional flow due to cascading vortex breakdown
●
A region where turbulent spots form
●
Fully turbulent flow

16
INCOMPRESSIBLE, TURBULENT BOUNDARY LAYER
A turbulent flow has several characteristics:
●
Chaotic and irregular—fundamentally unsteady and three dimensional
●
Enhanced mixing of momentum, heat, and mass
●
Large Reynolds numbers
●
Three-dimensional vorticity fluctuations

17
Derivation of the Momentum Equation for Turbulent Boundary Layer
Continuity equation;

Now we can perform the same decomposition for the momentum equation.

Now take the time average of the terms in this equation, remembering that terms which contain
only one fluctuating parameter vanish, since their time-averaged value is zero.

18
Now we can multiply the fluctuating portion of the continuity equation (4.45b) by

Taking the time average of these terms, we find that:

Adding equation (4.47) to (4.46) and rearranging the terms gives us:

We need to further examine the last two terms in equation (4.49):

19
Neumann (1989) notes, “Turbulence models employed in computational schemes to specify
the character of turbulent flows are just that . . . models, non-physical ways of describing the
character of the physical situation of turbulence. The models are the result of generalizing and
applying fundamental experimental observations; they are not governed by the physical
principles of turbulence and they are not unique.”
20
Turbulent Boundary Layer for a Flat Plate

21
EDDY VISCOSITY AND MIXING LENGTH CONCEPTS
In the late nineteenth century, Joseph Boussinesq introduced the concept of eddy viscosity to
model the Reynolds shear stress. Boussinesq assumed that the Reynolds stresses act like the
laminar shear stresses and are proportional to the transverse gradient of the mean streamwise
velocity component. The coefficient of proportionality is called the eddy viscosity and
the Reynolds stresses were assumed to take the form:

22
Inner region: Outer region: Between inner and outer layer

23
INTEGRAL EQUATIONS FOR A FLAT-PLATE
BOUNDARY LAYER

24
25
Application of the Integral Equations of Motion to a Turbulent, Flat-Plate Boundary Layer
Experimental measurements have shown that the time-averaged velocity for a turbulent boundary
layer on a flat plate may be represented by the 1/7th power law:

So, although the correlation given in equation (4.76) provides a reasonable representation of the
actual velocity profile, we need another piece of experimental data: a correlation for the shear
acting at the wall. Blasius found that the skin-friction coefficient for a turbulent boundary layer on a
flat plate where the local Reynolds number is in the range 5 x 105 to 1 x 107 is given by:

26
We can now find the boundary layer thickness by substituting equations (4.76) and (4.77) into
equation (4.78) to obtain:

which eventually becomes:

If we assume that the boundary-layer thickness is zero when x = 0, we find that

By rearranging, we can find the various thicknesses of a turbulent boundary layer on a flat plate
since Schlichting and Gersten (2000) showed that
and where

27
Substituting the boundary layer thickness from equation (4.79) into equation (4.77) yields:

As with the laminar skin-friction coefficient found in Sec. 4.3.1, a total skin-friction coefficient can
be found for turbulent flow by integrating equation (4.80) over the length of a flat plate:

28
The calculation of the skin-friction drag for a flat plate with transition theoretically would require
using the local skin-friction coefficients, equation (4.21) for the laminar portion of the flow and
equation (4.80) for the turbulent part of the flow. Taking care to only integrate each relation over
the laminar and turbulent lengths, respectively, gives us:

29
●
The value A = 1700 represents the laminar correction for a transition Reynolds number of
Rex,tr = 500,000.
●
You can see from this formulation that if the Reynolds number at the end of the plate is very high,
then the laminar correction term plays a fairly insignificant role in the total skin-friction drag on the
plate.
●
A good rule of thumb is to assume that if transition takes place at less than 10% of the length of
the plate, then the laminar correction usually can be ignored, since it is relatively small.
●
Fig. 4.19 shows how the total skin-friction coefficient varies from the laminar value in equation
(4.32), through transition, and finally to the fully turbulent value, in equation (4.82).

30
In reality, the flow is continuous at the transition location and the boundary-layer thickness
does not change instantaneously. Furthermore, since we are at the transition location, it is not
realistic to use the assumption that the boundary layer is turbulent all the way from the leading
edge. (This assumption would be reasonable far downstream of the transition location so that
x >> xtr .)
31
Example 4.4
Air at standard day sea-level atmospheric pressure and 5o C flows at 200 km/h across a flat plate.
Compare the velocity distribution for a laminar boundary layer and for a turbulent boundary layer
at the transition point, assuming that the transition process is completed instantaneously at that
location.

32