Chapter 2

Boundary-Layer Approximations for Laminar Flow

2.1 Introduction
Consider the flow of a viscous fluid over a stationary smooth plate surface as
illustrated in Fig 2.1. As we have already discussed, at the surface the fluid
particles adhere to it and the frictional forces between the fluid layers retard
the motion of the fluid within a thin layer near the surface. In this thin layer,
which is termed the boundary layer, the velocity of the fluid decreases form
its free-stream value U

to a value of zero at the surface (no-slip condition).

Fig 2.2 shows diagrammatically the variation of the velocity component
parallel to the surface in the boundary layer at a given location on the
surface.
) (x
U, T

x
y
U
T
w
T

Fig 3.1 Flow of a viscous fluid

parallel to a flat surface.

y
0.99
u/U
1.0
0
Fig. 3.2 Velocity boundary layer
thickness
According to Prandtl's boundary-layer concept, under certain conditions
viscous forces are of importance only in the immediate vicinity of a solid
surface where velocity gradients are large. This region near the surface is
referred to as the boundary layer. In regions removed from the solid surface
where there exist no large gradients in fluid velocity, the fluid motion may
be consider frictionless i.e. potential flow. In fact, there is no precise
dividing line between the potential flow region, where friction is negligible,
and the boundary region, because the velocity component parallel to the
surface, u, approaches the free-steam value U

asymptotically. However, it
is customary to define the boundary layer as that region where the velocity
component parallel to the surface is less than 99% of the free-steam velocity.
Since the velocity component parallel to the surface, u, is varying, the
continuity requires that there should also be velocity component v normal to
the surface. So the originally parallel flow becomes, at least, a two-
dimensional flow in the boundary layer.
Initially the boundary layer development is laminar, but at some critical
distance from the leading edge small disturbance in the flow begin to be
amplified and a transition process takes place until the flow in the boundary
layer becomes turbulent as illustrated in Fig. 2.3. Experiments have shown
that the transition will normally occur at a distance x
cr
measured from the
leading edge, where the value of a local Reynolds number based on x
cr
i.e.,
Re
cr
= U

x
cr
/ v has a value approximately equal to 510
5
. Depending on the
flow conditions, fluid properties and surface conditions the transition to
turbulent flow may be start at a Reynolds number as small as 10
5
or the
laminar boundary layer may exist up to a Reynolds number 310
6
.
The structure of turbulence is a complex phenomenon, which is still not fully
understood and the theory that has been developed is so limited that semi-
empirical methods are invariably used for the solution of energy equations
for laminar boundary layer and obtain the laminar boundary-layer
equations.
T
r
a
n
s
i
t
i
o
n

r
e
g
i
o
n
Laminar
boundary layer
Turbulent
boundary layer
x
a
x
y
U
Fig 3.3 Laminar and turbulent boundary layers on a flat plate
2.2 Momentum Equations of the Boundary Layer
Consider now the case of steady two-dimensional incompressible laminar
boundary layer over a surface with a free steam velocity U

(x), which may

vary with x in an arbitrary manner. Assume that the body forces are
negligible and the viscosity of the fluid is constant.
The continuity equation (1.13) and the Navier-Stokes equations (1.33) for
this case reduce to
0 =
c
c
+
c
c
y
v
x
u (2.1)
|
|
.
|

\
|
c
c
+
c
c
+
c
c
=
c
c
+
c
c
2
2
2
2
1
y
u
x
u
x
p
y
u
v
x
u
u

(2.2)
and
|
|
.
|

\
|
c
c
+
c
c
+
c
c
=
c
c
+
c
c
2
2
2
2
1
y
v
x
v
y
p
y
v
v
x
v
u

(2.3)
These equations may now be written in the following dimensionless forms:
0 =
c
c
+
c
c
y
v
x
u

(2.4)
|
|
.
|

\
|
c
c
+
c
c
+
c
c
=
c
c
+
c
c
2
2
2
2
Re
1
y
u
x
u
x
p
y
u
v
x
u
u
(2.5)
and
|
|
.
|

\
|
c
c
+
c
c
+
c
c
=
c
c
+
c
c
2
2
2
2
Re
1
y
v
x
v
y
p
y
v
v
x
v
u
(2.6)
where

L U
L
y
y
L
x
x
0
Re , , = = =
2
0
0 0
, ,
U
p
p
U
v
v
U
u
u

= = =
Here U
0
= U

(0) is taken to be reference velocity and L is the length of the

surface in x-direction.
We now wish to simplify Eqs. (2.4), (2.5) and (2.6) by utilizing the
experimentally observed fact that 1 << = L . We do this by subjecting
these equations to an order-of-magnitude analysis with the purpose of
dropping those terms which we feel are negligibly small in the computations
as compared o the other terms in the same equations.
In most of the boundary layer ) 1 ( O u = and ) ( O y = , and in locations
removed from the leading edge ) 1 ( O x = , where the symbol O( ) stands for
the "order of.". Considering the order of magnitudes of u and x we observe
that
) 1 (
) 1 (
) 1 (
O
O
O
x
u
= ~
c
c

and, therefore, it follows that

) 1 (
2
2
O
x
u
=
c
c

Similarly, since ) ( O y = , we also have

, , 1 O
y
u
=
c
c
and , ,
2
2
2
1 O
y
u
=
c
c
On the other hand, from the continuity equation (2.4) we know that x u c c
and y v c c are to be of the same order-of-magnitude. Thus,
, , 1 O
y
v
=
c
c
and, therefore, ) ( O v = . Hence, we also conclude that
, , 1 ), (
2
2
O
y
v
O
x
v
=
c
c
=
c
c

and , , O
x
v
=
c
c
2
2
We now rewrite Eqs. (2.4), (2.5) and (2.6) and indicate under each term its
order-of-magnitude:

1
1
0 =
c
c
+
c
c
y
v
x
u

(2.7)
|
.
|

\
|
+ +
|
|
.
|

\
|
c
c
+
c
c
+
c
c
=
c
c
+
c
c
2
2
2
2
2
1
1
1 1
1
1
Re
1

y
u
x
u
x
p
y
u
v
x
u
u
(2.8)
and
|
|
.
|

\
|
+ +
|
|
.
|

\
|
c
c
+
c
c
+
c
c
=
c
c
+
c
c
2
2
2
2
2
1 1
1
Re
1

y
v
x
v
y
p
y
v
v
x
v
u
(2.9)
Since 1 << , from Eq. (2.8) we conclude that
2
2
2
2
y
u
x
u
c
c
<<
c
c

and therefore the first term in the parenthesis on the right-hand side can be
neglected compared to the second term. Note that the order of magnitude of
left-hand side of Eq. (2.8) is unity. In order for the viscous effects to be of
the same order of magnitude as the inertia effects on the left-hand side of
this equation, the Reynolds number should be of order
2
1 :
, ,
2
1 Re O =
which means that the Reynolds number must be very large for the boundary
layer approximation to be applicable. Inserting this order of magnitude into
Eq. (2.9), we see that all terms are of order of magnitude and consequently
conclude that
, , O
y
p
=
c
c
and, therefore, the change in pressure across the boundary layer is of the
order of
2
which is very small. This implies that at any x the pressure is
practically constant in the y-direction within the boundary layer, but can
vary in the x-direction. Thus, the pressure within the boundary layer may be
considered to be determined by the outer potential flow.
As a result of the forgoing order of magnitude analysis, the momentum
equations (2.2) and (2.3) reduce to
|
|
.
|

\
|
c
c
+
c
c
=
c
c
+
c
c
2
2
1
y
u
x
p
y
u
v
x
u
u

(2.10)
0 =
c
c
y
p
(2.11)
which shows that pressure can be assumed as constant across the boundary
layer. Then Eqs. (2.10) and (2.11) can be combined into one as
|
|
.
|

\
|
c
c
+
c
c
=
c
c
+
c
c
2
2
1
y
u
x
p
y
u
v
x
u
u

(2.12)
Equation (2.12) in the limit as y (outside the edge of the velocity
boundary layer) reduces to
x
p
x
U
U
c
c
=
c
c

1
(2.13)
in the outer flow region. In the boundary layer analysis, the free-steam
velocity U

(x) is assumed to be available from the solution of the potential

flow outside the boundary layer, and thus the pressure gradient term dp/dx is
considered to be known from Eq. (2.13).
Hence, Eqs. (2.1), (2.2) and (2.3) may now be replaced by the following
Prandtl's boundary-layer equations:
0 =
c
c
+
c
c
y
v
x
u
(2.14)
|
|
.
|

\
|
c
c
+
c
c
=
c
c
+
c
c
2
2
1
y
u
x
p
y
u
v
x
u
u

(2.15)
with the following boundary conditions:
0 : 0 = = = v u y at
(2.16a,b)
) ( : x U u y as

=
(2.16c)
In addition, the velocity distribution at x = 0 must also be specified.
It may be noted that, although one of the viscous terms in Eq. (2.8) has been
dropped, the order of this equation has not been reduced. Also, one of the
equations of motions has been dropped completely. As a result, the number
of unknowns has been reduced by one.
A similar analysis has been carried out for the boundary layer flow along a
curved wall and it has been concluded that the above equations may be
applied to a curved wall as long as no large variations in curvature occur.
It should be noted that the boundary layer approximations are valid for large
values of the Reynolds number and the no-slip condition is assumed on the
solid surface i.e., the fluid layer at y = 0 sticks to the solid surface.
2.3 Boundary-Layer Energy Equations
If the plate temperature is different from the fluid temperature a thermal
boundary layer also develop over the plate as illustrated in Fig. 2.1,
indicating a significant temperature variation over a narrow zone in the
immediate vicinity of the plate. The thermal boundary layer thickness
T
may be defined as that distance from the surface (T-T
w
) = 0.99(T

-T
w
) as
shown in Fig. 2.4.
y
0.99

T T
T T
w
1.0
0

T
Fig 2.4 Thermal boundary layer thickness
For steady, two-dimensional and incompressible viscous flow with constant
thermophysical properties, the energy equation (1.61) reduce to
1
1
]
1

|
|
.
|

\
|
c
c
+
c
c
+
|
|
.
|

\
|
c
c
+
|
.
|

\
|
c
c
+
|
|
.
|

\
|
c
c
+
c
c
=
|
|
.
|

\
|
c
c
+
c
c
2 2
2
2
2
2
2
2
1
2
y
u
x
v
y
v
x
u
y
T
x
T
y
T
v
x
T
u c
p

(2.17)
Introducing the dimensionless variables
w w
T T
T T
T
U
v
v
U
u
u
U
p
p
L
y
y
L
x
x

= = = = = =

; ; ; ; ;
0 0
2
0

Where U
0
= U

(0), the energy equation (2.17) may be written in the

following dimensionless form:
1
1
]
1

|
|
.
|

\
|
c
c
+
c
c
+
|
|
.
|

\
|
c
c
+
|
.
|

\
|
c
c
+
|
|
.
|

\
|
c
c
+
c
c
=
c
c
+
c
c
2 2
2
2
2
2
2
2
1
Re
2
Pr Re
1
y
u
x
v
y
v
x
u Ec
y
T
x
T
y
T
v
x
T
u
(2.18)
where we have introduced
, ,
w p
p
T T c
U
Ec
C
L U

= = =

2
0
, Pr , Re

Following the dimensions of the previous section, we now again conclude

that in the energy equation (2.18)
) 1 ( ), 1 ( ), 1 (
2
O
y
u
O
x
u
O u =
c
c
=
c
c
=
) 1 ( ), ( ), ( O
y
v
O
x
v
O v =
c
c
=
c
c
=
and , ,
2
1 Re O = , where, as before ) 1 ( O x = , ) ( O y = and L = .
Comparing the order of magnitudes of the dissipation terms, the energy
equation (2.18) can be rewritten as
2 2
2
2
2
2
2
1 1 1 1 1 1
Re Pr Re
1

T
x
u Ec
y
T
x
T
y
T
v
x
T
u
|
.
|

\
|
c
c
+
|
|
.
|

\
|
c
c
+
c
c
=
c
c
+
c
c
(2.19)
Furthermore, we note that in the thermal boundary layer ) 1 ( O T = because T
varies from zero at the surface of the plate to almost unity at y =
T
and also
) (
T
O y = , where L
T T
/ = . Hence, we observe that
) 1 ( and ) 1 (
2
2
O
x
T
O
x
T
=
c
c
=
c
c
) 1 ( and ) 1 (
2
2
2
T T
O
y
T
O
y
T
=
c
c
=
c
c
The order of magnitude of various terms in Eq. (2.19) is indicated under
each term. Comparing the orders-of-magnitude of the conduction terms we
conclude that
2
2
2
2
y
T
x
T
c
c
<<
c
c
It can also be seen from Eq. (2.19) that the conduction and convection terms
are to be of same order of magnitude if the produce RePr is of order
2
1
T
;
that is,
, ,
2
1 Pr Re
T
O =
Combining the previously obtained estimation for the order of magnitude of
Re number with the above result gives
Pr
1
~

T
(2.20)
In Eq. (2.19), the dissipation term will be important if
) (
Re
2
O
Ec
=
Or, since , ,
2
1 Re O =
, ,
) 1 (
2
0
O
T T c
U
Ec
w p
=

In view of the forgoing order of magnitude analysis, the boundary layer

energy equation for a steady, two-dimensional and incompressible laminar
flow with constant thermophysical properties becomes
2
2
2
2
|
|
.
|

\
|
c
c
+
c
c
=
c
c
+
c
c
y
u
c
y
T
c y
T
v
x
T
u
p p

(2.21)
with boundary conditions:
w w
q
y
T
or T T y at ' ' =
c
c
= = : 0
(2.22a,b)

= T T y as :
(2.22c)
As stated in Eqs. (2.22a,b), at the surface of the plate either the wall
temperature T
w
or the heat flow rate per unit area from the surface to the
fluid q
w
is specified.
In conclusion, for two-dimensional steady and incompressible laminar
boundary-layer flows with constant thermophysical properties, we have only
three equations, that is, Eqs. (2.14), (2.15) and (2.21), for the three
unknowns u, v and T.