CHAPTER 3

BOUNDARY LAYER

CHAPTER 3 : BOUNDARY LAYER

Introduction

Internal flow vs External flow

External flow : object is completely surrounding by fluid flow.

Examples
- flow of air around airplanes.
- falling snow flakes.
- flow of water around submarine and fish

Important of external flow study : aerodynamics
- lift and drag on surface vehicles
- decrease vehicles fuel consumption
- improve vehicle handling

Two approaches (theoretical and experimental) used to obtain

information about fluid forces developed by external flow.

Because of the complexities of the governing equations and the
complexities of the geometry of the objectinvolved,theamountof
information obtained from theoretical method is limited. Much
information about external flow comes from experiments carried
out, for most part, on scale models of the actual objects.

In this chapter, we consider characteristics of external flow past an
objects and learn how to determine the various forces on objects
surrounded by a moving liquid.

General External Flow Considerations

A body immersed in a moving fluid experiences a resultant force
due to the interaction between the body and the fluid surrounding
it.

Fluid mechanics phenomena
- body moves through the fluid with velocity U.
- fluid flows past the stationary body with velocity U.

In any case we can treat the situation as fluid flowing past a

stationary body with velocity U, the upstream velocity.

Assume : the upstream velocity is constant in both time and
location (uniform and constant velocity).

The structure of an external flow and the ease with which the flow
can be described and analyzed often depend on the nature of the
body in the flow. Three general categories of bodies are
- two dimensional object
- axisymmetric bodies
- three dimensional bodies

Another classification of body shape

- streamlined body
- blunt body

When an object is immersed in a free stream of fluid flow, an

interaction between the body and the fluid occur.

 This can be described in terms of the stresses wall shear stress
due to viscous effects and normal stress due to the pressure.

Typical shear stressanddistributionsareshowninfigurebelowand

both shear stress and pressure vary in magnitude and direction
along the surface.

It is often useful to know the detailed distribution of shear stress

and pressure over the surface of the body although such
information is difficult to obtain. However only the integrated or
resultant effects of these distribution is needed.

Inaflowfield,thedragforceresultinginflowpassesoveranobject
is
D = DP +
DF
where
Dp - drag force resulting from pressure
Df - drag force resulting from friction

The resultant of the shear stress and pressure distribution can be

obtained by integrating the effect of these two quantities on the
body surface as indicated in figure below.


The components of the fluid force on the small area element dA

are
dFx = (pdA) cos + (wdA)
sin
dFy =
-(pdA) sin + (wdA)
cos

The net components of the force on the object are

Drag, D = dFx= p cos dA + w sin
dA
Lift, L = dFy = - p sin dA + w cos
dA

Dragistheresistingforceactingagainstthebodyinthedirectionof
flow.

Lift is the force acting in the vertical direction.

Obtaining these values of drag and lift for complex shapes is
difficult because we have to know as a function of locationalong
the body and the distribution of w and
p along the surface.

In this chapter, we will focus on a flat surface that is parallel tothe
direction of flow.

Characteristic of Flow Past an Object

Reynolds Number represents the ratio of inertial effect to viscous

effects.

In the absenceofallviscouseffect,theReynoldsNumberisinfinite
while in the absence of all inertial effects, the Reynolds Number is
zero.

Most external flows are associated with
- moderately sized objects (0.01m<l <10m)
- typical upstream velocities (0.01m/s<U <100m/s)
- fluid involved are typically water or air

The resulting Reynolds Number range for such flows is

approximately 10<Re<109. Flows with Re>100 are dominated by
inertial effects and flows with Re<1 are dominated by viscous
effects. Hence most familiar external flows are dominated by
inertial.

Flows past three flat plates of length l with Reynolds Number 0.1,
10, and 107 are shown in figure below.

 If the Reynolds Number is small, the viscous
effects are relatively strong and the plate affects
the uniform upstream flow far ahead, above,
below and behind the plate.

As the Reynolds Number increased, the region
in which viscous effects are important becomes
smaller in all direction except downstream. The
streamlined are displaced from their original
uniform upstream condition but the
displacement is not as great as for Re = 0.1
situation.

If the Reynolds Number is large, the flow is
dominated by inertial effects and the viscous
effects are neglected except in the region very
close to the plate and in the thin wake region
behind the plate

The streamlined of the flow outside the boundary layer are nearly
parallel to the plate.

The slide displacement of the external streamlines that are outside
of the boundary layer isduetothethickeningoftheboundarylayer
in the direction of the flow.

Flow past a blunt object also varies with Reynolds Number. In

general, the larger the Reynolds Number, the smaller the region of
the flow field in which viscous effects are important. For an object
that are not sufficiently streamlined, an additional characteristic of
the flow is observed and call flow separation and is illustrated in
figure below.

For Re = 0.1, the viscous effects areimportantseveraldiametersin
any direction from the cylinder. A somewhat surprising
characteristic of this flow is that the streamlines are essentially
symmetric aboutthecenterofthecylinder(thestreamlinepatternis
the same in front of the cylinder as it is behind the cylinder.

As the Reynolds Number is increased, the region ahead of the
cylinder inwhichviscouseffectsareimportantbecomesmallerwith
the viscous region extending only a short distance ahead of the
cylinder. With the increased of Reynolds Number, the fluid inertia
becomes more important and at some location on the body,
denoted the separation location, the fluid inertia cannot follow the
curved path around to the rear of the body. The result is a
separation bubble behind the cylinder in which some of thefluidis
actually flowing upstream, against the direction of the upstream
slow.

At still larger Reynolds Number, the area affected by the viscous

forces is forced farther downstream until it involves only a thin
(<<D) boundary layer on the front portionofthecylinderandan
 irregular, unsteady wake region that extends far downstream of the
cylinder.

Boundary Layer Characteristics

A boundary layer is a volume of fluidthatissloweddowncompare

to the free stream.

In the boundary layer, the local velocity (u) is less that the free
stream velocity (U), i.e u < U.

The boundary layer thickness is shown here as .

We assume here that the formation of the boundary layer takes
place at the front end of the flat plate.

The boundary layer forms at the top and bottom of the plate
surface.

The boundary layer can be divided into laminar boundarylayerand
turbulent boundary layer.

The Reynolds number for a boundary layer is given by
Rex = Ux/

 The transition from laminar boundary layer to turbulent boundary
layer takes place at a critical distance x = xcrit in the x-direction (2 x
105<Rex<3 x 106).

Turbulent boundary layer produces larger drag on an object

because the viscous effect is higher.

The sub-layer (similar to the one we discussed in chapter 2) is
located under the turbulent boundary layer on the surface.

This sub-layer will be destroyed when the surface roughness of the
plate is high.

Consider 2-dimensional flat plate on an x-y plane.

 The shear stress within fluid is given by

= (u/y)

where
= shear stress
= dynamic viscosity
(u/y) = velocity gradient across the boundary layer in
y-direction

The purpose of the boundary layer on plate is to allow the fluid to
change its velocity from upstream value of U to zero on the plate.
Thus u=0aty=0andu=Uaty=,withthevelocityprofile,u=
u(x,y) bridging the boundary layer thickness. In actuality there is no
sharp edge to the boundary layer (it is not precisely u = 0 at y = ).

Several parameters has been defined to develop the theory of
boundary layers namely
- boundary layer thickness,
- displacement thickness *
- momentum thickness

Boundary Layer Thickness,

Defined as the distance perpendicular from the surface of the flat

plate to the position where the local velocity is 99% of the free
stream velocity
= y when u = 0.99U

Displacement Thickness *

Defined as the distance for a deflected streamline due to the

boundary layer. Mathematically it is expressed by

= [1 (u/U)]dy

The displacement thickness represents the amount that the
thickness of the body must be increase so thattheuniforminviscid
flow has the same mass flowrate properties as the actual viscous
flow.

Figure below shows two velocity profiles for flow past a flat plate,
one if there were no viscosity and the other ifthereisviscosityand
zero slip at the wall. Because of the velocity deficit, U u with in
the boundary layer, the flowrate across section b-b is less than that
acrosssectiona-a.Howeverifwedisplacetheplateatsectiona-aby
an appropriate amount * (boundary layer displacement thickness),
the flowrates across each section will be identical. This is true if

*bU = (U u)bdy

* = [1 (u/U)]dy



Example : Air flow into a 2 ft square duct with uniform velocity of

10 ft/s forms a boundary layer on the walls. The fluid
within the core region flows as if it were inviscid. From
advanced calculation it is determined that for this flow
the boundary layer displacement thickness is given by
* = 0.007(x)1/2
Determine the velocity U = U(x,y) and pressure drop of
the air within the duct but outside of the boundary
layer.

Solution :

If we assumed incompressible flow, it follows that the volume
flowrate across any section of the duct is equal to that at the
entrance. That is

U1A
1 = (10)(2)(2) = 40 ft /s = 2uda
3

According to the definition of displacement thickness, the flowrate

across section 2 is the same asthatforauniformflowwithvelocity
U through a duct whose walls have been moved inward by *.
Therefore

 40 = 2uda
= U(2 - 2*)2

40 = 4U(1 - *)2

40 = 4U[1 - 0.007(x)1/2
]2

U = 10/[1 - 0.007(x)1/2
]2 ft/s

From Bernoulli equation

p1 + (1/2)U12 = p2 + (1/2)U22

with = 2.34 x 10-3 slug/ft3 and p1 = 0

p2 = (1/2) ( U12 U22 )
= (1/2)( 2.34 x 10-3 )[(10)2 (10/[1 - 0.007(x)1/2
]2 )2 ]
= 0.119 [1 - (1/[1 - 0.007(x)1/2 ]4 ) lb/ft2

Momentum Thickness,

Defined as the loss of the momentum capacity rate compared to

the rate without the boundary layer. Mathematically it can be
expressed by

= (u/U) [1 (u/U)]dy
Momentum thickness is often used when determining the drag on
an object.

Boundary Layer Equation and Solution

Governing equation for boundary layer flow - Prandtl Momentum
Equation (2 Dimensional)


Governing equation for Prandtl Momentum Equation
- Continuity equation

- Momentum equation

Solution for Prandtl Momentum Equation - Blasius Solution

- this solution is for laminar flow
- introduced the dimensional similarity variable
= (U/x)1/2

- from stream function equation, = ( x U)1/2 , the velocity
components are
u = Uf ()
v = (U/4x)1/2 (f - f )

- substitute velocity components into continuity and
momentum equation, the third order ordinary differential
equation
2f + ff = 0

- boundary condition
f = f = 0 at = 0 and f 1 as

 - numerical solution termed as Blasius solution is tabulatedin
table below

- from the solution it is found that u/U 0.99 when
= 5,
thus
i. Wall shear stress

ii. Boundary layer thickness



iii. Displacement thickness

iv. Momentum thickness

v. Friction coefficient, cf
cf = tw/0.5
U2

where
cf = 0.664(Rex) -1/2

vi. Friction drag coefficient, CDf as

CDf = Df/(1/2)
U2 bl

where
CDf = 1.328(Rel) -1/2

*** where
Reynolds number, Re = Ux/

**** page 17 31 shows the derivation of Prandtl Momentum Equation

and Blasius Solution (shear stress, boundary layer,
displacement thickness, momentum thickness equation)








 Themomentumintegralequation,w=U2 d/dxcanbeusedalong
with an assumed velocity profile to obtain reasonable, approximate
boundary layer results. The accuracy of these results depends on
how closely the shape of the assumedvelocityprofileapproximates
the actual profile.

Most simple approximation
(u/U) = y/ for 0 y


Other approximations are

Several assumed velocity profiles and the resulting values of are

given in figure and table below.

Profile Character Rex1/2
/x cfRe
x
1/2
CDfRe
l
1/2

Blasius solution 5.00 0.664 1.328

Linear
3.46 0.578 1.156
u/U = y/
Parabolic
5.48 0.730 1.460
u/U = 2y/ - (y/)2
Cubic
4.64 0.646 1.292
u/U = 3(y/)/2 (y/)3 /2
Sine wave
4.79 0.655 1.310
u/U = sin[(y/)/2]

It is often convenient to use the dimensionless local friction
coefficient, cf to express the wall shear stress
cf = tw/0.5
U2

The approximate value of local friction coefficient, cf using Blasius
solution is
cf = 0.664/(Rex) 1/2

For a flat plate of length, l and width, b the net frictiondragDfcan

be expressed in terms of the friction drag coefficient, CDf as
CDf = Df/(1/2)
U2 bl

The approximate value of friction drag coefficient, CDf using
Blasius solution is
CDf = 1.328/(Rel) 1/2


Example : A flat plate with a length of 2.4 m and width 0.9 m
moves horizontally with velocity 6 m/sinstillairwitha
density 1.21 kg/m3 and kinematic viscosity of 1.49 x
10-5m2/s. Assuming that the boundary layer is laminar
determine , *, , , cf , CDf
and Df .




Solution :

Laminar flow, Re = 2 x 105
Re = Ux/
2 x 10 =6(x)/(1.49 x 10-5 )
5

x= (2 x 105 )(1.49 x 10-5 )/6

= 0.497 m

t (/x)1/2
= 0.332U3/2
= 0.332(6 )[(1.21)(1.49 x 10-5 )]1/2
3/2

= 0.021Pa

= 5(x)/(Re)1/2
= 5(0.497)/(2 x 105 )1/2

= 5.56 x 10-3 m
= 5.56 mm

* = 1.72(x)/(Re)1/2
= 1.72(0.497)/(2 x 105 )1/2

= 1.91 x 10 m
-3

= 1.91 mm

= 0.664(x)/(Re)1/2
= 0.664(0.497)/(2 x 105 )1/2

= 7.38 x 10-4

cf = 0.664/(Rex) 1/2

= 0.664/(2 x 105 )1/2

= 1.48 x 10
-3

CDf = 1.328/(Rel)1/2
= 1.328/(2 x 105 )1/2

= 2.97 x 10-3

Df = CDf (1/2)U2 bl
= 2.97 x 10-3 (0.5)(1.21)(62 )(0.9)(2.4)
= 0.155 N
**** page 41 45 shows the derivation of Friction Coefficient and
Friction Drag Coefficient




Transition from Laminar to Turbulent Flow

When the flat plate is long enough there will be a transition from
laminar to turbulent boundary layer

This transition depends on Rex.

Transition takes place at Rex = Uxcr/

where x is the critical
distance along the flat plate.

The transition normally occurs at Rex = 5 x 10 although the value
5

might vary because of the instability of the process

Example : A fluid flows steadily past a flat plate with velocity U =

3.048 m/s. At approximately what location will the
boundary layer become turbulent, and how thick is the
boundary layer at that point if the fluid is a
a. water 15.6 oC
b. standard air
c. glycerin at 20 C
o



Turbulent Boundary Layer

Like turbulent flow in pipes, turbulent flow in boundary layer is
complex.

There are no exact solutions for a turbulent boundary layer.

To describe the turbulent boundary layer, we use the power law

u/U = (y/) 1/n

Momentum integral equation can be used to obtain results for

turbulent boundary layer.

From momentum integral solution, if Re < 107 , n = 7, thus

i. Wall shear stress, w
w = 0.0293U2 /(Rex) 1/5

ii Boundary layer thickness,

/x = 0.378/(Rex) 1/5

iii. Displacement thickness, *

*/x = 0.0473/(Rex) 1/5

iv. Momentum thickness,

/x = 0.0368/(Rex )1/5

v. Drag Force, Df
Df = 0.0366U2 A/(Rex ) 1/5

vi. Friction drag coefficient, CDf

CDf = Df/(1/2)
U2 A = 0.0732/(Rex ) 1/5

Different Re and n will give different solution/equation.

**** page 50 56 shows the derivation of an equation related to
turbulent boundary layer.





Comparison between Laminar and Turbulent Boundary Layer

Laminar Flow Turbulent Flow

(Blasius Solution)
Wall shear stress Wall shear stress
w = 0.0293U2 /(Rex) 1/5

Boundary layer thickness

Boundary layer thickness
/x = 0.378/(Rex) 1/5


Displacement thickness
Displacement thickness */x = 0.0473/(Rex) 1/5



Momentum thickness Momentum thickness
/x = 0.0368/(Rex) 1/5


Drag Force
Friction coefficient Df = 0.0366U2 A/(Rex ) 1/5
cf = tw/0.5
U
2

where
cf = 0.664(Rex) -1/2

Friction drag coefficient
Friction drag coefficient CDf = Df/(1/2)
U2 A
CDf = Df/(1/2)
U2 bl = 0.0732/(Rex )1/5

where
CDf = 1.328(Rel ) -1/2





Effect of Pressure Gradient

Previously we discussed the boundary layer on flat surface.

We assumed no pressure drop along the plate.

For curved surfaces the effect of pressure gradient is more

profound.

Consider flow over a curved surface shown in figure above.

At point A, velocity normal to surface is zero.

This is the stagnation point. Boundary layer grows from A.

Boundary layer keeps growing at B and C.

Over the rear section of the surface, the flow decelerates,therefore

the pressure increases.

The fluid particles in the boundary layer are slowed down due to
pressure gradient. (Adverse pressure gradient dp/dx > 0)

At D, dp/dx = 0

To overcome this effect, energy of fluid diffuses from free stream

to the surface of the body.
But if the pressure gradient is too large, the diffusion isinsufficient
to sustain a forward movement and the flow along the surface
reverses.

This forces the mainstream to separate givingthevelocityprofileat

E and F.

The region where flow is reversed is called the separation region.

Drag Forces over Various Types of Bodies

External flow over a body can be divided into 3 types

- free surface flow (e.g. ships)
- immersed flow (e.g. airplane, cars, submarines etc)
- compressible flow (e.g. high speed objects)

We will focus on flow over immersed bodies.

Objects fully immersed in fluid can be classified as :-
- Blunt objects
- Streamlined objects

Blunt objects
- flow is characterized by separation
- normally associated with high drag

Streamlined objects
- flow does not separate and follow the contour of the body.

 Drag is defined as a resistance force acting on the body in the
direction of flow.

Drag force is given as

D = U2A.CD
where
= fluid density
U = free stream velocity
A = area
CD = drag coefficient

CD depends
on the shape of the body and Reynolds Number.

For blunt objects, sometimes CD becomes independent of Re as

shown for spheres in figure below.



Drag can be divided into 2 types
- Friction drag discussed earlier
- Pressure drag due to difference in pressure over the body

Pressure Drag, Dp

where

Coefficient of Pressure, CP
- Defined as the ratio of static pressure to the dynamic force.

where
 P = reference pressure
P = static pressure
U = dynamic pressure
2

- Substituting in equation above

In many practical cases, we often look at the total drag force onan
object rather than in fraction (i.e friction and pressure drag).

One of the application is drag on airfoils.

The drag force acting on the airfoil is as follows
 FD = U2 A.CD
where
= fluid density
U = free stream velocity
A = planform area
CD = drag coefficient

Here A = Cl .For other objects A might be referred to as the
projected frontal area.

The drag coefficient for 2 dimensional objects

 The drag coefficient for 3 dimensional objects


The drag coefficient for objects of interest

Lift Force

The other force acting on the airfoil is the lift force

Lift acts in the direction perpendicular to the flow.

where CL is
the lift coefficient

Lift force results from the pressure difference over the top and
lower surface of the airfoil. (or any other shapes considered).

We discussed earlier how separation occurs due to the adverse
pressure gradient.

On airfoils, when (angle of attack) is increased, the separation
point moves from the rear end towards the leading edge figure
above causing wake at the rear.

When the separation point is almost at the leading edge, the entire
upper region is in the wake. (entire upper region in separation
flow).

 At this point, the pressure on the top of the airfoil is equal to
pressure at the bottom side.

Therefore, no more lift force is produced.

This condition is called stall and the angle of attack where stall
occurs is the stall angle

Lift and drag coefficient for wings/airfoils are functions of the

angle of attack.

Forces on Ground Vehicles

Thedragandliftforcesactingongroundvehiclesplayanimportant
role in terms of fuel consumption and stability.

Because ground vehicles travel in air (wind), we often call the drag
and lift as aerodynamic forces.

Vehicles are aerodynamically developed using wind tunnels.

Higher drag means higher resistance is sustained by the vehicle.To
overcome this resistance the vehicle draws power from the engine.
(Power = Force * Velocity)

Since the power to the engine is provided by fuel, then higherdrag
results in higher fuel consumption.

Lift is important when we consider the stability of the vehicle.
Lower lift is requiredsothatthevehiclestaysonthegroundathigh
speeds or when cornering. Too much lift however will make the
vehicle heavier, thus increasing fuel consumption.