The Navier-Stokes equations

Fundamental equations:

Constitutive equations:
 Equations of incompressible flow
The mass balance equation:

The momentum balance equation:

The internal energy balance equation:

External incompressible
viscous flow
 What is the difference between
internal and external viscous flows?
• Fully developed flow, in which the fluid does not accelerate,
can be obtained in internal flow. =0

- the material derivative

 What is the difference between
internal and external viscous flows?
• For external flow, the flow never becomes fully developed so that the
nonlinear convective acceleration term in the Navier-Stokes equations
cannot be neglected. The boundary layers continue to grow along the
streamwise direction.
≠0
 Definition of external flows
“Flows over bodies immersed in an unbounded fluid.”
 Objective

“To quantify the behaviour of viscous,

incompressible fluids in external flow.”
 A number of phenomena occur

• Stagnation point, No-slip condition, Separation point, Viscous wake

• Boundary layer – Laminar, Transitional, Turbulent

Viscous flow around an airfoil

A number of phenomena occur

Flow over a flat plate

 Outline
• Part A
The boundary-layer concept
Boundary layer equations
Boundary layer parameters
von-Kármán’s integral equation
Laminar boundary layer
- Blasius velocity profile
- Use of the von-Kármán’s integral equation
Turbulent boundary layer
Pressure gradients in boundary-layer flow
• Part B
Streamlined body aerodynamics
Bluff-body aerodynamics
Part A
 Equations of incompressible fluid flow
• Euler’s equation was first published by Leonhard Euler
in 1755. It was derived using a purely mathematical approach.

(The motion of non-viscous fluid)

Results have discrepancy with experimental observations.

• The complete equations, the Navier-Stokes equations,
describing the motion of a viscous fluid were later developed
by Navier in 1827 and independently by Stokes in 1845.

These equations are very difficult to solve due to nonlinearity.

 The boundary-layer concept
• The boundary-layer concept was introduced by Ludwig
Prandtl in 1904.
Irrotational flow

Rotational flow

• “It is possible that the effects of viscosity be ‘felt’ only

within a thin layer adjacent to the body surfaces (the grey
regions), where vorticity is presented, while in the outer
irrotational flow the fluid particles move as if they belonged
to a non-viscous fluid.”
Rotational and irrotational flows
Prandtl’s simplified mathematical model
• Inside the boundary layer – The boundary layer equations
are employed. Their derivation will be presented in the next
section.
• Outside the boundary layer– Without vorticity, the motion
of the fluid may be considered as irrotational. Potential flow
theory can be applied to obtain the velocity field.
, where a scalar function is called the velocity potential.

The non-penetration boundary condition, , is applied. is

the normal unit vector at the boundary. Then, the pressure field can
be computed using Bernoulli’s equation.
 Key points from Prandtl’s proposal
• Prandtl’s proposal marked a major breakthrough in the
historical development of fluid dynamics.

• Since then, predicting the dynamics of boundary layers has been a

fundamental step in understanding fluid flows and mastering their
behavior.

• It provided a theoretical basis to explain why a non-viscous

flow model may give useful results. Moreover, it permitted
procedures to be devised to estimate the fluid dynamic loads
on certain bodies. Further details will be given in Part B.

• It can be applied not only to the regions adjacent to solid

surfaces, but also to all zones in the flow where vorticity is
confined within thin layers such as jets, wakes, and mixing layers.
 Recapitulation of vorticity
• The curl of the velocity field

• The rotation of a fluid particle is defined by the rotation

vector, which is the average angular velocity of two
originally perpendicular lines attached to the fluid particle.
 Deformation of a fluid element
• The deformation of the elementary volume of fluid
is defined by the nine components of the tensor .
 Decomposition of
• By decomposing into the symmetric and anti-
symmetric parts,

Symmetric part –
Linear and angular
deformations

Anti-symmetric part
- Rotation
 The origin of vorticity

Impulsive start of an infinite flat plate in a reference system fixed to the

plate: (a) Non-viscous flow (b) Viscous flow.
Impulsive start of a flat plate through water
 The origin of vorticity
• In non-viscous flow, no rotation of any fluid element
adjacent to the plate will occur. Thus, vorticity will remain
zero everywhere.

• In viscous flow, the deformation of the fluid element

corresponds not only to a distortion but also to a rotation.
Therefore, the fluid element acquires vorticity due to the
viscous no-slip boundary condition, which prevents the
presence of a velocity difference between fluid particles
and the solid wall.
The applicability of Bernoulli’s theorem

Momentum equations Continuity equation

Forces due to
gravitational, electrical,
and magnetic fields are
examples of body forces.
Flow restrictions : 1.) Incompressible flow
If is conservative, it
2.) Steady flow can be written as the
3.) Irrotational flow negative gradient of a
4.) Flow along a streamline potential, .
 Bernoulli trinomial
From the aforementioned flow restrictions,

.
Thus,

As a result,

with
 Incompressible flow over a flat plate

The development of boundary layer on a flat plate

(vertical thickness exaggerated greatly)

• Laminar flow - motion in which different layers of fluid

move by sliding over one another, exchanging momentum
and energy by diffusion through molecular mechanisms.
• Turbulence are always unsteady and three-dimensional.
Exchanging momentum and energy are caused by the
macroscopic fluctuations of the fluid particles, which migrate
into different flow regions.
 Incompressible flow over a flat plate
• In the boundary layer, both viscous and inertia forces are
significant. Thus, the Reynolds number is significant in
characterising boundary-layer flows.

• The characteristic length used can be either the flow direction

over which the boundary has developed or some measure of
the boundary-layer thickness.
• Rex is introduced to classify flow regime weather it is laminar
or turbulence.
 Boundary layer transition
• At low Rex, the viscous forces are so strong that they damp down any
perturbations in the fluid motion and therefore the boundary layer
remains laminar.

• At increasing Rex, the inertial effects become more dominant and cause
the boundary layer to become unstable.

• Above a certain Reynolds number, Tollmien-Schlichting waves,

boundary layer disturbances, grow and become unstable. Turbulence can
also be initiated by transient growth of small velocity fluctuations.

• Under typical flow condition, transition usually is considered to occur at

Rex = 500,000. This Rex is so-called the critical Reynolds number. In the
absence of heat transfer, transition in the boundary layer flow over a
smooth flat plate (zero pressure gradient) can be delayed up to Rex = 106.
 Boundary layer transition
• There is no unique value of the Reynolds number since transition
occurs over a region of a plate rather than at a single line across the
plate.

• The transition region extends downstream to the location where the

boundary-layer flow becomes completely turbulent.

• This increases the rate of momentum transfer in the boundary layer

and therefore the rate of the boundary layer's growth.

• Turbulent boundary layer grows at a faster rate than a laminar

boundary layer.
Factors related to boundary layer transition

• Freestream disturbances
• Surface roughness
• Pressure gradients
• Heat transfer
• Body forces
• Noise
• Vibration
 Boundary layer equations
Boundary layer and relevant
reference system -
δ and L are the typical reference
dimensions of the flow, in the y and
in the x directions, respectively.

Two-dimensional incompressible Navier-Stokes equations

 Assumptions
• The boundary layer is very thin. In other words, the thickness, δ, is
very small compared to any linear dimension L in which velocity
variations occur in the x direction.

• Since , the velocity varies much more rapidly in the y

direction than in the x direction.

• It is reasonable to assume that the streamlines within the layer are

nearly parallel since the boundary layer grows slowly. Therefore, the
pressure across the layer is nearly constant.
 An order of magnitude analysis
From the continuity equation, the absolute value of is given by

.
Then, the orders of magnitude of all terms in the x- momentum
equation are considered.

The inertia terms:

 An order of magnitude analysis
The viscous term is comparable in magnitude with the inertia terms :

.
Consequently,
.

It shows that Re must be sufficiently high for the assumption .

The maximum order of magnitude of the pressure variation is given by

.
The momentum equation in the x direction then becomes
 An order of magnitude analysis
Next, the orders of magnitude of all terms in the y- momentum
equation are considered.

The maximum order of magnitude of the pressure variation is given by

.
All the above terms have a maximum order of magnitude that is equal
to that of the terms in the x- momentum equation multiplied by .
The momentum equation in the y direction then becomes
Results from the y- momentum equation
Since inside the boundary layer, .

At the edge of the boundary layer, where is the x-

component of the velocity in the potential flow,

and viscous term can be neglected.

The x- momentum equation at the edge of the boundary layer is

.
Boundary layer equations for steady flow
For steady flow, the x- momentum equation becomes

This equation is parabolic instead of elliptic as characterised in the

complete momentum equation. Thus, the solution can be constructed by
proceeding from upstream to downstream. In other words, the solution
at a certain x depends on the solution for smaller values of x only. No
downstream condition is required. The boundary conditions are
Boundary layer equations for steady flow
The pressure variation is given by the relation

.
 Boundary layer parameters
• Characteristics of a boundary layer can be
described using 4 main parameters as follows:

1. Boundary-layer thickness, δ
2. Displacement thickness, δ*
3. Momentum thickness, θ
4. Shape factor, H
 Boundary-layer thickness, δ
• Boundary-layer thickness, δ – it is defined as the distance
from the surface to the point where the velocity is within 1
percent of the freestream velocity.
 Displacement thickness, δ*
• Displacement thickness, δ* – it is defined as follows:

Since u ≈ Ue at y = δ, the integrand is essentially zero for y ≥ δ.

Consequently,
 What is the purpose of defining δ* ?
From the relation,

It shows that the mass flux passing through the distance δ* in

the absence of the boundary layer is the same as the deficit in
mass flux due to the presence of the boundary layer.
 What is the purpose of defining δ* ?
From the point of view of the flow outside the boundary layer, δ*
can be interpreted as the distance that the presence of the
boundary layer appears to “displace” the flow outward from
the plate. To the external flow, this streamline displacement also
looks like a slight thickening of the body shape.
 Momentum thickness, θ
• Momentum thickness, θ – it is defined as follows:

From the relation,

It shows that the momentum flux passing through the

distance θ in the absence of the boundary layer is the same as
the deficit in momentum flux due to the presence of the
boundary layer.
 Momentum thickness, θ
• Manipulating of the previous equation, we obtain

It shows that the actual momentum flux through the boundary

layer is equal to the momentum flux of fluid moving at a
constant velocity Ue through a distance

Both δ* and θ are integral thicknesses since their definitions

are in terms of integrals across the boundary layer.
 Shape factor, H
• Shape factor is defined as the ratio of the displacement
thickness to the momentum thickness,

• It is an important parameter characterising the velocity profile

within the boundary layer. Smaller values of H indicate a
“fuller” velocity profile (that is, one with higher velocities near
the wall). We see that , and so
 von-Kármán’s integral equation
From the boundary layer equations for steady flow,

Introducing the terms and leads to ,

 von-Kármán’s integral equation
With the aid of integration by parts and the continuity equation,
the last term on the left-hand side can be written as follows:

Since is not a function of y,

 von-Kármán’s integral equation
Due to the fact that

von-Kármán’s integral equation can be obtained as follows:

Note that the ordinary differentiation symbol can be used since all
the quantities appearing in the equation are the function of x only.
Notes on von-Kármán’s integral equation
• It is also denoted as the momentum-integral equation.
• von-Kármán’s integral equation can be applied to both
laminar and turbulent flow.
• The growth rate of the boundary layer and the drag exerted by
the flow on the solid surface can be estimated.
• The prediction of the tangential viscous stress acting over the
solid surface can be obtained by following these three steps:

1. First, the velocity distribution along the x- direction is

determined from inviscid flow theory (the velocity that would exist
in the absence of a boundary layer). Thus, the pressure
distribution inside the boundary layer can be approximated.
2. Assume a reasonable velocity profile shape inside the boundary layer.
3. Relate the wall shear stress to the velocity field.
Laminar boundary layer – Blasius velocity profile

For steady, incompressible flow with zero pressure gradient, the

governing equations of motion reduce to

with boundary conditions

,
,
Laminar boundary layer – Blasius velocity profile
Paul Blasius, one of Prandtl’s students, showed that this equation has a
“similarity” solution. He proposed that the velocity distribution u is a
function only of the freestream velocity Ue , the density ρ, the viscosity
μ, the distance from the wall y, the distance along the plate x. From
dimensional analysis, it gives

Blasius further showed that u/Ue can be written as a function of only

one composite variable η.

, where
Laminar boundary layer – Blasius velocity profile
The variables u/Ue and η are called similarity variables, which means that
if the velocity distribution is plotted using these non-dimensional
variables (instead of dimensional variables such as u and y), it is defined
by a universal curve, for any Reynolds number and any position along
the plate.
Laminar boundary layer – Blasius velocity profile
By introducing the stream function , where and ,
the governing equations of motion reduces the velocity components u, v
into only one variable . The velocity components are given by
Laminar boundary layer – Blasius velocity profile
By differentiating the velocity components,

and

After carrying out various mathematical manipulations, the following

nonlinear, third-order ordinary differential equation is obtained.

with boundary conditions,

Laminar boundary layer – Blasius velocity profile
The previous equation was first solved by Blasius using a power series
expansion, but nowadays it can be solved more precisely using
numerical methods such as the fourth-order Rugge-Kutta numerical
integration. The numerical values of f (η), df/dη, and d2f/dη2 are
Laminar boundary layer – Blasius velocity profile
Once the velocity distribution is obtained, the growth rate of the
boundary layer and the drag exerted by the flow on the solid surface can
be estimated. From the aforementioned table u/Ue= 0.9915 at η = 5.0
and the definition of the boundary layer thickness u = 0.99Ue at y = δ,

or

The wall shear stress can be computed as follows:

Laminar boundary layer – Blasius velocity profile
The friction coefficient Cf is given by

The total force on a plate of length L can be obtained by integrating

Cf from x = 0 to x = L,

The global friction force coefficient CF is

Use of the von-Kármán’s integral equation
For steady, incompressible flow with zero pressure gradient,
the von-Kármán’s integral equation reduces to

Since Ue is constant along the x- direction,

The velocity distribution u/Ue inside the boundary layer is normally

specified as a function of y/δ where δ is a function of x only.
Consequently, it is convenient to define the dimensionless variable η,
and

Clearly, η varies from 0 to 1 across the boundary layer.

 Use of the von-Kármán’s integral equation
The von-Kármán’s integral equation can be rewritten as follows:

u/Ue can be any shape we choose, although it should be a reasonable

approximation to the boundary layer profile. For examples, a polynomial
in y is considered,

The physical boundary conditions are

 Use of the von-Kármán’s integral equation
Evaluating the constants, a, b, and c gives

The wall shear stress is given by

According to the assumed velocity profile,

 Use of the von-Kármán’s integral equation
Substituting τw and u/Ue into the von-Kármán’s integral equation results in

or

Further mathematical manipulations give

or
 Use of the von-Kármán’s integral equation
Integrating again gives

At x = 0, δ = 0, then c = 0 and thus

or

The friction coefficient Cf is given by

Use of the von-Kármán’s integral equation

Typical approximate
boundary layer
profiles used
in the von-Kármán’s
integral equation
Use of the von-Kármán’s integral equation
 Turbulent boundary layer
For incompressible flow with zero pressure gradient, the velocity profile
for turbulent boundary layers are very similar to those for turbulent flow
in pipes and channels. The universal velocity profile is shown below.

where

is the friction velocity.

However, the universal velocity profile is too complicated to apply with

the von-Kármán’s integral equation.
 Turbulent boundary layer
For a smooth pipe, the velocity profile for turbulent flow can be
represented by the empirical power-law equation,

, where U is the centerline velocity and R is the pipe radius.

The ratio of the average velocity to the centerline velocity U for the
power-law profiles is given by
 Turbulent boundary layer
As n increases (due to increasing Reynolds number), the ratio of the
average velocity to the centerline velocity increases. The velocity profile
becomes more blunt or “fuller”.
 Turbulent boundary layer
A suitable velocity profile for turbulent boundary layers on smooth flat
plates with zero pressure gradient is the empirical power-law profile. An
exponent of 1/7 , n = 7, is typically used to model the velocity profile,

Clearly, this profile does not hold in the immediate vicinity of the wall
since it predicts at the wall and we cannot use this profile to
obtain τw in terms of δ as we did for laminar boundary-layer flow.
Therefore, the expression developed for turbulent flow in smooth pipes
is adopted as follows:
 Turbulent boundary layer
For a 1/7-power profile in a smooth pipe, /U = 0.817 and R = δ. Thus,

The von-Kármán’s integral equation can then be applied,

This results in
 Turbulent boundary layer
Further mathematical manipulations give

At x = 0, δ = 0, then c = 0 and thus

, and
 Turbulent boundary layer
As a result,

, and
Pressure gradients in boundary-layer flow
From the von-Kármán’s integral equation without the pressure gradient
term, it shows that the wall shear stress is balanced by a decrease in fluid
momentum,

As we move along the plate, the boundary-layer thickness continues to

increase and the fluid close to the wall is continually slowed down
(losing momentum). τw will never be zero since

for laminar flow,

for turbulent flow

As a result, the fluid close to the wall cannot be brought to zero velocity.
Pressure gradients in boundary-layer flow
However, with the pressure gradient term,

the fluid close to the wall can be brought to rest. Note that

.
Pressure gradients in boundary-layer flow
The presence of pressure gradients of different sign along the surface:
Pressure gradients in boundary-layer flow
Pressure gradients in boundary-layer flow
• When , the pressure gradient is said to be favourable. In
other words, the pressure decreases in the direction of the flow. The
pressure behind the particle is greater than that opposing its motion;
the particle is “sliding down a pressure hill,” without danger of being
slowed to zero velocity. The outer flow Ue accelerates along the x-
direction. The thickness of the boundary layer will increase less
rapidly along the surface, compared to the zero pressure gradient case.

• When , the pressure gradient is said to be adverse. In other

words, the pressure increases in the direction of the flow. The particle
must “climb a pressure hill”. The fluid particle near the wall could be
brought to rest, thus causing the neighboring fluid to be deflected
away from the boundary; when this occurs, the flow is said to separate
from the surface. This phenomenon is known as boundary layer
separation. At the separation point, and τw vanishes.
Pressure gradients in boundary-layer flow
Beyond that point, the particles close to the surface move backwards,
i.e. in the opposite direction relative to the outer stream due to the
increased pressure downstream. The outer flow Ue decelerates along
the x- direction. The thickness of the boundary layer will increase
more rapidly along the surface, compared to the zero pressure gradient
case.

• Clearly, an adverse pressure gradient, is a necessary

condition for separation. In other words, the separation cannot occur
unless . However, it is not a sufficient condition, and this
means that if the pressure gradient along the surface is positive but
not too strong, or if it does not act over a sufficiently long surface
extension, then it is possible that the boundary layer remain attached
to the surface.
Boundary layer separation at a sharp edge

Natural separation Separation at a sharp edge

With the presence of a sharp edge, the outer potential flow would be
characterised by a strong convex curvature of the streamlines
adjacent to that region, followed by a concave curvature to realign the
flow towards the upstream direction. Thus, a high acceleration, and
thus a very low pressure, would occur at the sharp edge, and would
then be followed by an abrupt deceleration causing so strong an
adverse pressure gradient that it would be incompatible with the
existence of an attached boundary layer.
Laminar vs. turbulent boundary layers
Laminar vs. turbulent boundary layers
 Laminar vs. turbulent boundary layers
• Figure 9.6 and 9.7 show that the turbulent profile is much fuller (more
blunt) than the laminar profile. At the same freestream speed, the
momentum flux within the turbulent boundary layer is greater than
within the laminar layer. Separation occurs when the momentum of
fluid layers near the surface is reduced to zero by the combined action
of pressure gradient and viscous forces. Since the momentum of the
fluid near the surface is significantly greater for the turbulent profile,
the turbulent layer is better able to resist separation in an adverse
pressure gradient.
 The effects of pressure gradients to H
• Typically, for a laminar boundary layer, the Blasius solution gives H =
2.59, and for a turbulent boundary layer experiment shows that 1.15 <
H < 1.40, where H decreases with increasing Reynolds number.
Smaller values of H indicate a “fuller” velocity prole (that is, one with
higher velocities near the wall).

• For turbulent boundary layer, H increases from 1.30 for a zero

pressure gradient to approximately 2.50 at separation. For laminar
boundary layer, H increases from 2.59 for a zero pressure gradient to
approximately 3.50 at the separation point.