MSG 322 FLUID MECHANICS

NOTE 5
THE INTEGRAL FORMS OF THE FUNDAMENTAL LAWS
Introduction
Quantities of interest to engineers can often be expressed in terms of integrals, such as

Volume flow rate is the integral of the velocity over an area

Heat transfer is the integral of the heat flux over an area
Force is the integral of a stress over an area
Mass is the integral of the density over a volume
Kinetic energy is the integral of V 2 /2 over each mass element in a volume

There are, of course, many other integral quantities. To determine an integral quantity, the integrand
must be known, or information must be available so that a good approximation to the integrand can
be made. If the integrand is not known or cannot be approximated with any degree of certainty,
appropriate differential equations must be solved yielding the needed integrand the integration is
then performed giving the engineer the desired integral quantity.
There are many quantities of interest that are not integral in nature, such as
The point of separation of the flow around a body
The concentration of a pollutant in a stream at a certain location
The pressure distribution on the side of a building
where to study subjects such as these, it is necessary to consider the differential equations that describe
the flow.
The Three Basic Laws
The integral quantities of primary interest in fluid mechanics are contained in three basic laws:
1. Conservation of mass
2. First law of thermodynamics
3. Newtons second law

These basic laws are expressed using a Lagrangian

description in terms of a system, a fixed collection of material particles. For example, if we consider flow through a pipe, we could identify a fixed
quantity of fluid at time t as the system (Figure
1) this system would then move due to velocity
to a downstream location at time t + t.
Figure 1: Example of a system in
fluid mechanics

Any of the three basic laws could be applied to this system. This is, however, not an easy task. First
let us state the basic laws in their general form.
Conservation of Mass
The law stating that mass must be conserved is: The mass of a system remains constant.
The mass of a fluid particle is d
V where d
V is the volume occupied by the particle and is its
density. Knowing that the density can change from point to point in the system, the conservation
of mass can be expressed in integral form as
Z
D
d
V=0
(1)
Dt sys
where D/ Dt is used since we are following a specified group of material particles a system.
1

First Law of Thermodynamics

The law that relates heat transfer, work and energy change is the first law of thermodynamics. It
states: The rate of heat transfer to a system minus the rate at which the system does
work equals the rate at which the energy of the system is changing.
Recognizing that both density and specific energy may change from point to point in the system,
it may be expressed as
Z
D

e d
V
(2)
QW =
Dt sys
where the specific energy e accounts for kinetic energy, potential energy and internal energy
per unit mass. Eq.(2) is often referred to as the energy equation. In its basic form stated here,
the first law of thermodynamics applies only to a system, a collection of fluid particles therefore
D/ Dt is used.
Newtons Second Law
Newtons second law also called the momentum equation states: The resultant force acting
on a system equals the rate at which the momentum of the system is changing.
The momentum of a fluid particle of mass is a vector quantity given by V d
V consequently,
Newtons second law may be expressed in an inertial reference frame as
Z
X
D
V d
V
(3)
F=
Dt sys
recognizingP
that both density and velocity may change from point to point in the system. Eq.(3)
F = ma if V and are constant throughout the entire system is often constant
reduces to
but in fluid mechanics the velocity vector invariably changes from point to point. Again D/ Dt is
used to provide the rate-of-change since Newtons second law ia applied to a system.
Moment-of-Momentum Equation
The moment-of-momentum equation results from Newtons second law. It states: The resultant
moment acting on a system equals the rate of change of the angular momentum of
the system.
In equation form this becomes, relative to an inertial reference frame:
Z
X
D
M=
r V d
V
(4)
Dt sys
where r V d
V represents the angular momentum of a fluid particle with mass d
V. The vector
r locates the volume element d
V and is measured from the origin of the coordinate axes, the point
relative to which the resultant moment is measured.
Note that in each of the basic laws the integral quantity is an extensive property of the system. We
will use the symbol Nsys to denote this extensive property. For example Nsys could represent the mass,
the momentum or the energy of the system. The left-hand side of Eq.(1) and the right-hand sides of
Eqs.(2), (3) and (4) may all be expressed as
DNsys
Dt

(5)

where Nsys represents an integral quantity, either a scalar quantity or a vector quantity.
It is also useful to introduce the variable for the intensive property, the property of the system per
unit mass. The relation between Nsys and is given by
Z
Nsys =
d
V.
(6)
sys

As an example, the extensive property of Newtons second law is the momentum

Z
momentumsystem =
V d
V

(7)

sys

which is a vector quantity. The corresponding intensive property would be the velocity vector V. Note
that the density and velocity, which may vary from point to point within the system, may also be
functions of time, as in unsteady flow.
2

 Our interest is most often focused on a device or a region of space, into which fluid enters and/or from
which fluid leaves we identify this region as a control volume.
An example of a fixed control volume is shown in Figure 2(a). A control volume need not be fixed
it could deform as in a piston-cylinder during exhaust or in a balloon as it deflates. We will however
consider only fixed control volumes. This will not limit us in most problem situations. The difference
between a control volume and a system is illustrated in Figure 2(b). The figure indicates that the
system occupies the control volume at time t and has partially moved out of it at time t + t.

(a)

(b)

Figure 2: Example of a fixed control volume and a system: (a) time t (b) time t + t

Since it is often more convenient to focus on a control volume rather than on a system, the first order
of business is to find a transformation that will allow us to express the substantial derivative of a
system (a Lagrangian description) in terms of quantities associated with a control volume (an Eulerian
description) so that the basic laws can be applied directly to a control volume. This will be done in
general and then applied to the specific laws.
System-to-Control-Volume Transformation
We are interested in the time rate of change of the extensive property Nsys as we follow the system
along, that is, DNsys / Dt, and we would like to express this in terms of quantities that pertain to the
control volume.
The derivation involves fluxes of the extensive property in and out of the control volume. A flux is a
measure of the rate at which an extensive property crosses an area. For example, a mass flux is the
rate at which mass crosses an area.
It is useful to introduce vector notation to describe these fluxes. Consider an area element dA of the
control surface, the surface area that completely encloses a control volume.

Figure 3: Illustration of the flux of an extensive property

The property flux across an elemental area dA (see Figure 3) may be expressed by:
flux across dA =
n V dA

(8)

a unit vector normal to the area element dA, always points out of the control volume and
where n
represents the intensive property associated with Nsys .
3

 Note that this expression yields a negative value if it concerns a property influx. Only the normal
V of the velocity vector contributes to this flux term. If there is no normal component
component n
of velocity on a particular area such as the wall of a pipe, no flux occurs across that area.
V indicates a flux out of the volume. A negative n
V, that is, V has a component
A positive n
indicates a flux into the volume. In this note we use n
pointing out of
in the opposite direction of n
the volume.
; the dot product n
V accounts for
The velocity vector V may be at some angle to the unit vector n
the appropriate component of V that produces a flux through the area.
The net property flux out of the control surface is then obtained by integrating over the entire control
surface:
Z
net flux of property =

n V dA.
(9)
c.s.

If the flux is positive, the flux out is larger than the flux in.
Let us return now to the derivative DNsys / Dt. The definition of a derivative from calculus allows us
to write
Nsys (t + t) Nsys (t)
DNsys
= lim
.
(10)
t0
Dt
t

Figure 4: System and fixed control volume

The system is shown in Figure 4 at times t and t + t. Assume that the system occupies the full
control volume at time t. If we were considering a device, such as a pump, the particles of the system
would just fill the device at time t. Since the device, the control volume shown in Figure 4, is assumed
to be fixed in space, the system will move through the device. Equation (10) can then be written
DNsys
Dt

N3 (t + t) + N2 (t + t) N2 (t) N1 (t)
t
N2 (t + t) + N1 (t + t) N2 (t) N1 (t)
= lim
t0
t
N3 (t + t) N1 (t + t)
+ lim
t0
t

lim

t0

(11)

(12)

where, in this second expression, we have simply added and subtracted N1 (t + t) in the numerator.
In the equations above, the numerical subscript denotes the region; for example, N2 (t) signifies the
extensive property in region 2 at time t.
Now, we observe that the first limit on the right-hand side of Eq.(12) refers to the control volume, so
we can write
DNsys
Nc.v. (t + t) Nc.v. (t)
N3 (t + t) N1 (t + t)
= lim
+ lim
.
t0
t0
Dt
t
t

(13)

The first ratio on the right-hand side is dNc.v. / dt, where we use an ordinary derivative since we are
not following specific fluid particles. Thus there results
DNsys
dNc.v.
N3 (t + t) N1 (t + t)
=
+ lim
.
t0
Dt
dt
t
4

(14)

Figure 5: Differential volume elements

Now, we must find expressions for the extensive quantities N3 (t + t) and N1 (t + t). They, of course,
depend on the mass contained in the volume elements dV1 and dV3 , shown in Figure 4 and enlarged
in Figure 5.
always point out of the volume and hence to obtain a positive differential
Note that the unit vector n
volume, a negative sign is required for region 1. Also, note that the cosine of the angle between the
velocity vector and the normal vector is required, thus the presence of the dot product. Referring to
Figures 4 and 5, we have
Z
N3 (t + t) =

n V t dA3
A3
Z
N1 (t + t) =

n V t dA1
(15)
A1

recognizing that A3 plus A1 completely surrounds the control volume, we combine the two integrals
into one integral. That is
Z
N3 (t + t) N1 (t + t) =

n V t dA
(16)
c.s.

where the control surface, denoted by c.s. is the area that completely surrounds the control volume. Substituting Eq.(16) back into Eq.(14) yields the desired result, the system-to-control-volume
transformation, or equivalently the Reynolds transport theorem:
Z
Z
d
DNsys
=
d
V+

n V dA.
(17)
Dt
dt c.v.
c.s.
This is a Lagrangian-to-Eulerian transformation of the rate of change of an extensive integral quantity.
The first integral represents the rate of change of the extensive property in the control volume. The
second integral represents the flux of the extensive property across the control surface; it is nonzero
only where fluid crosses the control surface. Thus we can now express the basic laws in terms of a fixed
volume in space.
We can move the time derivative of the control volume term inside the integral since, for a fixed
control volume, the limits on the volume integral are independent of time; we then write the Reynolds
transport theorem as
Z
Z
DNsys

=
() d
V+

n V dA.
(18)
Dt
c.v. t
c.s.
In this form we have used /t since and are, in general, dependent on the position variables.
Simplifications of the Reynolds Transport Theorem
Many flows of interest are steady flows, so that ()/t = 0. Our system-to-control-volume
transformation then takes the form
Z
DNsys
=

n V dA.
(19)
Dt
c.s.

 Furthermore, there is often only one area A1

across which fluid enters the control volume and
one area A2 across which fluid leaves the control volume; assuming that the velocity vector
is normal to the area (Figure 6), we can write
V1 = V1 over area A1 and n
V2 = V2 over
n
area A2 . Then Eq.(19) becomes
Z
Z
DNsys
=
2 2 V2 dA
1 1 V1 dA. (20)
Dt
A2
A1
Finally, there are many situations that are modeled acceptably by assuming uniform properties
over each plane area (see Figure 7); then the equation simplifies to
DNsys
= 2 2 V2 A2 1 1 V1 A1 .
Dt

(21)

We will find that the system-to-control-volume

Figure 6: Flow into and from a
transformation in this simplified form is most ofdevice
ten used in the application of the basic laws to
problems of interest in fluid mechanics. Some applications will, however, be included that will illustrate non-uniform distributions and unsteady
Figure 7: Uniform velocity profiles
flows.
If we generalize Eq.(23) to include several areas across which the fluid flows, we could write
N

X
DNsys
=
i i Vi ni Ai
Dt
i=1

(22)

V would provide us with the appropriate

where N is the number of areas. The dot product n
sign at each area:
V introduces a negative sign.
For an inlet area, n
V introduces a positive sign.
For and exit area, n
For an unsteady flow in which flow properties are assumed to be uniform throughout the control
volume, the system-to-control-volume equation takes the form
DNsys
d()
c.v.
=V
+ 2 2 V 2 A2 1 1 V 1 A1
Dt
dt

(23)

for one inlet and one outlet with uniform properties.

Conservation of Mass
A system is a given collection of fluid particles; hence its mass remains fixed. This is expressed as
Z
Dmsys
D
=
d
V = 0.
(24)
Dt
Dt sys
In Eq.(6), Nsys represents the mass of the system, so we simply let = 1. Thus the conservation of
mass, referring to Eq.(17), becomes
Z
Z
d
0=
d
V+

n V dA
(25)
dt c.v.
c.s.
or, if we prefer, it takes the equivalent form for a fixed control volume,
Z
Z

0=
d
V+

n V dA.
c.s.
c.v. t
If the flow is steady, there results (see Eq.(19))
Z

n V dA = 0
c.s.

(26)

(27)

which, for a uniform flow with one entrance and one exit, takes the form (see Eq.(23))
2 A2 V2 = 1 A1 V1

(28)

1 V1 = V1 and for an exit n

2 V2 = V2 . Recall n
always points
where for an inlet we have used n
out of the control volume.
If the density is constant in the control volume, the derivative /t = 0 even if the flow is unsteady;
the continuity equation (26) then reduces
A2 V 2 = A1 V 1 .

(29)

This form of the continuity equation is used quite often, particularly with liquids and low-speed gas
flows.
Suppose that the velocity profiles at the entrance and the exit are not uniform. Furthermore, suppose
that the density is uniform over each area. Then the continuity equation takes the form
Z
Z
1
V1 dA = 2
V2 dA
(30)
A1

A2

or letting an overbar denote an average, we can write

1 V1 A1 = 2 V2 A2

(31)

where V1 and V2 are the average velocities over the areas at sections 1 and 2, respectively. In
examples and problems the overbar is often omitted. It should be kept in mind, however, that actual
velocity profiles are usually not uniform; Eqs. (28) and (29) are used with the velocities representing
average velocities.
Any one of the above equations (25) through (31) is referred to as the continuity equation.
Before presenting some examples applying the continuity equation, two fluxes are defined that will be
useful in specifying the quantity of flow.
The mass flux m,
or the mass rate of flow, is
Z
m
=

Vn dA

(32)

and has units of kg/s; Vn is the normal component of velocity.

The flow rate Q, or the volume rate of flow, is
Z
Q=

Vn dA

(33)

and has units of m3 /s.

The mass flux is usually used in specifying the quantity of flow for a compressible flow and the flow
rate for an incompressible flow. We often refer to the flow rate as discharge.
In terms of average velocity, we have
Q = AV
= AV

(34)
(35)

where for the mass flux we assume a uniform density profile; we also assume that the velocity is normal
to the area.

Energy Equation
Many problems involving fluid motion demand that the first law of thermodynamics, often referred to
as the energy equation, be used to relate quantities of interest. If the heat transferred to a device
(a boiler or compressor) or the work done by the device (a pump or turbine), is desired, the energy
equation is obviously needed. It is also used to relate pressures and velocities when Bernoullis equation
is not applicable; this is the case whenever viscous effects cannot be neglected, such as flow through a
piping system in an industrial plant or a golf course.
Let us express the energy equation in control volume form. For a system it is
Z
D

e d
V
QW =
Dt sys

(36)

where the specific energy e includes specific kinetic energy V 2 /2, specific potential energy gz and
specific internal energy u
; that is
V2
e=
+ gz + u
.
(37)
2
We will not include other forms of energy such as energy due to magnetic or electric field-flow field
interactions or those due to chemical reactions.
In terms of a control volume, Eq.(36) becomes
Z
Z
= d
dA.
Q W
e d
V+
eV n
dt c.v.
c.s.

(38)

This can be put in simplified forms for certain restricted flows but first let us discuss the rate-of-heat
. The term Q represents the rate-of-energy transfer across
transfer term Q and the work-rate term W
the control surface due to a temperature difference.
Work-Rate Term
The work-rate term results from work being done by the system or work being done by the control
volume. Work of interest in fluid mechanics is due to a force moving through a distance while it
acts on the control volume.
or power is given by the dot product of a force F with its velocity:
The rate of doing work W
= F VI
W

(39)

where VI is the velocity measured with respect to a fixed inertial reference frame. The negative
sign results because we have selected the convention that work done on the control volume is
negative.
If the force results from a variable stress
acting over the control surface, we must
integrate
Z
=
W
VI dA
(40)
c.s.

where is the stress vector acting on

the elemental area dA, the differential
force being represented by dF = dA,
as shown in Figure 8.

Figure 8: Stress vector acting on the

control surface

For a moving control volume, such as a car, we have to evaluate the velocity with respect to a
fixed reference frame. For example, let us consider an automobile traveling at constant speed. If
we want to apply the energy equation, we could make the car the control volume. In that case,
the velocity in Eq.(39) would be measured relative to a fixed reference and not relative to the car.
If the velocity relative to the car were used, the drag force would have a zero velocity which would
result in no work done; but we know that at high speed, energy from the petrol goes primary to
overcome the drag. Thus the stationary reference frame is needed.

Figure 9: Motion relative to a noninertial reference frame

In general, for moving control volumes the velocity vector VI is related to a relative velocity V,
observed in a reference frame attached to the control volume by
VI = V + S + r

(41)

where S is the velocity of the control volume (see Figure 9). We can now write the work rate of
Eq.(40) as
Z

I
W = V dA + W
(42)
where the inertial work-rate term is given by
Z
I =
W
(S + r) dA.

(43)

c.s.

Next, express the stress vector as the sum of a normal component and a shear component, as
shown in Figure 8, that is
= p
n + s
(44)
where the normal stress component is assumed to be negative the pressure p. Then
Z
Z

I.
W =
p
n V dA
s V dA + W
c.s.

(45)

c.s.

We will allow the shear stress term to consist of two parts:

S : transmitted by a rotating shaft that is cut by the control surface; this term
1. Shaft work W
is important when we deal with flows in pumps and turbines.
shear : results from moving boundaries; this term is required if the control
2. Shear work W
surface itself moves relative to the control volume, as occurs with a moving belt.
Hence the work-rate term becomes
Z

S +W
shear + W
I.
W =
p
n V dA + W
c.s.

The terms are summarized as follows:

R
p
n V dA Work rate resulting from the force due to pressure moving at the
control surface. It is often referred to as flow work
S
W
Work rate resulting from rotating shafts such as that of a pump
or turbine, or the equivalent electric power

Wshear
Work rate due to the shear acting on a moving boundary, such
as a moving belt
I
W
Work rate that occurs when the control volume moves relative
to a fixed reference frame

(46)

General Energy Equation

When the work-rate term of Eq.(46) is substituted into Eq.(38), we obtain the energy equation
in the form

Z
Z 
d
p

Q WS Wshear WI =
e d
V+

n V dA.
(47)
e+
dt c.v.

c.s.
The work-rate term needed to move the pressure force has been moved to the right-hand side, as
is common and is treated like an energy flux term.
Substitution of Eq.(37) into Eq.(47) results in


Z  2
Z  2
VI
p
VI
W
S W
shear W
I = d
dA. (48)
+ gz + u
d
V+
+ gz + u
+
V n
Q
dt c.v. 2
2

c.s.
This general form of the energy equation is useful in analyzing fluid flow problems that may
include time-dependent effects and nonuniform profiles.
In many fluid flows, useful forms of energy (kinetic energy and potential energy) and flow work
are converted into unusable energy forms (internal energy or heat transfer).
If we assume that the temperature of the control volume remains unchanged, the internal energy
does not change and the losses are balanced by heat transfer across the control surface. This heat
transfer can be the result of convection, radiation or conduction at the control surfaces.
Thus we define losses as the sum of all the terms representing unusable forms of energy:
Z
Z
d

dA.
u
d
V+
u
V n
losses = Q +
dt c.v.
c.s.

(49)

We can now rewrite the energy equation as



Z  2
Z  2
VI
VI
p
S W
shear W
I = d
dA + losses. (50)
+ gz d
V+
+ gz +
W
V n
dt c.v. 2
2

c.s.
Losses are due to two primary effects:
1. Viscosity causes internal friction that results in increased internal energy (temperature increase) or heat transfer.
2. Changes in geometry result in separated flows that require useful energy to maintain the
resulting secondary motions in which viscous dissipation occurs.
In a conduit, the losses due to viscous effects are distributed over the entire length, whereas the
loss due to a geometry change (a valve, an elbow, an enlargement) is concentrated in the vicinity
of the geometry change.
The analytical calculation of losses is rather difficult, particularly when the flow is turbulent. In
general, the prediction of losses is based on empirical formulas. For a pump or turbine the losses
are expressed in terms of efficiency. For example, if the efficiency of a pump is 80%, the losses
would be 20% of the energy input to the pump.
It may be that the objective in a particular fluid flow is to change the internal energy of the fluid,
such as in the steam generator (boiler) of a power plant, by the transfer of heat; then the definition
of losses above must be altered so that the loss term includes only the dissipative effects of the
viscosity of the fluid. Generally, for problems of interest in fluid mechanics, Eq.(50) is acceptable.
Steady Uniform Flow
Consider a steady-flow situation in which there is one entrance and one exit across which uniform
shear = W
I = 0 with VI = V . For such a flow the
profiles can be assumed. Also, assume that W
term (V 2 /2 + gz + p/) in Eq.(50) is constant across the cross section because V is constant (we
assume a uniform velocity profile) and the sum of p/ + gz is constant if the streamlines at each
section are parallel. The energy equation (Eq. (50)) then simplifies to
 2

 2

S = 2 V2 A2 V2 + p2 + gz2 1 V1 A1 V1 + p1 + gz1 + losses
(51)
W
2
2
2
1
where the subscripts 1 and 2 refer to the entrance and exit, respectively.

10

 The mass flux is given by m

= 1 A1 V1 = 2 A2 V2 . After dividing by mg
we have

S
V 2 V12
p2
p1
W
= 2
+

+ z2 z1 + hL
mg

2g
2
1

(52)

where we have introduced the head loss hL , defined to be

hL =

u
2 u
1
Q

.
g
mg

(53)

It is often written in terms of a loss coefficient K as

hL = K

V2
2g

(54)

where V is most often V1 or V2 .

The head loss is referred to as a head since it has dimension of length. We may also refer to
V2 /2g as the velocity head and p/ as the pressure head since those terms also have dimensions
of length. Also recall from previous note that p/ + z is called the piezometric head. Further,
the sum of the piezometric head and the velocity head is called the total head.
The energy equation, in the form of Eq.(52) is useful in many applications and is, perhaps, the
most often used form of the energy equation. If the losses are negligible and if there is no shaft
work, we note that the energy equation takes the form
V22
p2
V2
p1
+
+ z2 = 1 +
+ z1 .
2g
2
2g
1

(55)

Observe that the energy equation has been reduced to a form identical with Bernoullis equation
when 2 = 1 (a constant density flow). We must remember, however, that Bernoullis equation
is a momentum equation applicable along a streamline and the equation above is an energy
equation applied between two sections of a flow. It is not surprising that both should predict
identical results from the conditions stated because the velocity head is constant over a cross
section and the sum of pressure head and elevation remains constant over a cross section.
The energy equation (52) may be applied to any steady, uniform flow with one entrance and
one exit. The control volume is usually selected such that the entrance and exit sections have a
uniform total head. For example, it may be applied to water flow through a long pipeline; the
total head at the entrance and exit may then be evaluated conveniently at the center of the pipe
entrance and exit.

Figure 10: Application of the energy equation to a gate in an open channel

The energy equation may be applied to the flow passing a gate (Figure 10). An appropriate
control volume is shown. The total head at the entrance and exit can be evaluated at any point
at the entrance and exit, respectively. However, a convenient choice would be the points at the
water surface. Thus the energy equation becomes
V12
p1
V2
p2
+
+ h1 = 2 +
+ h2 + hL
2g

2g

where the pressures have been set to zero (p1 = p2 = 0).

11

(56)

 If we had picked the centroids of the entrance and exit, as shown in Figure 10, we would have
obtained
V12
p1
h1
V2
p2
h2
+
+
= 2 +
+
+ hL .
(57)
2g

2
2g

2
We see that this result is the same as in Eq. (56) if we substitute p1 = h1 /2 and p2 = h2 /2. For
completeness, it is noted that the losses between 1 and 2 in Figure 10 could be neglected because
internal viscous effects occur only over a relatively short distance and no significant secondary
flows are generated.

Figure 11: Application of the energy equation to a T-section

The energy equation (50) can be applied to any control volume. For example, consider steady,
uniform, incompressible flow through a T-section in a pipe (Figure 11) in which there is one
entrance and two exits. The energy equation can be applied to each of two control volumes, one
for the mass flux that exits section 2 and the other for the mass flux that exits section 3:
V12
p1
V2
p2
+
+ z1 = 2 +
+ z2 + hL12 ,
2g

2g

V12
p1
V2
p3
+
+ z1 = 3 +
+ z3 + hL13 .
2g

2g

(58)

The loss terms in Eq.(58) include the losses between the inlet and the respective exits. If the
losses are negligible, the energy equation reduces to a from similar to the Bernoulli equation
being applied along a streamline going from 1 to 2 or a streamline going from 1 to 3.
S /mg)
It is often conventional to call the energy term (W
associated with a pump the pump head

HP and the term (WS /mg)

associated with a turbine the turbine head HT . Then the energy

equation, for an incompressible flow, takes the form
HP +

V12
p1
V2
p2
+
+ z1 = HT + 2 +
+ z2 + hL .
2g

2g

(59)

In this form we have equated the energy at the inlet plus added energy to the energy at the exit
plus extracted energy (energy per unit weight). If any of the quantities is zero (e.g. there is no
pump), the appropriate term is simply omitted. The term HP and HT above represent the energy
that is transferred to and from the fluid, respectively. If the energy delivered by the turbine or
required by the pump is desired, the efficiency of each device must be used.
The power generated by the turbine with an efficiency of T is simply
T = mgH
W

T T = QHT T .

(60)

The power requirement by a pump with an efficiency of P would be

QHP
P
P = mgH
W
=
.
P
P

(61)

The unit for power is watts, ft-lb/sec or horsepower. 1 horsepower is equivalent to 746 W or 550
ft-lb/sec.

12

Steady Nonuniform Flow

If the assumption of uniform velocity profiles is not acceptable for a problem of interest, as is
sometimes the situation, we have to consider the control-surface integral in Eq.(50) with the
proper expression for the velocity distribution.
In practice, a velocity distribution can be accounted for by introducing the kinetic-energy
correction factor , defined by
R 3
V dA
= 3
(62)
V A
where V is the average velocity over the area A, given by Q = AV . Then the term that accounts
for the flux of kinetic energy in Eq.(50) is
Z
1
1

(63)
V 3 dA = V 3 A
2 A
2
where we have used V
n = V and VI = V . Using this factor, we can account for nonuniform
velocity distributions by modifying Eq.(59) to read
HP + 1

p1
V 2
p2
V12
+
+ z 1 = HT + 2 2 +
+ z2 + hL
2g

2g

(64)

where V1 and V2 are the average velocities at section 1 and 2, respectively.

For a flow with a parabolic profile in a pipe we can calculate = 2.0. For most internal turbulent
flows, however, the profile is nearly uniform with ' 1.05. Hence we simply let = 1 since it is
so close to unity; this will always be done unless otherwise stated, since most of the internal flows
that we encounter are, in fact, turbulent flows.

13

Momentum Equation
Derived from Newtons second law.
General Momentum Equation
Newtons second law, often called the momentum equation, states that the resultant force
acting on a system equals the rate of change of momentum of the system when measured in an
inertial reference frame, that is
Z
X
D
F=
V d
V.
(65)
Dt sys
Using Eq.(17)
DNsys
d
=
Dt
dt

n V dA

d
V+
c.v.

c.s.

with replaced by V, this is written for a control volume as

Z
Z
X
d
) dA
V d
V+
V(V n
F=
dt c.v.
c.s.

(66)

is simply a scalar for each differential area dA. The control surface integral on the
where V n
right represents the net momentum flux across the control surface of the fluid entering and/or
leaving the control volume.
P
When applying Newtons second law the quantity
F represents all forces acting on the control
volume. The forces include the surface forces resulting from the surroundings acting on the control
surface and body forces that result from gravity and magnetic fields.
The momentum equation is often used to determine the forces induced by the flow. For example,
the equation allows us to calculate the force on the support of an elbow in a pipeline or the force
on a submerged body in a free-surface flow.

Figure 12: Forces acting on the c.v. of a horizontal nozzle:

(a) c.v. includes nozzle and fluid in nozzle;
(b) c.v. includes fluid in nozzle only. Body forces are neglected

When we apply the momentum equation the surrounding fluid and sometimes the entire conduit
or container is separated from the control volume. For example, in the horizontal nozzle of Figure
12(a), the nozzle and the fluid in the nozzle are isolated. Thus, care must be taken to include
the pressure forces shown and the force Fjoint . It is convenient to use gage pressures so that the
pressure acting on the exterior of the pipe is then zero.
Alternatively, we could have selected a control volume that includes only the fluid in the nozzle
(Figure 12(b)). In that case we have to consider the pressure forces at the entrance and exit and
the resultant pressure force Fnozzle of the interior wall of the nozzle on the fluid.
A free body of the nozzle excluding the fluid shows that the force Fjoint and Fnozzle are equal in
magnitude. If the problem is to determine the forces exerted by the flow on the nozzle (Figure
12(b)), we have to reverse the direction of the calculated force Fnozzle .

14

Steady Uniform Flow

Eq.(66) can be simplified considerably if a device has entrances and exits across which the flow
may be assumed to be uniform and if the flow is steady. Then there results
X

F=

N
X

)
i Ai Vi (Vi n

(67)

i=1

where N is the number of flow exit/entrance areas.

= V since the unit vector points out of the volume and at the exit V n
=V.
At an entrance V n
If there is only one entrance and one exit, as in Figure 12, the momentum equation becomes
X
F = 2 A2 V 2 V 2 1 A1 V 1 V 1 .
(68)
Using continuity,
m
= 1 A1 V 1 = 2 A2 V 2
the momentum equation takes the simplified form
X
F = m(V

2 V1 ).

(69)

Note that the momentum equation is a vector equation which represents the three scalar equation
X
Fx = m(V
2x V1x ),
X
Fy = m(V
2y V1y ),
(70)
X
Fz = m(V
2z V1z ).
If we consider the nozzle of Figure 12(a) and we want to determine the xcomponent of the force
of the joint on the nozzle, (V1 )x = V1 and (V2 )x = 0 so that the momentum equation for the
xdirection becomes
X
Fx = (Fx )joint + p1 A1 = mV
1
(71)
Similarly, we could write the ycomponent equation that would contain the term (Fy )joint .

Figure 13 : Force of the flow on a gate in a free-surface flow

An example of a free-surface flow in a rectangular channel is shown in Figure 13. If the force
of the gate on the flow is desired, the following expression can be derived from the momentum
equation
X
Fx = Fgate + F1 F2 = m(V
2 V1 )
(72)
where F1 and F2 are pressure forces.

15

Momentum Equation Applied to Deflectors

The application of the momentum equation to deflectors forms an integral part of the analysis of
many turbomachines, such as turbines, pumps and compressors. In this section, we will consider
two separated parts:
1. Fluid jets deflected by stationary deflectors
2. Fluid jets deflected by moving deflectors
For both problems we will assume the following
1. The pressure external to the fluid jets is everywhere constant so that the pressure in the fluid
as it moves over a deflector remains constant.
2. The frictional resistance due to the fluid-deflector interaction is negligible so that the relative speed between the deflector surface and the jet stream remains unchanged, a result of
Bernoullis equation.
3. Lateral spreading of a plane jet is neglected.
4. The body force, the weight of the control volume, is small and will be neglected.
Stationary Deflector
Let us first consider the stationary deflector, illustrated in Figure 14. Bernoullis equation
allows us to conclude that the magnitudes of the velocity vectors are equal (i.e. V2 = V1 ),
since the pressure is assumed to be constant external to the fluid jet and elevation changes
are negligible (refer Bernoullis equation in Note 4)

Figure 14: Stationary deflector

Assuming steady, uniform flow, the momentum equation takes the form of Eq.(69), which for
the x and ydirections becomes
Rx
Ry

= m(V
2 cos V1 ) = mV
1 (cos 1),
=

mV
2 sin = mV
1 sin .

(73)

For given jet conditions the reactive force components can be calculated.
Moving Deflectors
The situation involving a moving deflector depends on whether a single deflector is moving
(the blade on a snowplow or a water scoop used to slow a high-sped train) or whether a series
of deflectors is moving (the vanes on a turbine).
Let us first consider the single deflector shown in Figure 15 to be moving in the positive x
direction with the speed VB . In a reference frame attached to the stationary nozzle, from
which the fluid jet issues, the flow is unsteady; that is, at a particular point in space, the flow
situation varies with time.
A steady flow is observed, however, from a reference frame attached to the deflector. From
this inertial reference frame, moving with the constant velocity VB , we observe the relative
speed Vr1 entering the control volume to be V1 VB , as shown. It is this relative speed
that remains constant as the fluid flows relative to the deflector; it does not change since
the pressure does not change. Hence, from this moving frame, the momentum equation (69)
takes the forms
Rx
Ry

= m
r (V1 VB )(cos 1),
= m
r (V1 VB ) sin ,

(74)

where m
r represents only that part of the mass flux exiting the fixed jet that has its momentum
changed.
16

Figure 15: Moving deflector

Since the deflector moves away from the fixed jet some of the fluid that exits the fixed jet
never experiences a momentum change; this fluid is represented by the distance VB t, shown
in Figure 15. Hence
m
r = A(V1 VB )
(75)
where the relative speed (V1 VB ) is used in the calculation; the mass flux AVB is subtracted
from the exiting mass flux AV1 to provide the mass flux m
r that experiences a momentum
change.

Figure 16 : Fluid striking a series of vanes

Figure 17 : Detail of the flow situation involving a series of vanes:

(a) average position of jet; (b) entrance velocity polygon; (c) exit velocity polygon

For a series of vanes (a cascade), the jets may be oriented to the side as shown in Figure 16.
The actual force on a particular vane would be zero until the jet strikes the vane; then the
force would increase to a maximum and decrease to zero as the vane leaves the jet.
We will idealize the situation as follows: Assume that, on the average, the jet is deflected by
the vanes as shown in Figure 16 and Figure 17(a) as viewed from a stationary reference frame;

17

the fluid jet enters the vanes with an angle 1 and exits with an angle 2 . What is desired,
however, is that the relative velocity enter the vanes tangent to the leading edge of the vanes,
that is, Vr1 in Figure 17(b) is at angle 1 . The relative speed then remains constant as the
fluid travels over the vane with the exiting relative velocity Vr2 leaving with the vane angle
2 . The relative and absolute velocities are related with the velocity equations which are
displayed by the velocity polygons of Figures 17(b) and 17(c).
Assuming that all of the mass exiting the fixed jet has its momentum changed, we can write
the momentum equation as
Rx = m(V
2x V1x ).
(76)
Interest is usually focused on the xcomponent of force since it is this component that is
related to the power output (or requirement). The power would be found by multiplying the
xcomponent force by the blade speed for each jet; this takes the form
= N Rx VB
W

(77)

where N represents the number of jets. The ycomponent force does not move in the
ydirection so it produces no power.
Momentum Equation Applied to Propellers

Figure 18 : Propeller in a fluid flow

Consider the propeller of Figure 18 with the streamlines shown forming the surface of a control
volume in which the fluid enters with a uniform velocity V1 and exits with a uniform velocity V2 .
The outer streamlines just touch the tips of the propeller. This flow situation can be seen to be
identical to that of a propeller moving with velocity V1 in a stagnant fluid by adding V1 to the
left in Figure 18.
The momentum equation, applied to the large control volume shown, gives
F = m(V
2 V1 ).

(78)

This control volume is not sufficient, however, since the areas A1 and A2 are unknown. We know
the flow area A of the propeller. So a control volume is drawn close to propeller such that V3
= V4
and A3
= A4 = A. The momentum equation (69) in the xdirection gives
F + p3 A p4 A = 0

(79)

F = (p4 p3 )A.

(80)

or
Now, since viscous effects would be quite small in this flow situation the energy equation up to
the propeller and then downstream from the propeller is used to obtain
V12 V32
p1 p3
+
=0
2

and

V42 V22
p4 p2
+
= 0.
2

Adding these equations together, recognizing that p1 = p2 = patm , we have

(V22 V12 ) = p4 p3 .
2
Inserting this and Eq.(80) into Eq.(78) results in

(81)

(82)

1
(V2 + V1 )
(83)
2
where we have used m
= AV3 since the propeller area is the only area known. This result shows
that the velocity of the fluid moving through the propeller is the average of the upstream and
downstream velocities.
V3 =

18

 The input power needed to produce this effect is found by applying the energy equation between
sections 1 and 2, where the pressures are atmospheric; neglecting losses, Eq.(52) takes the form
2
2
fluid = V2 V1 m

W
2

(84)

fluid is the energy input between the two sections. The moving propeller requires power
where W
given by
prob
W

= F V1
= mV
1 (V2 V1 ).

(85)

The theoretical propeller efficiency is then

P =

prop
V1
W
=
.
fluid
V3
W

(86)

In contrast to the propeller, a wind machine extracts energy from the airflow; the downstream
velocity is reduced and the diameter is increased.
Steady Nonuniform Flow
If we cannot assume uniform velocity profiles, we can let the momentum flux be expressed as
Z
V 2 dA = V 2 A
(87)
A

where we have introduced the momentum-correction factor , expressed explicitly as

R 2
V dA
.
= 2
V A

(88)

The momentum equation (69), for a steady flow with one inlet and one exit, can then be written
as
X
F = m(
2 V2 1 V1 ).
(89)
For a laminar flow with a parabolic profile in a circular pipe, = 4/3. If a profile is given,
however, the integral is usually integrated and Eq.(66) is used.
Noninertial Reference Frames
In certain situations it may be necessary to choose a noninertial reference frame in which the
velocity is measured. This would be the case if we were to study the flow through a dishwater
arm, around a turbine blade, or from a rocket.
Relative to a noninertial reference frame, Newtons second law takes the form (refer Note 4):
Z
X
D
F =
V d
V
Dt sys

Z  2
d S
d
is volume
+
+ 2 V + ( r) +
r d
V, V
(90)
2
dt
sys dt
where V is the velocity relative to the noninertial frame; the acceleration a of each particle in the
system is already accounted for in the first integral. Equation (90) is often written as
Z
X
D
F FI =
V d
V
Dt sys
Z
Z
d
) dA
=
V d
V+
V(V n
(91)
dt c.v.
c.s.
where FI is called the inertial body force, given by

Z  2
d S
d
FI =
+
2

V
+

r)
+

r
d
V.
2
dt
sys dt

(92)

Since the system and control volume are identical at time t the system integration can be replaced
with a control volume integration in the integral of Eq.(92).
19

Moment-of-Momentum Equation
In the preceding section we determined the magnitude of force components in a variety of flow situations.
To determine the line of action of a given force component, it is often necessary to apply the moment-ofmomentum equation. Also, in analyzing the flow situation in devices that have rotating components the
moment-of-momentum equation is needed to relate the rotational speed to the other flow parameters.
Since it may be advisable to attach the reference frame to the rotating component, we will write the
general equation with the inertial forces included. It is
Z
X
D
r V d
V
(93)
M MI =
Dt sys
where

Z
MI =


d2 S
d
r
+ 2 V + ( r) +
r d
V.
dt2
dt


(94)

This inertial moment MI accounts for the fact that a noninertial reference frame was selected; it is
simply the moment of FI (see Eq.(92)). Applying the system-to-control volume transformation, the
moment-of-momentum equation for a control volume becomes
Z
Z
X
d
) dA.
r V d
V+
r V(V n
(95)
M MI =
dt c.v.
c.s.

20