Southern Methodist University

Bobby B. Lyle School of Engineering

CEE 2342/ME 2342 Fluid Mechanics
Roger O. Dickey, Ph.D., P.E.

III. BASIC EQS. OF HYDRODYNAMICS
A. Fluid Flow Concepts (continued)
Reading Assignment:
Chapter 3 Elementary Fluid Dynamics ,
Sections 3.1 and 3.3
Acceleration Field
Let (x,y,z,t) be any physical quantity, it may be
either a scalar or a vector, that is a function of
position in xyz-space and of time, t (e.g., the
velocity field of a flowing fluid). Then, is
mathematical short-hand for the phrase the rate of
change of the function with respect to the
variable x, i.e., with respect to changing position in
the x-direction.
x c
c
The partial derivative operator, , is used
instead of to emphasize that is a
function of other variables in addition to x.
is evaluated by treating y, z, and t as
constants within the function , and then
differentiating with respect to x in the same way
as for . Evaluation of , , and is
done in a similar manner.
( )
x c
c
( )
dx
d
x c
c
dx
d
y c
c
t c
c
z c
c
Imagine that an observer is riding along on an
infinitesimal fluid particle while measuring the
change in , d, over the small time interval, dt.
Since = (x,y,z,t), the total change observed,
d, is the sum of the changes due to motion in
the x-, y-, and z-directions and the changes over
time, t:
dt
t
dz
z
dy
y
dx
x
d
c
c
+
c
c
+
c
c
+
c
c
=

Then the total rate of change, , is obtained by

dividing the previous equation through by dt:


Simplifying and rearranging,
dt
dt
t dt
dz
z dt
dy
y dt
dx
x dt
d
c
c
+
c
c
+
c
c
+
c
c
=

1
dt
dz
z dt
dy
y dt
dx
x t dt
d
c
c
+
c
c
+
c
c
+
c
c
=

Chain Rule for Partial Derivatives
dt
d
Since the observer is riding along with the
fluid particle, the observer has the same velocity
as the particle in xyz-space:

where,
dt
dz
w
dt
dy
v
dt
dx
u = = = and , ,
k j i V

w v u + + =
Substituting u, v, and w for the appropriate
derivatives in the Chain Rule, and replacing
with the special operator to emphasize that
the resulting expression applies to a specific
fluid particle:
dt
d
z
w
y
v
x
u
t Dt
D
c
c
+
c
c
+
c
c
+
c
c
=

Lagrangian
Reference
Frame
Dt
D
Eulerian
Reference
Frame
The expression for relates Lagrangian and
Eulerian reference frames and is variously called
the material derivative, substantial derivative,
particle derivative, or total derivative depending
upon the reference cited. All standard procedures
and techniques for differentiating functions also
apply to .
Dt
D
Dt
D
The material derivative is viewed as a special
operator of differential calculus defined as:


Once again, u, v, and w are the components of
the velocity vector along the x-, y-, and z-
coordinate axes, respectively.
( ) ( ) ( ) ( ) ( )
z
w
y
v
x
u
t Dt
D
c
c
+
c
c
+
c
c
+
c
c
=
[Equation (4.5), p. 167]
*Basics of Vector Calculus
1. Vectors obey the same principles of differential
calculus as scalars. For example:
Sum rule
Scalar multiplication
Product rule
( )
dt
d
dt
d
dt
d b a
b a + = +
( )
dt
d
a a
dt
d b
b =
( )
dt
d
dt
d
dt
d b
a b
a
ab + =
*Basics of Vector Calculus (continued)
2. Vectors may change in either magnitude or
direction, or both. The derivative of a vector
constant is zero, but only when the vector is
constant in both magnitude and direction.
Now, consider the specific case where is the
velocity field of a flowing fluid,

Then, the acceleration field is obtained by taking
the material derivative of the velocity vector:
k j i V

w v u + + =
Dt
DV
a =
Evaluating the material derivative on the right-
hand side of the acceleration equation:
z
w
y
v
x
u
t c
c
+
c
c
+
c
c
+
c
c
=
V V V V
a
Unsteady
effect,
called local
acceleration
Non-uniform
or spatial
effect, called
convective
acceleration
[Equation (4.3), p. 167]
Acceleration is a vector quantity that may be written
in component form as three scalar equations:
z
w
w
y
w
v
x
w
u
t
w
a
z
v
w
y
v
v
x
v
u
t
v
a
z
u
w
y
u
v
x
u
u
t
u
a
z
y
x
c
c
+
c
c
+
c
c
+
c
c
=
c
c
+
c
c
+
c
c
+
c
c
=
c
c
+
c
c
+
c
c
+
c
c
=
(iii)
(ii)
(i)
[Equation (4.4), p. 167]
Where a
x
, a
y
, and a
z
are the components of the
acceleration vector in the x-, y-, and z-
coordinate directions, respectively.

Streamline Coordinates
Consider the steady flow of an ideal fluid, i.e., a
fluid that is incompressible and inviscid, in xyz-
space. Remember that streamlines correspond to
particle pathlines for the special case of steady
flow. Streamlines are, in general, 3-dimensional
curvilinear paths.
When using a Lagrangian reference frame to study
the motion of an arbitrary infinitesimal fluid
particle moving along its streamline, it is
convenient to use a streamline coordinate system
having one axis parallel to (i.e., along) the
particles streamline (designated the s-axis) with a
second axis normal to the streamline (designated
the n-axis) as illustrated in Figure 4.8, p. 173:
Figure 4.8, p. 173 Streamline Coordinate System
y
s
s > 0 in direction of flow
n > 0 toward the center of
curvature of streamline
x
Top View
A streamline coordinate systemhaving unit
vectors along the s-axis and n-axis,
respectivelyis convenient because the fluid
particle moves only along the s-axis, hence
neither the displacement vector, s, nor the
velocity vector, V, have n-components. On the
other hand, acceleration may have a centrifugal
n-component as shown later.
n s

and

If the particle is located an arbitrary distance s

along the s-axis at time t, and then moves to a
new position at s+s at time t+t , then the
magnitude of the velocity vector V = ds/dt, is
defined as,
( ) ( )
(

+

t
t s t t s
dt
ds
o
o
o 0 t lim
By definition of a streamline, the velocity vector
has a zero n-component, thus:

For an individual particle, the acceleration is
given by the material derivative of the velocity,
s V

V =
( )
Dt
V D s
a

=
Recognizing that unit vector has a constant
magnitude of 1, but changes in direction as the
particle travels along the streamline, i.e., ,
the derivative of the velocity is established by the
product rule from differential calculus:
( )
Dt
D
V
Dt
DV
Dt
V D s
s
s
a

+ = =
( ) t s s

=
s

Separately considering the material derivatives on

the right-hand side of the equation,
dt
dn
n
V
dt
ds
s
V
t
V
Dt
DV
c
c
+
c
c
+
c
c
= (i)
0, Steady
Flow
0, No Flow Normal
to Streamline
V
s
V
V
Dt
DV
c
c
=
dt
dn
n dt
ds
s t Dt
D
c
c
+
c
c
+
c
c
=
s s s s

(ii)
0, Steady
Flow
V
0, Steady
Flow
s
V
Dt
D
c
c
=
s s

Substituting (i) and (ii) into the acceleration
equation,
|
.
|

\
|
c
c
+
|
.
|

\
|
c
c
=
s
V V
s
V
V
s
s a

Only for
Steady Flow
(i)
(ii)
Now consider a differential change in distance, s,
along the s-axis as illustrated in Figure 4.9, p.174:
O
R = radius of curvature
( ) s s o + n

directed from
the streamline
toward O, the
center of curvature
n

Now, consider a geometric argument concerning

as illustrated in Figure 4.9, p. 174:
s c
cs

Similar
Triangles
R
s o
ou
( ) ( )
( ) ( ) s s s
s s s

i.e.,

o o
o o
+ = +
+
s s s
s s s
, change in direction
of unit vector : s

R
s

o
By similar triangles,


Substituting 1 for and rearranging,
is a unit vector, thus its
magnitude in the
denominator is identically
1, by definition
s

( ) s s

( ) s
s
s
s

o
o
=
R
R R
1

1

1

= = s
s
o
o o
o
s s
In the limit as , it can be seen from the similar
triangles shown above that the vectors and
, the latter vector being the scalar product
must both be normal to the streamline, i.e., having
direction . Therefore,
s o
os

o
0 s o
s

1
o
os
n

R
R
n s
n
s

0 lim
=
c
c
=
(

s
s
s
o
o
o
[Figure 4.9, p. 174]
Substituting into the acceleration
equation,

Simplifying,
|
.
|

\
|
+
|
.
|

\
|
c
c
=
R
n
s a

V V
s
V
V
Only for
Steady
Flow
s c
cs n

for

R
R
R
2
2
and i.e.,

V
a
s
V
V a
V
s
V
V
n s
=
c
c
=
+
c
c
= n s a
[Equation (4.7), p. 174]
Notice that,

Substitute this result into the acceleration equation,
( )
s
V
s
V
V
c
c
=
c
c 2
2
( )
( )
R
R
2 2
2 2
and
2
i.e.,

2
V
a
s
V
a
V
s
V
n s
=
c
c
=
+
c
c
= n s a
centrifugal acceleration
normal to path
acceleration
tangent to path
*Important Point
These acceleration components tangent and
normal to the curvilinear fluid streamline should
already be familiar from study of rigid body
motion along curved paths from classical
dynamics.
Pressure Variation in a Flowing Fluid
Pressure Variation Parallel to Streamlines
Still for steady flow of an ideal fluid, consider the
forces producing the previously derived tangent
and normal acceleration components on a fluid
particle in the shape of a parallelepiped that is
moving along a streamline as shown in Figure 3.3,
p. 104 Modified:
Figure 3.3, p. 104
Modified
z
x

s s
A p o
( )
s s s
A p p o +

n n
A p o
( )
n n n
A p p o +
ny A
s
=
sy A
n
=
Side View
Dimensions of the fluid element are s n
tangent and normal to the streamline,
respectively, having thickness y perpendicular to
the figure. Thus, pressure forces parallel to the
streamline act on surfaces of the fluid element
having an area oriented normal to the
streamline, and pressure forces normal to the
streamline act on surfaces having an area
oriented parallel to the streamline.
ny A
s
=
sy A
n
=
Total volume, weight, and mass of the fluid
element are given by:


The vector components of the weight along the
streamline and normal to the streamline are, then:
n A V s A V
n s
= = o o or
n A W s A W
n s
o o = = or
( )
( ) u o o
u o o
cos
sin
W W
W W
n
s
=
=
n A m s A m
n s
o o = = or
Apply Newtons Second Law of motion parallel
to the streamline, i.e., along the s-axis
s s s s s
s axis s
a m F W F
a m F


o o
o
o
=
=
+

Pressure force
at s
Pressure force
at s+s
Substituting appropriate expressions for the forces:

Expand the term for pressure force at s+s:

Simplifying and substituting appropriate
expressions for W and m:
( ) ( ) ( )
s s s s s s
a m A p p W A p o o o u o o = + + sin
( ) ( )
s s s s s s s
a m A p A p W A p o o o o u o o = + sin
( ) ( ) ( )
s s s s s
a s A A p s A o o o o u o o = sin
Dividing through by A
s
s:

Notice in the previous figure that and
substitute:
( )
s
z
o
o
u = sin
( )
s
s
a
s
p

o
o
u = sin
s
s
a
s
p
s
z

o
o
o
o
=
Taking , i.e. both s and n
approaching zero, and noting that p
s
and p
n

become the pressure, p, at a common point in the
limit:

Divide through by using that :
s
a
s
p
s
z
=
c
c

c
c

0 lim V
g
g
1
= =

s
a
g s
p
s
z 1 1
=
c
c

c
c

Substituting the result previously derived for the

acceleration field, :

Since and g are both constants for
incompressible fluids, they may be moved inside
the derivative terms and the equation rearranged:
( )
s
V
a
s
c
c
=
2
2
( ) ( )
0
2
2
=
c
c
+
c
c
+
c
c
s
g V
s
p
s
z
( )
s
V
g s
p
s
z
c
c
=
c
c

c
c

2 1 1
2

Applying the sum rule from differential calculus:

If the derivative of the quantity in square brackets
equals zero, the quantity must be a constant, thus:



This is the Bernoulli Equation, first published by
Daniel Bernoulli in 1738.
0
2
2
=
(

+ +
c
c
g
V p
z
s
) streamline given any for (constant C
2
2
= + +
g
V p
z

[Equation (3.7), p. 107 Rearranged; equation

shown in this equivalent form on p. 111]
*Implications of the Bernoulli Equation
(i) Bernoullis constant, C, remains fixed along a
given streamline, implying that the equation
may be applied between any two points (1)
and (2) along a streamline in steady, inviscid,
incompressible flow as follows:
g
V p
z
g
V p
z
2 2
2
2 2
2
2
1 1
1
+ + = + +

[Equation (3.17), p. 117 Rearranged]
(ii) The various terms in the Bernoulli equation
are called;
z, elevation head
p/, pressure head
when combined, (z + p/) is called the
piezometric head because it is the height
to which the fluid would rise in a
piezometer connected to the flow field
V
2
/2g, velocity head
(iii) For horizontal flow, i.e., (z
1
= z
2
), the equation
written between Points (1) and (2) becomes;

implying that an increase in velocity in the
direction of flow (V
2
> V
1
) must be accompanied
by a decrease in pressure (p
2
< p
1
). This is
known as the Bernoulli Effect ships passing
too close are drawn toward each other possibly
causing a collision, reducing pipe size causes a
decrease in pressure, etc.
g
V p
g
V p
2 2
2
2 2
2
1 1
+ = +

(iv) In a static fluid, V
2
= 0 and V
1
= 0, yielding
one form of the Basic Equation of
Hydrostatics derived previously,


proving that the Basic Equation of
Hydrostatics is simply a special case of the
more general Bernoulli Equation for a
flowing fluid!

2
2
1
1
p
z
p
z + = +
(v) Rearranging the equation, solving for the
pressure difference between Points 1 and 2,




Pressure changes in the direction of flow
differ from that caused by the hydrostatic
pressure distribution by an amount
( ) ( ) ( )
2
1
2
2 1 2 1 2
2
V V
g
z z p p =

/g =
h
( ) ( )
2
1
2
2 1 2
2
V V h p p =

hydrostatic pressure difference

( )
2
1
2
2
2
V V

(vi) Each term in the equation has dimensions [L],

which can be interpreted as energy [FL] per
unit weight [F] of the fluid, ,
kinetic energy,

potential energy,
lb
ft lb
N
m N
or
g
V
mg
mV
weight
E K
2
2 1 . .
2 2
= =
z
weight
z weight
weight
E P
=

=
. .
pressure energy, i.e., pressure force acting
through distance d per unit fluid weight,
( )
( )
p
Ad
d pA
volume
d orce pressure f
=

(vii) The Bernoulli Equation may also be derived
from the work-energy principle from
classical mechanics. Derivation of a similar,
but more general, Energy Equation
applicable to viscous fluids and that is not
constrained to a given streamline will be
considered later.
Pressure Variation Normal to Streamlines
Reconsider the free-body diagram of the fluid
particle in the shape of a parallelepiped that is
moving along a streamline as shown in Figure 3.3,
p. 104 Modified:
Figure 3.3, p. 104
Modified
z
x

s s
A p o
( )
s s s
A p p o +

n n
A p o
( )
n n n
A p p o +
ny A
s
=
sy A
n
=
Side View
Apply Newtons Second Law of motion normal
to the streamline, i.e., along the n-axis,
n n n n n
n axis n
a m F W F
a m F


o o
o
o
=
=
+

Pressure force
at n
Pressure force
at n+n
Substituting appropriate expressions for the forces:

Expand the term for pressure force at n+n:

Simplifying and substituting appropriate
expressions for W and m:
( ) ( ) ( )
n n n n n n
a m A p p W A p o o o u o o = + + cos
( ) ( )
n n n n n n n
a m A p A p W A p o o o o u o o = + cos
( ) ( ) ( )
n n n n n
a n A A p n A o o o o u o o = cos
Dividing through by A
n
n:

Notice in the previous figure that and
substitute:
( )
n
z
o
o
u = cos
( )
n
n
a
n
p

o
o
u = cos
n
n
a
n
p
n
z

o
o
o
o
=
Taking , i.e. both s and n
approaching zero, and noting that p
s
and p
n

become the pressure, p, at a common point in the
limit:

Divide through by using that :
n
a
n
p
n
z
=
c
c

c
c

0 lim V
g
g
1
= =

n
a
g n
p
n
z 1 1
=
c
c

c
c

Substituting the result previously derived for the

acceleration field, :
R
2
V
a
n
=
|
|
.
|

\
|
=
c
c

c
c

R
2
1 1 V
g n
p
n
z

[Equation (3.10a), p. 108 Modified]

*Implications of Normal Acceleration Equation
(i) When streamlines are straight, and thus parallel,
(i.e., steady, uniform flow) then :

Then,

Rearranging and solving for pressure gradient in
n-direction:
R
0
1
lim
2
=
(

|
|
.
|

\
|

R
R
V
g
0
1
=
c
c

c
c

n
p
n
z

n
z
n
p
c
c
=
c
c

By the Chain Rule, , then substitute
for :

Dividing out yields:


This proves that the pressure distribution is
hydrostatic normal to streamlines when the
streamlines are straight and parallel.
n
z
z
p
n
p
c
c
c
c
=
c
c
n
p
c
c
n
z
n
z
z
p
c
c
=
c
c
c
c

n
z
c
c
=
c
c
z
p
Basic Equation of
Hydrostatics
The assumption of approximately straight
streamlines is acceptable in most practical
applications, except for extreme cases of flow
around tight corners and sharp bends.
Since the hydrostatic distribution applies
normal to streamlines, a piezometer attached to
conduits having steady, uniform flow measures
the piezometric head
|
|
.
|

\
|
+

p
z
(ii) If the flow is in the horizontal plane,
and Newtons Second Law reduces to,

Recognizing that

Since n > 0 toward center of curvature,
then indicating that the pressure
decreases toward the center of curvature.
0 =
c
c
n
z
|
|
.
|

\
|
=
c
c

R
2
1 1 V
g n
p

R
2
V
n
p
=
c
c
g
g

= =
0 <
c
c
n
p

This is the reason that the tornadoes have
dangerously low partial vacuums near the
center, or eye, of the storm.