Equations of Motion for Viscous,

Incompressible Fluids
INTRODUCTION
Fluid Kinematics  analysis of fluid flow by considering first the description of motion in
terms of displacement, velocity, and acceleration but without regard to the forces causing it 
Conservation of Mass [Continuity Equation].

Fluid Dynamics  determine the force produced on a solid body by fluid flowing (steadily)
over or through it.

EULERIAN AND LAGRANGIAN COORDINATES

In the Lagrangian description of fluid flow, individual fluid particles are "marked," and their
positions, velocities, etc. are described as a function of time. In the example shown, particles
A and B have been identified. Position vectors and velocity vectors are shown at one instant of
time for each of these marked particles. As the particles move in the flow field, their positions
and velocities change with time.
In the Eulerian description of fluid flow, individual fluid particles are not identified. Instead, a
control volume is defined, as shown in the diagram. Pressure, velocity, acceleration, and all
other flow properties are described as fields within the control volume. In other words, each
property is expressed as a function of space and time, as shown for the velocity field in the
diagram. In the Eulerian description of fluid flow, one is not concerned about the location or
velocity of any particular particle, but rather about the velocity, acceleration, etc. of whatever
particle happens to be at a particular location of interest at a particular time.

In the Lagrangian approach, attention is fixed on a particular mass of fluid as it flows. Suppose
we could color a small portion of the fluid without changing its density. Then in the Lagrangian
framework we follow this colored portion as it flows and changes its shape, but we are always
considering the same particles of fluid. In this reference frame the independent variables are
x0, y0, z0, and t, where x0, y0, z0 are the initial coordinates which a specified fluid element passed
through at time t0 (=0). That is, the coordinates (x0, y0, z0) identify which fluid element is being
considered, and the time t identifies its instantaneous location. So, the instantaneous state of
any variable  is expressed as  (t).

In the Eulerian framework the independent variables are the spatial coordinates x, y, and z
(defining the control volume location) and time t. So, the instantaneous state of any variable 
is expressed as  (x, y, z, t).

THE MATERIAL DERIVATIVE

Let  be any field variable such as the density or temperature of the fluid. In the Lagrangian
approach, the net change of  over a time interval t is expressed as . However, according
to the Eulerian approach the field variable is expressed as  (x, y, z, t). Hence, it can be written
as

𝜕𝛼 𝜕𝛼 𝜕𝛼 𝜕𝛼
∆𝛼 = ∆𝑡 + ∆𝑥 + ∆𝑦 + ∆𝑧
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧
∆𝛼 𝜕𝛼 𝜕𝛼 ∆𝑥 𝜕𝛼 ∆𝑦 𝜕𝛼 ∆𝑧
or, = + 𝜕𝑥 ∆𝑡 + 𝜕𝑦 ∆𝑡 + 𝜕𝑧 ∆𝑡
∆𝑡 𝜕𝑡

∆𝛼 𝜕𝛼 ∆𝑥 𝜕𝛼 ∆𝑦 𝜕𝛼 ∆𝑧 𝜕𝛼
or, lim ( ∆𝑡 ) = + lim ( ∆𝑡 ) 𝜕𝑥 + lim ( ∆𝑡 ) 𝜕𝑦 + lim ( ∆𝑡 ) 𝜕𝑧
∆𝑡→0 𝜕𝑡 ∆𝑡→0 ∆𝑡→0 ∆𝑡→0

𝐷𝛼 𝜕𝛼 𝜕𝛼 𝜕𝛼 𝜕𝛼
or, 𝐷𝑡 = + 𝑢 𝜕𝑥 + 𝑣 𝜕𝑦 + 𝑤 𝜕𝑧 ,
𝜕𝑡

∆𝑥 ∆𝑦 ∆𝑧 𝐷𝛼 ∆𝛼
where, 𝑢 = lim ( ∆𝑡 ), 𝑣 = lim ( ∆𝑡 ), 𝑤 = lim ( ∆𝑡 ) and = lim ( ∆𝑡 ).
∆𝑡→0 ∆𝑡→0 ∆𝑡→0 𝐷𝑡 ∆𝑡→0
 𝜕 𝜕 𝜕
⃗ = 𝑖̂𝑢 + 𝑗̂𝑣 + 𝑘̂𝑤 and ∇
So, considering 𝑉 ⃗ = 𝑖̂ + 𝑗̂ + 𝑘̂
𝜕𝑥 𝜕𝑦 𝜕𝑧

𝐷𝛼 𝜕𝛼 𝜕𝛼 𝜕𝛼 𝜕𝛼
= + 𝑢 𝜕𝑥 + 𝑣 𝜕𝑦 + 𝑤 𝜕𝑧  differential form
𝐷𝑡 𝜕𝑡

𝐷𝛼 𝜕𝛼
= ⃗ .∇
+ (𝑉 ⃗ )𝛼  vector form
𝐷𝑡 𝜕𝑡

 
Lagrangian OR Eulerian Derivative
Material OR
Total Derivative

Eulerian time derivative Eulerian space derivative

(Temporal part) (Spatial/Convective part)

REYNOLDS’ TRANSPORT THEOREM

Consider a fixed mass of fluid and follow it for a short period of time t as it flows. Let  be a
specific property (property per unit volume) of the fluid – (mass or momentum in any direction
or energy) per unit volume. In this Lagrangian approach  is only a function of time and can
be expressed as (t). The control volume containing the specified mass of the fluid is ∀(𝑡)
which can change in shape and size with the flow. Hence the total content of the property is of

the fixed mass of the fluid is ∫∀(𝑡) 𝛼(𝑡)𝑑∀. The rate of change of this property can be expressed

as

𝐷 1
∫ 𝛼(𝑡)𝑑∀ = lim { [ ∫ 𝛼(𝑡 + ∆𝑡)𝑑∀ − ∫ 𝛼(𝑡)𝑑∀]}
𝐷𝑡 ∆𝑡→0 ∆𝑡
∀(𝑡) ∀(𝑡+∆𝑡) ∀(𝑡)

1
= lim { [ ∫ 𝛼(𝑡 + ∆𝑡)𝑑∀ − ∫ 𝛼(𝑡 + ∆𝑡)𝑑∀]
∆𝑡→0 ∆𝑡
∀(𝑡+∆𝑡) ∀(𝑡)

1
+ [ ∫ 𝛼(𝑡 + ∆𝑡)𝑑∀ − ∫ 𝛼(𝑡)𝑑∀]}
∆𝑡
∀(𝑡) ∀(𝑡)

1 𝜕𝛼
= lim { [ ∫ 𝛼(𝑡 + ∆𝑡)𝑑∀]} + ∫ 𝑑∀
∆𝑡→0 ∆𝑡 𝜕𝑡
∀(𝑡+∆𝑡)−∀(𝑡) ∀(𝑡)
 ⃗
𝑽
̂
𝒏

Figure (a) above shows the arbitrary control volume at times t [∀(𝑡)] and t+t [∀(𝑡∆𝑡)]. Figure
(b) shows the superposition of these 2 volumes. It can be concluded from the above figures
that,

⃗ . 𝑛̂∆𝑡𝑑𝑆 ,
𝑑∀= 𝑉

⃗ is the velocity vector on the control surface element dS having an outward normal 𝑛̂.
where, 𝑉

Hence the last equation can be written as,

𝐷 𝜕𝛼
⃗ . 𝑛̂𝑑𝑆]} + ∫
∫ 𝛼(𝑡)𝑑∀ = lim {[ ∫ 𝛼(𝑡 + ∆𝑡)𝑉 𝑑∀
𝐷𝑡 ∆𝑡→0 𝜕𝑡
∀(𝑡) 𝑆(𝑡) ∀(𝑡)

𝜕𝛼
⃗ . 𝑛̂𝑑𝑆 + ∫
= ∫ 𝛼(𝑡)𝑉 𝑑∀
𝜕𝑡
𝑆(𝑡) ∀(𝑡)

⃗ . 𝑎)𝑑∀]
By 𝐺𝑎𝑢𝑠𝑠’𝑠 𝑇ℎ𝑒𝑜𝑟𝑒𝑚 [∫𝑆 𝑎. 𝑛̂𝑑𝑆 = ∫∀ (∇

⃗ . 𝑛̂𝑑𝑆 = ∫ ⃗∇(𝛼𝑉
∫ 𝛼(𝑡)𝑉 ⃗ )𝑑∀
𝑆(𝑡) ∀(𝑡)

Hence, the Reynolds Transport Theorem becomes

𝐷 𝜕𝛼
∫ 𝛼(𝑡)𝑑∀ = ∫ ⃗∇(𝛼𝑉
⃗ )𝑑∀ + ∫ 𝑑∀
𝐷𝑡 𝜕𝑡
∀(𝑡) ∀(𝑡) ∀(𝑡)

or,

𝐷 𝜕𝛼
∫ ⃗ (𝛼𝑉
𝛼𝑑∀ = ∫∀ [ 𝜕𝑡 + ∇ ⃗ )] 𝑑∀  vector form
𝐷𝑡 ∀

𝐷 𝜕𝛼 𝜕(𝛼𝑢) 𝜕(𝛼𝑣) 𝜕(𝛼𝑤)

∫ 𝛼𝑑∀ = ∫∀ [ 𝜕𝑡 + + + ] 𝑑∀  differential form
𝐷𝑡 ∀ 𝜕𝑥 𝜕𝑦 𝜕𝑧

CONSERVATION OF MASS  CONTINUITY EQUATION

The Continuity Equation is a mathematical statement of the principle of the conservation of
mass  matter can neither be created nor destroyed.

Consider a specific mass of fluid of arbitrary volume ∀. If this mass of fluid is followed as it
moves around (Lagrangian frame), then the mass will remain unaltered. If  is the density of a

differential volume d∀, the total mass of the arbitrary volume is ∫∀ 𝜌𝑑∀. Hence the
conservation of mass along with Reynolds Transport Theorem can be written as,

𝐷 𝜕𝜌
⃗ (𝜌𝑉
∫ 𝜌𝑑∀ = ∫ [ + ∇ ⃗ )] 𝑑∀ = 0
𝐷𝑡 𝜕𝑡
∀ ∀

Since the volume ∀ was arbitrarily chosen, the only way in which the above equation can be
satisfied for all possible choices of ∀ is for the integrand to be zero. Then the equation
expressing conservation of mass becomes

𝜕𝜌
+ ⃗∇(𝜌𝑉
⃗)=0
𝜕𝑡
𝜕𝜌 𝜕(𝜌𝑢) 𝜕(𝜌𝑣) 𝜕(𝜌𝑤)
or, + + +
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧

The terms in the last equation above can be rearranged and written as,
𝜕𝜌 𝜕𝜌 𝜕𝜌 𝜕𝜌 𝜕𝑢 𝜕𝑣 𝜕𝑤
+ 𝑢 𝜕𝑥 + 𝑣 𝜕𝑦 + 𝑤 𝜕𝑧 + 𝜌 [𝜕𝑥 + 𝜕𝑦 + ]=0
𝜕𝑡 𝜕𝑧

𝐷𝜌 𝜕𝑢 𝜕𝑣 𝜕𝑤
+ 𝜌 [𝜕𝑥 + 𝜕𝑦 + ]=0
𝐷𝑡 𝜕𝑧

For incompressible fluids  is constant and the conservation of mass can be expressed as
𝜕𝑢 𝜕𝑣 𝜕𝑤
𝜕𝑥
+ 𝜕𝑦 + 𝜕𝑧
=0

⃗∇𝑉
⃗ =0
Applying the principle to a fixed region (control volume) (Eulerian frame) within a fluid as
shown below:

The rate at which mass enters the region = The rate at which mass leaves the region

+ The rate of accumulation of mass in the region

If the flow is steady (i.e., unchanging with time) the rate at which mass is accumulated within
the region is zero. The expression then reduces to:

The rate at which mass enters the region = The rate at which mass leaves the region; which can
be expressed as

⃗ . 𝑑𝐴) = constant,
∫𝐴 𝜌(𝑉

where A is the control surface area. In this integration the normal to a differential area 𝑑𝐴 is
⃗ is causing flow
always outward away from the control volume. Hence if the velocity vector 𝑉
⃗ . 𝑑𝐴) > 0. If the velocity vector 𝑉
to enter the control volume, the term (𝑉 ⃗ is causing flow to

⃗ . 𝑑𝐴) < 0. Hence summation of (rate at which mass enters

exit the control volume, the term (𝑉
the region) + (rate at which mass leaves the region) = 0.

For incompressible flow the above equation becomes,

⃗ . 𝑑𝐴) = constant.
∫𝐴 (𝑉

⃗ . 𝑑𝐴) = 𝑄, where Q is the volumetric

Now for a differential element 𝑑𝐴, the term ∫𝐴 (𝑉
discharge crossing the area. Also 𝑄 = 𝐴𝑉̅, where 𝑉̅ is the average velocity of flow. Hence it
can be concluded that

1
⃗ . 𝑑𝐴) = 𝑄.
𝑉̅ = 𝐴 ∫𝐴 (𝑉 𝐴
CONSERVATION OF LINEAR MOMENTUM
The principle of conservation of linear momentum is an application of Newton’s second law
of motion to an element of the fluid consisting of a fixed mass (of arbitrary volume ∀ and total
surface area S) as it moves (Lagrangian frame) and it is stated as the rate at which the linear
momentum of the fluid mass is changing is equal to the net external force acting on the mass.

The forces acting on a fluid element can be classified into body force and surface force. Body
forces act on each and every fluid particle uniformly and can be expressed as force per unit
mass (or weight). Examples of body force are s gravitational or electromagnetic forces. Surface
forces on the other hand act on surfaces surrounding the fluid volume and can be expressed as
force per unit area. Examples of surfaces force include pressure forces (acting normal to an
area) or viscous stresses (acting tangential to an area).

Consider a differential element of arbitrary volume d∀. The volume is small enough so that the
forces, velocity, and density are uniform over it. However, it is not so small so as to violate the
Continuum Hypothesis. Let the density, velocity, body force per unit mass and surface force
per unit area acing on this element be , 𝑉
⃗ , 𝑓𝐵 and 𝑓𝑆 respectively. Let this differential element
have a total surface area 𝑑𝑆. Then following the Lagrangian formulation and the above
definitions, the conservation of linear momentum can be expressed as

𝐷
⃗ 𝑑∀ = ∫ 𝑓𝑆 𝑑𝑆 + ∫ 𝜌𝑓𝐵 𝑑∀
∫ 𝜌𝑉
𝐷𝑡
∀ 𝑆 ∀

or,

𝐷
∫ 𝜌𝑢𝑑∀ = ∫ 𝑓𝑆𝑥 𝑑𝑆 + ∫ 𝜌𝑓𝐵𝑥 𝑑∀
𝐷𝑡
∀ 𝑆 ∀

𝐷
∫ 𝜌𝑣𝑑∀ = ∫ 𝑓𝑆𝑦 𝑑𝑆 + ∫ 𝜌𝑓𝐵𝑦 𝑑∀
𝐷𝑡
∀ 𝑆 ∀

𝐷
∫ 𝜌𝑤𝑑∀ = ∫ 𝑓𝑆𝑧 𝑑𝑆 + ∫ 𝜌𝑓𝐵𝑧 𝑑∀
𝐷𝑡
∀ 𝑆 ∀

where,
𝑓𝑆 = 𝑖̂𝑓𝑆𝑥 + 𝑗̂𝑓𝑆𝑦 + 𝑘̂𝑓𝑆𝑧

𝑓𝐵 = 𝑖̂𝑓𝐵𝑥 + 𝑗̂𝑓𝐵𝑦 + 𝑘̂𝑓𝐵𝑧

𝑟 = 𝑖̂𝑥 + 𝑗̂𝑦 + 𝑘̂𝑧

⃗ = 𝑖̂𝑢 + 𝑗̂𝑣 + 𝑘̂𝑤

𝑉

However, to fully describe for surface force per unit area or stress at a point, one requires the
magnitude of the stress and 2 directions – one is the direction along which the surface force
acts and the other is the direction of the normal to the surface on which it acts. Hence a pair of
subscripts m and n are required to specify a stress component as ij where i refers to the
direction of the force and j refers to the normal to the surface – each referring to any one of the
3 principal axes. So, the stress at a point is fully defined by the 9 components as
𝜎𝑥𝑥 𝜎𝑥𝑦 𝜎𝑥𝑧
𝜎
𝝈 = [ 𝑦𝑥 𝜎𝑦𝑦 𝜎𝑧𝑧 ]
𝜎𝑧𝑥 𝜎𝑧𝑦 𝜎𝑧𝑧

To visualize the force and planes of action the stress is represented by use of a cubical element
in which the faces of the cube are orthogonal to the cartesian coordinates as shown below and
correspondingly their areas are denoted as dSx, dSy and dSz – which represent the 3 components
of arbitrary differential area 𝑑𝑆 along the three principal axes. The dimensions of the cube are
(∆𝑥 × ∆𝑦 × ∆𝑧) = ∆∀.

So,

The distribution of stress is shown on the 6 faces of the cube. On the left surface since the
outward normal is in the negative x-direction, so the normal stress (acting normal to the
surface) 𝜎𝑥𝑥 also acts in the negative x-direction. Consistent with the sign convention the 2
tangential stresses (acting tangential to the surface), 𝜎𝑥𝑦 and 𝜎𝑥𝑧 also acts along the negative y-
and z-directions respectively. Similarly on the right surface, the outward normal is in the
positive x-direction. Hence, the normal stress as well as the tangential stresses act along the
positive directions of the principal axes. The representation reduces to stress at a point when
the sides of the cube are tending to zero i.e. ∆∀→ 0.
 𝝏𝝈𝒚𝒚
𝝈𝒚𝒚 + ∆𝒚 dSx
dSy 𝝏𝒚

y ∆𝑥
𝝏𝝈𝒚𝒙
𝝈𝒚𝒙 + ∆𝒚
𝝏𝒚
𝝈𝒛𝒛
𝝏𝝈𝒙𝒚
𝝏𝝈𝒚𝒛 𝝈𝒙𝒚 + ∆𝒙
𝝈𝒚𝒛 + ∆𝒚 𝝏𝒙
𝝏𝒚
𝝈𝒛𝒙
𝝈𝒙𝒛 𝝏𝝈𝒙𝒙
𝝏𝝈𝒛𝒚 𝝈𝒙𝒙 + ∆𝒙
𝝈𝒙𝒙 𝝈𝒛𝒚 + ∆𝒛 𝝈𝒛𝒚 𝝏𝒙
𝝏𝒛
𝝏𝝈𝒛𝒙 𝝏𝝈
∆𝒛 𝝈𝒙𝒛 + 𝒙𝒛 ∆𝒙
𝝈𝒙𝒚
𝝈𝒛𝒙 +
𝝏𝒛 𝝏𝒙 x
∆𝑦 𝝈𝒚𝒛
𝝏𝝈𝒛𝒛
𝝈𝒛𝒛 +
𝝏𝒛
∆𝒛
𝝈𝒚𝒙
∆𝑧

z
𝝈𝒚𝒚

dSz

Considering the forces in the 3 principal directions, one can write

𝜕𝜎𝑥𝑥 𝜕𝜎𝑦𝑥
∫ 𝑓𝑆𝑥 𝑑𝑆 = {(𝜎𝑥𝑥 + ∆𝑥) ∆𝑦∆𝑧 − 𝜎𝑥𝑥 ∆𝑦∆𝑧} + {(𝜎𝑦𝑥 + ∆𝑦) ∆𝑥∆𝑧 − 𝜎𝑦𝑥 ∆𝑥∆𝑧} +
𝜕𝑥 𝜕𝑦
𝑆

𝜕𝜎𝑧𝑥
{(𝜎𝑧𝑥 + ∆𝑧) ∆𝑥∆𝑦 − 𝜎𝑧𝑥 ∆𝑥∆𝑦}
𝜕𝑧

𝜕𝜎𝑥𝑥 𝜕𝜎𝑦𝑥 𝜕𝜎𝑧𝑥

={ + + } ∆𝑥∆𝑦∆𝑧
𝜕𝑥 𝜕𝑦 𝜕𝑧

or,

𝜕𝜎𝑥𝑥 𝜕𝜎𝑦𝑥 𝜕𝜎𝑧𝑥

∫ 𝑓𝑆𝑥 𝑑𝑆 = { + + } ∆∀
𝜕𝑥 𝜕𝑦 𝜕𝑧
𝑆

Similarly,

𝜕𝜎𝑥𝑦 𝜕𝜎𝑦𝑦 𝜕𝜎𝑧𝑦

∫ 𝑓𝑆𝑦 𝑑𝑆 = { + + } ∆∀
𝜕𝑥 𝜕𝑦 𝜕𝑧
𝑆

𝜕𝜎𝑥𝑧 𝜕𝜎𝑦𝑧 𝜕𝜎𝑧𝑧

∫ 𝑓𝑆𝑧 𝑑𝑆 = { + + } ∆∀
𝜕𝑥 𝜕𝑦 𝜕𝑧
𝑆

Substituting the above expression in the linear momentum conservation equation,

𝐷 𝜕𝜎𝑥𝑥 𝜕𝜎𝑦𝑥 𝜕𝜎𝑧𝑥
∫ 𝜌𝑢𝑑∀ = ∫ { + + } 𝑑∀ + ∫ 𝜌𝑓𝐵𝑥 𝑑∀
𝐷𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧
∀ 𝑆 ∀

𝐷 𝜕𝜎𝑥𝑦 𝜕𝜎𝑦𝑦 𝜕𝜎𝑧𝑦

∫ 𝜌𝑣𝑑∀ = ∫ { + + } 𝑑∀ + ∫ 𝜌𝑓𝐵𝑦 𝑑∀
𝐷𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧
∀ 𝑆 ∀

𝐷 𝜕𝜎𝑥𝑧 𝜕𝜎𝑦𝑧 𝜕𝜎𝑧𝑧

∫ 𝜌𝑤𝑑∀ = ∫ { + + } 𝑑∀ + ∫ 𝜌𝑓𝐵𝑧 𝑑∀
𝐷𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧
∀ 𝑆 ∀

Hence, using Reynold’s Transport Theorem, bringing all the terms to the left of the ‘is equal
to’ sign and using the tensor representation,

𝜕(𝜌𝑢) 𝜕(𝜌𝑢𝑢) 𝜕(𝜌𝑢𝑣) 𝜕(𝜌𝑢𝑤) 𝜕𝜎𝑥𝑥 𝜕𝜎𝑦𝑥 𝜕𝜎𝑧𝑥

∫[ + + + −{ + + } − 𝜌𝑓𝐵𝑥 ] 𝑑∀ = 0
𝜕𝑡 𝜕𝑥 𝜕𝑣 𝜕𝑧 𝜕𝑥 𝜕𝑦 𝜕𝑧
∀

𝜕(𝜌𝑣) 𝜕(𝜌𝑣𝑢) 𝜕(𝜌𝑣𝑣) 𝜕(𝜌𝑣𝑤) 𝜕𝜎𝑥𝑦 𝜕𝜎𝑦𝑦 𝜕𝜎𝑧𝑦

∫[ + + + −{ + + } − 𝜌𝑓𝐵𝑦 ] 𝑑∀ = 0
𝜕𝑡 𝜕𝑥 𝜕𝑣 𝜕𝑧 𝜕𝑥 𝜕𝑦 𝜕𝑧
∀

𝜕(𝜌𝑤) 𝜕(𝜌𝑤𝑢) 𝜕(𝜌𝑤𝑣) 𝜕(𝜌𝑤𝑤) 𝜕𝜎𝑥𝑧 𝜕𝜎𝑦𝑧 𝜕𝜎𝑧𝑧

∫[ + + + −{ + + } − 𝜌𝑓𝐵𝑧 ] 𝑑∀ = 0
𝜕𝑡 𝜕𝑥 𝜕𝑣 𝜕𝑧 𝜕𝑥 𝜕𝑦 𝜕𝑧
∀

Again, since the volume ∀ was arbitrarily chosen, the only way in which the above equation
can be satisfied for all possible choices of ∀ is for the integrand to be zero. Then the x-equation
expressing conservation of linear momentum becomes

𝜕(𝜌𝑢) 𝜕(𝜌𝑢𝑢) 𝜕(𝜌𝑢𝑣) 𝜕(𝜌𝑢𝑤) 𝜕𝜎𝑥𝑥 𝜕𝜎𝑦𝑥 𝜕𝜎𝑧𝑥

[ + + + ={ + + } + 𝜌𝑓𝐵𝑥 ]
𝜕𝑡 𝜕𝑥 𝜕𝑣 𝜕𝑧 𝜕𝑥 𝜕𝑦 𝜕𝑧

or,

𝜕𝜌 𝜕𝑢 𝜕𝜌 𝜕𝜌 𝜕𝜌 𝜕(𝑢𝑢) 𝜕(𝑢𝑣) 𝜕(𝑢𝑤)

𝑢 +𝜌 + 𝑢𝑢 + 𝑢𝑣 + 𝑢𝑤 +𝜌 +𝜌 +𝜌
𝜕𝑡 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑥 𝜕𝑦 𝜕𝑧
𝜕𝜎𝑥𝑥 𝜕𝜎𝑦𝑥 𝜕𝜎𝑧𝑥
={ + + } + 𝜌𝑓𝐵𝑥
𝜕𝑥 𝜕𝑦 𝜕𝑧

OR
 𝜕𝜌 𝜕𝜌 𝜕𝜌 𝜕𝜌 𝜕𝑢 𝜕(𝑢𝑢) 𝜕(𝑢𝑣) 𝜕(𝑢𝑤)
𝑢[ +𝑢 +𝑣 +𝑤 ]+𝜌[ + + + ]
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧
𝜕𝜎𝑥𝑥 𝜕𝜎𝑦𝑥 𝜕𝜎𝑧𝑥
={ + + } + 𝜌𝑓𝐵𝑥
𝜕𝑥 𝜕𝑦 𝜕𝑧

OR

𝐷𝜌 𝜕𝑢 𝜕𝑣 𝜕𝑤 𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑢
𝑢[ +𝜌 +𝜌 +𝜌 ]+𝜌[ +𝑢 +𝑣 +𝑤 ]
𝐷𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧
𝜕𝜎𝑥𝑥 𝜕𝜎𝑦𝑥 𝜕𝜎𝑧𝑥
={ + + } + 𝜌𝑓𝐵𝑥
𝜕𝑥 𝜕𝑦 𝜕𝑧
Zero from Continuity Eq.

So,

𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑢 1 𝜕𝜎𝑥𝑥 𝜕𝜎𝑦𝑥 𝜕𝜎𝑧𝑥

+𝑢 +𝑣 +𝑤 = [ + + ] + 𝑓𝐵𝑥
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜌 𝜕𝑥 𝜕𝑦 𝜕𝑧

Similarly

𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑣 1 𝜕𝜎𝑦𝑥 𝜕𝜎𝑦𝑦 𝜕𝜎𝑧𝑦

+𝑢 +𝑣 +𝑤 = [ + + ] + 𝑓𝐵𝑦
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜌 𝜕𝑥 𝜕𝑦 𝜕𝑧

𝜕𝑤 𝜕𝑤 𝜕𝑤 𝜕𝑤 1 𝜕𝜎𝑥𝑧 𝜕𝜎𝑦𝑧 𝜕𝜎𝑧𝑧

+𝑢 +𝑣 +𝑤 = [ + + ] + 𝑓𝐵𝑧
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜌 𝜕𝑥 𝜕𝑦 𝜕𝑧

Or

⃗
𝜕𝑉 1
⃗ . ⃗∇)𝑉
+ (𝑉 ⃗ = (∇
⃗ . 𝝈) + 𝑓𝐵
𝜕𝑡 𝜌

ROTATION AND RATE OF SHEAR

(1)Rotation of a fluid element about its own axis
} Obtain the repsective tensors
(2)Shearing of a fluid element

2D analysis – extend concluisions to 3D

Consider a 2D rectangular differential fluid element of dimensions (xy) at time t = 0 – with

one of the 2 pairs of opposite sides parallel to the x-axis and the other parallel to the y-axis. Let
the origin of the 2D coordinates system x-o-y coincide with the centroid of the rectangle – as
shown in the figure below. For purposes of identification, the corners of the fluid element have
been labeled A, B, C, and D and the centroid as O. After a short time-interval t, the centroid
of the fluid element will have moved downstream to some new location as shown in the figure
with the fluid element simultaneously undergoing deformation. Let the corners be labeled A,
B, C, and D and centroid O at this new location.

O

vB vA
uB E
uA
̅̅̅̅ ̅̅̅̅̅
DD′|𝑦
v CC′|𝑦
u
O
vC vD
uC uD

̅̅̅̅̅
DD′|𝑥
̅̅̅̅
CC′|𝑥

Let the velocities in the x- and y-directions at point O be u and v respectively. The
corresponding velocities of the 4 corners A, B, C, and D are indicated by the same symbols
with appropriate suffixes. Considering the smallness of the dimensions of the differential
element it can be written that
𝜕𝑢 𝛿𝑦 𝜕𝑢 𝛿𝑦
𝑢A = 𝑢B = 𝑢 + 𝜕𝑦 ( 2 ); 𝑢C = 𝑢D = 𝑢 − 𝜕𝑦 ( 2 )

and
𝜕𝑣 𝛿𝑥 𝜕𝑣 𝛿𝑥
𝑣A = 𝑣D = 𝑣 + 𝜕𝑥 ( 2 ); 𝑣B = 𝑣C = 𝑣 − 𝜕𝑥 ( 2 )

The x- and y-components of the displacements ̅̅̅̅

CC′ and ̅̅̅̅̅ ̅̅̅̅| , ̅̅̅̅
DD′ are indicated as (CC′ 𝑥
CC′|𝑦 ) and

̅̅̅̅̅| , ̅̅̅̅̅
(DD′ DD′|𝑦 ) respectively in the above figures and can be expressed as,
𝑥
̅̅̅̅ 𝜕𝑢 𝛿𝑦
CC′|𝑥 = 𝑢C 𝛿𝑡 = {𝑢 − 𝜕𝑦 ( 2 )} 𝛿𝑡

̅̅̅̅ 𝜕𝑣 𝛿𝑥
CC′|𝑦 = 𝑣C 𝛿𝑡 = {𝑣 − 𝜕𝑥 ( 2 )} 𝛿𝑡

̅̅̅̅̅ 𝜕𝑢 𝛿𝑦
DD′|𝑥 = 𝑢D 𝛿𝑡 = {𝑢 − 𝜕𝑦 ( 2 )} 𝛿𝑡

̅̅̅̅̅ 𝜕𝑣 𝛿𝑥
DD′|𝑦 = 𝑣D 𝛿𝑡 = {𝑣 + 𝜕𝑥 ( 2 )} 𝛿𝑡

Now in DEC,

C′ E = ̅̅̅̅̅̅
𝜕𝑢 𝛿𝑦 𝜕𝑢 𝛿𝑦
DD′| + CD − ̅̅̅̅̅
CC′ |𝑥 = {𝑢 − ( )} 𝛿𝑡 + 𝛿𝑥 − {𝑢 − ( )} 𝛿𝑡 = 𝛿𝑥
𝜕𝑦 2 𝜕𝑦 2
𝑥

D′ E = ̅̅̅̅̅̅
𝜕𝑣 𝛿𝑥 𝜕𝑣 𝛿𝑥 𝜕𝑣
DD′| − ̅̅̅̅̅
CC′ |𝑦 = {𝑣 + 𝜕𝑥 ( 2 )} 𝛿𝑡 − {𝑣 − 𝜕𝑥 ( 2 )} 𝛿𝑡 = 𝜕𝑥 𝛿𝑥𝛿𝑡
𝑦

Hence, for a small time-interval 𝛿𝑡 (implying small deformation), it can be written that,

𝜕𝑣
D′ E 𝜕𝑥 𝛿𝑥𝛿𝑡 𝜕𝑣
𝛿𝛼 ≈ ′ = = 𝛿𝑡
CE 𝛿𝑥 𝜕𝑥

Hence,

𝛿𝛼 𝜕𝑣 𝜕𝑣
𝛼̇ = lim ( ) = lim ( ) =
𝛿𝑡→0 𝛿𝑡 𝛿𝑡→0 𝜕𝑥 𝜕𝑥

Similarly, it can be shown that

𝛿𝛽 𝜕𝑢
𝛽̇ = lim ( )=
𝛿𝑡→0 𝛿𝑡 𝜕𝑦

In the figure  is measured counterclockwise and β is measured clockwise. Thus, the rate of
clockwise rotation of the fluid element about its centroid is given by

1 1 𝜕𝑢 𝜕𝑣
(𝛽̇ − 𝛼̇ ) = ( − )
2 2 𝜕𝑦 𝜕𝑥

Likewise, the shearing action is measured by the rate at which the sides BC and DC are
approaching each other and is therefore given by the quantity

1 1 𝜕𝑢 𝜕𝑣
(𝛽̇ + 𝛼̇ ) = ( + )
2 2 𝜕𝑦 𝜕𝑥

The foregoing analysis was carried out in two dimensions which may be considered as the
projection of a three-dimensional element on the xy plane. If the analysis is carried out in the
other planes, it may be verified that the rate of rotation of the element about its own axes and
the rate of shearing are given by
1 𝜕𝑢 𝜕𝑣 1 𝜕𝑢 𝜕𝑤
0 ( − 𝜕𝑥) ( − )
2 𝜕𝑦 2 𝜕𝑧 𝜕𝑥
1 𝜕𝑢 𝜕𝑣 1 𝜕𝑣 𝜕𝑤
𝝎= − 2 (𝜕𝑦 − 𝜕𝑥) 0 ( − 𝜕𝑦 )  Anti-symmetric tensor
2 𝜕𝑧
1 𝜕𝑢 𝜕𝑤 1 𝜕𝑣 𝜕𝑤
(− 2 ( 𝜕𝑧 − 𝜕𝑥
) − 2 (𝜕𝑧 − 𝜕𝑦 ) 0 )

and
𝜕𝑢 1 𝜕𝑢 𝜕𝑣 1 𝜕𝑢 𝜕𝑤
( + 𝜕𝑥) ( + )
𝜕𝑥 2 𝜕𝑦 2 𝜕𝑧 𝜕𝑥
1 𝜕𝑢 𝜕𝑣 𝜕𝑣 1 𝜕𝑣 𝜕𝑤
𝜺= ( + 𝜕𝑥) ( + )  Symmetric tensor
2 𝜕𝑦 𝜕𝑦 2 𝜕𝑧 𝜕𝑦
1 𝜕𝑢 𝜕𝑤 1 𝜕𝑣 𝜕𝑤 𝜕𝑤
(2 ( 𝜕𝑧 + 𝜕𝑥 ) 2 𝜕𝑧
( + 𝜕𝑦 ) 𝜕𝑧 )

The antisymmetric 𝝎 and the symmetric 𝜺 are added to obtain the rate of deformation, as

𝜕𝑢 𝜕𝑢 𝜕𝑢
𝜕𝑥 𝜕𝑦 𝜕𝑧
𝜕𝑣 𝜕𝑣 𝜕𝑣
𝒆=
𝜕𝑥 𝜕𝑦 𝜕𝑧
𝜕𝑤 𝜕𝑤 𝜕𝑤
( 𝜕𝑥 𝜕𝑦 𝜕𝑧 )

CONSTITUTIVE EQUATIONS
As is evident from the previous discussion, for a three-dimensional flow field, the stress 𝝈 and
the deformation rate 𝒆 have nine elements each.

For a Newtonian fluid considering the following conditions,

Condition 1: When the fluid is at rest, the stress is hydrostatic and the pressure exerted by the
fluid is the thermodynamic pressure p.

Condition 2: The stress 𝝈 is linearly related to the deformation-rate 𝒆 This is the distinguishing
feature of Newtonian fluids.

Condition 3: Since there is no shearing action in a solid-body rotation of the fluid, no shear
stresses will act during such a motion.

Condition 4: There are no preferred directions in the fluid, so that the fluid properties are point
functions. This is the so-called condition of isotropy, which guarantees that the results obtained
should be independent of the orientation of the coordinate system chosen.
The constitutive relation for stress in a Newtonian fluid becomes
𝜕𝑢 𝜕𝑣 𝜕𝑤 𝜕𝑢 𝜕𝑢
𝜎𝑖𝑗 = −𝑝𝛿𝑖𝑗 + 𝜆𝛿𝑖𝑗 (𝜕𝑥 + 𝜕𝑦 + ) + 𝜇 (𝜕𝑥 𝑖 + 𝜕𝑥𝑗 ), for i, j = x OR y OR z and
𝜕𝑧 𝑗 𝑖

𝛿𝑖𝑗 (𝐾𝑟𝑜𝑛𝑒𝑐𝑘𝑒𝑟 𝐷𝑒𝑙𝑡𝑎) = 1 if i = j

= 0 if i  j

where, 𝜇 is the dynamic viscosity of the fluid and 𝜆 is called the second viscosity coefficient.

So,
𝜕𝑢 𝜕𝑣 𝜕𝑤 𝜕𝑢 𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑤
−𝑝 + 𝜆 ( + + ) + 2𝜇 𝜇( + ) 𝜇( + )
𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑧 𝜕𝑥
𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣 𝜕𝑤 𝜕𝑣 𝜕𝑣 𝜕𝑤
𝝈= 𝜇( + ) −𝑝 + 𝜆 ( + + ) + 2𝜇 𝜇( + )
𝜕𝑦 𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑦 𝜕𝑧 𝜕𝑦
𝜕𝑢 𝜕𝑤 𝜕𝑣 𝜕𝑤 𝜕𝑢 𝜕𝑣 𝜕𝑤 𝜕𝑤
𝜇( + ) 𝜇( + ) −𝑝 + 𝜆 ( + + ) + 2𝜇
( 𝜕𝑧 𝜕𝑥 𝜕𝑧 𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑧 )

For incompressible fluids, from Continuity Equation,

𝜕𝑢 𝜕𝑣 𝜕𝑤
+ 𝜕𝑦 + = 0.
𝜕𝑥 𝜕𝑧

Hence the stress for incompressible fluids can be expressed as

𝜕𝑢 𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑤 pressure
−𝑝 + 2𝜇𝜇( + ) 𝜇( + )
𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑧 𝜕𝑥 motion related
𝜕𝑢 𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑤
𝝈 = 𝜇 ( + ) −𝑝 + 2𝜇 𝜇( + ) = −𝑝𝐈 + 𝝉
𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑦
𝜕𝑢 𝜕𝑤 𝜕𝑣 𝜕𝑤 𝜕𝑤
𝜇( + ) 𝜇( + ) −𝑝 + 2𝜇
( 𝜕𝑧 𝜕𝑥 𝜕𝑧 𝜕𝑦 𝜕𝑧 )

where
𝜕𝑢 𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑤
2𝜇 𝜕𝑥 𝜇 (𝜕𝑦 + 𝜕𝑥) 𝜇 ( 𝜕𝑧 + )
𝜏𝑥𝑥 𝜏𝑥𝑦 𝜏𝑥𝑧 𝜕𝑥
1 0 0 𝜕𝑢 𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑤
𝜏
𝐈 = [0 1 0] and 𝝉 = [ 𝑦𝑥 𝜏𝑦𝑦 𝜏𝑦𝑧 ] = 𝜇 (𝜕𝑦 + 𝜕𝑥) 2𝜇 𝜕𝑦 𝜇 (𝜕𝑧 + 𝜕𝑦 )
0 0 1 𝜏𝑧𝑥 𝜏𝑧𝑦 𝜏𝑧𝑧 𝜕𝑢 𝜕𝑤 𝜕𝑣 𝜕𝑤 𝜕𝑤
(𝜇 ( 𝜕𝑧 + 𝜕𝑥
) 𝜇(
𝜕𝑧
+
𝜕𝑦
) 2𝜇
𝜕𝑧 )

NAVIER-STOKES EQUATIONS (incompressible fluids)

The equation of momentum conservation together with the constitutive relation for a
Newtonian fluid yield the famous Navier-Stokes equations, which are the principal conditions
to be satisfied by a fluid as it flows. From the above equation, it can be obtained that

𝜕𝜎𝑥𝑥 𝜕 𝜕𝑢 𝜕𝑝 𝜕 2𝑢
= {−𝑝 + 2𝜇 } = − + 2𝜇 2
𝜕𝑥 𝜕𝑥 𝜕𝑥 𝜕𝑥 𝜕𝑥
𝜕𝜎𝑦𝑥 𝜕 𝜕𝑢 𝜕𝑣 𝜕 2𝑢 𝜕 2𝑣
= {𝜇 ( + )} = 𝜇 2 + 𝜇
𝜕𝑦 𝜕𝑦 𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑥𝜕𝑦

𝜕𝜎𝑧𝑥 𝜕 𝜕𝑢 𝜕𝑤 𝜕 2𝑢 𝜕 2𝑤
= {𝜇 ( + )} = 𝜇 2 + 𝜇
𝜕𝑧 𝜕𝑧 𝜕𝑧 𝜕𝑥 𝜕𝑧 𝜕𝑥𝜕𝑧
Zero from
Hence, Continuity Eq.
𝜕𝜎𝑥𝑥 𝜕𝜎𝑦𝑥 𝜕𝜎𝑧𝑥 𝜕𝑝 𝜕 2𝑢 𝜕 2𝑢 𝜕 2𝑢 𝜕 𝜕𝑢 𝜕𝑣 𝜕𝑤
[ + + ]=− + 𝜇 ( 2 + 2 + 2) + 𝜇 ( + + )
𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑧

So, finally,

𝜕𝜎𝑥𝑥 𝜕𝜎𝑦𝑥 𝜕𝜎𝑧𝑥 𝜕𝑝 𝜕 2𝑢 𝜕 2𝑢 𝜕 2𝑢

[ + + ]=− + 𝜇 ( 2 + 2 + 2)
𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑧

Similarly,

𝜕𝜎𝑦𝑥 𝜕𝜎𝑦𝑦 𝜕𝜎𝑧𝑦 𝜕𝑝 𝜕 2𝑣 𝜕 2𝑣 𝜕 2𝑣

[ + + ]=− + 𝜇 ( 2 + 2 + 2)
𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑧

𝜕𝜎𝑥𝑧 𝜕𝜎𝑦𝑧 𝜕𝜎𝑧𝑧 𝜕𝑝 𝜕 2𝑤 𝜕 2𝑤 𝜕 2𝑤

[ + + ]=− +𝜇( 2 + + )
𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑧 𝜕𝑥 𝜕𝑦 2 𝜕𝑧 2

Substituting the above relations in the linear momentum conservation equations lead to the
NAVIER-STOKES EQUATIONS for incompressible fluids.

𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑢 1 𝜕𝑝 𝜕 2𝑢 𝜕 2𝑢 𝜕 2𝑢
+𝑢 +𝑣 +𝑤 =− + 𝜈 ( 2 + 2 + 2 ) + 𝑓𝐵𝑥
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜌 𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑧

𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑣 1 𝜕𝑝 𝜕 2𝑣 𝜕 2𝑣 𝜕 2𝑣
+𝑢 +𝑣 +𝑤 =− + 𝜈 ( 2 + 2 + 2 ) + 𝑓𝐵𝑦
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜌 𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑧

𝜕𝑤 𝜕𝑤 𝜕𝑤 𝜕𝑤 1 𝜕𝑝 𝜕 2𝑤 𝜕 2𝑤 𝜕 2𝑤
+𝑢 +𝑣 +𝑤 =− +𝜈( 2 + + ) + 𝑓𝐵𝑧
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜌 𝜕𝑧 𝜕𝑥 𝜕𝑦 2 𝜕𝑧 2

Or

⃗
𝜕𝑉 1
⃗ .∇
+ (𝑉 ⃗ )𝑉
⃗ = − (∇
⃗ 𝑝) + 𝜈(∇2 𝑉
⃗ ) + 𝑓𝐵
𝜕𝑡 𝜌

where, 𝜈 = 𝜇 ⁄𝜌  kinematic viscosity of fluid

𝜕2 𝜕2 𝜕2
and ∇2 = 𝜕𝑥 2 + 𝜕𝑦 2 + 𝜕𝑧 2  Laplacian operator.
In the special case of negligible viscous effects, the above equation becomes

𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑢 1 𝜕𝑝
+𝑢 +𝑣 +𝑤 =− + 𝑓𝐵𝑥
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜌 𝜕𝑥

𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑣 1 𝜕𝑝
+𝑢 +𝑣 +𝑤 =− + 𝑓𝐵𝑦
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜌 𝜕𝑦

𝜕𝑤 𝜕𝑤 𝜕𝑤 𝜕𝑤 1 𝜕𝑝
+𝑢 +𝑣 +𝑤 =− + 𝑓𝐵𝑧
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜌 𝜕𝑧

These are known as Euler equations and the fluid is called inviscid.

REFERENCES
• Mechanics of Fluids – B.S. Massey
• Fundamental Mechanics of Fluids – I.G. Currie
• Introduction to Fluid Mechanics – Haberman
• Fluid Mechanics – White

QUESTIONS
[1] Describe the stress field for a generalized flow in Cartesian frame of reference.

[2] Derive the continuity equation from first principles using an infinitesimal control volume
of rectangular shape and having dimensions ∆𝑥 × ∆𝑦 × ∆𝑧.Identify the net mass flow rate
through each surface of this element as well as the rate at which the mass of the element is
increasing. The resulting equation should be expressed in terms of the cartesian coordinates (x,
y, z, t), the cartesian velocity components (u, v, w), and the fluid density .

[3] What are the Constitutive conditions for a Newtonian fluid? Obtain the independent
components of stress for a fluid in cartesian coordinates (x, y, z) using the cartesian
representation (u, v, w) for the velocity vector, for the case a Newtonian fluid.