Advanced Fluid Mechanics April

2016

MCE2121
ADVANCED FLUID MECHANICS

LECTURE NOTES
Module-II

Prepared By
Dr. Prakash Chandra Swain
Professor in Civil Engineering
Veer Surendra Sai University of Technology, Burla

Branch - Civil Engineering

Specialization-Water Resources Engineering
Semester – 1st Sem

Department Of Civil Engineering

VSSUT, Burla
 Advanced Fluid Mechanics April
2016

Disclaimer
This document does not claim any originality and cannot be
used as a substitute for prescribed textbooks. The information
presented here is merely a collection by Prof. P.C.Swain with
the inputs of Post Graduate students for their respective
teaching assignments as an additional tool for the teaching-
learning process. Various sources as mentioned at the
reference of the document as well as freely available materials
from internet were consulted for preparing this document.
Further, this document is not intended to be used for
commercial purpose and the authors are not accountable for
any issues, legal or otherwise, arising out of use of this
document. The authors make no representations or warranties
with respect to the accuracy or completeness of the contents of
this document and specifically disclaim any implied warranties
of merchantability or fitness for a particular purpose.
 Advanced Fluid Mechanics 2016

Course Content
Module II

Viscous Flow and Boundary Layer Theory: Study of Local Behavior, Differential

Approaches in Analysis of Viscous Flows, Equations of Motion of a Viscous

Flow, Navier – Stokes Equations, Exact and Approximate Solution of N – S

Equations, Hele – Shaw Flow, Creeping Flow past a Sphere, Boundary Layer

Concept, Prandtl’s Boundary Layer Equations, Laminar Boundary Layer Along a

Flat Plate, Integral Equation, Blassius Solution.

Prof P.C.Swain Page 3

 Advance Fluid Mechanics 2016

Lecture Note 1
Boundary Layer Theory

Introduction

 The boundary layer of a flowing fluid is the thin layer close to the wall
 In a flow field, viscous stresses are very prominent within this layer.
 Although the layer is thin, it is very important to know the details of flow within it.
 The main-flow velocity within this layer tends to zero while approaching the wall (no-
slip condition).
 Also the gradient of this velocity component in a direction normal to the surface is large
as compared to the gradient in the streamwise direction.

Boundary Layer Equations

 In 1904, Ludwig Prandtl, the well known German scientist, introduced the concept of
boundary layer and derived the equations for boundary layer flow by correct
reduction of Navier-Stokes equations.

 He hypothesized that for fluids having relatively small viscosity, the effect of
internal friction in the fluid is significant only in a narrow region surrounding solid
boundaries or bodies over which the fluid flows.

 Thus, close to the body is the boundary layer where shear stresses exert an
increasingly larger effect on the fluid as one moves from free stream towards the
solid boundary.

 However, outside the boundary layer where the effect of the shear stresses on the
flow is small compared to values inside the boundary layer (since the velocity
gradient is negligible),---------

1. the fluid particles experience no vorticity and therefore,

2. the flow is similar to a potential flow.

 Hence, the surface at the boundary layer interface is a rather fictitious one,
that divides rotational and irrotational flow. Fig 1 shows Prandtl's model regarding
boundary layer flow.

 Hence with the exception of the immediate vicinity of the surface, the flow is frictionless
(inviscid) and the velocity is U (the potential velocity).

 In the region, very near to the surface (in the thin layer), there is friction in the flow which
signifies that the fluid is retarded until it adheres to the surface (no-slip condition).

Prof P.C.Swain Page 1

 Advance Fluid Mechanics 2016

 The transition of the mainstream velocity from zero at the surface (with respect to the
surface) to full magnitude takes place across the boundary layer.

About the boundary layer

 Boundary layer thickness is which is a function of the coordinate direction x .

 The thickness is considered to be very small compared to the characteristic
length L of the domain.
 In the normal direction, within this thin layer, the gradient is very large
compared to the gradient in the flow direction .

Now we take up the Navier-Stokes equations for : steady, two dimensional, laminar,
incompressible flows.

Considering the Navier-Stokes equations together with the equation of continuity, the following
dimensional form is obtained.

(1)

(2)

(3)

Fig 1 Boundary layer and Free Stream for Flow Over a flat plate

 u - velocity component along x direction.

 v - velocity component along y direction
 p - static pressure
 ρ - density.

Prof P.C.Swain Page 2

 Advance Fluid Mechanics 2016

 μ - dynamic viscosity of the fluid

 The equations are now non-dimensionalised.

 The length and the velocity scales are chosen as L and respectively.
 The non-dimensional variables are:

where is the dimensional free stream velocity and the pressure is non-
dimensionalised by twice the dynamic pressure .

Using these non-dimensional variables, the Eqs (1) to (3) become

(4)

(5)

(6)

where the Reynolds number,

Order of Magnitude Analysis

 Let us examine what happens to the u velocity as we go across the boundary layer.
At the wall the u velocity is zero [ with respect to the wall and absolute zero for a
stationary wall (which is normally implied if not stated otherwise)]. The value of u on
the inviscid side, that is on the free stream side beyond the boundary layer is U.
For the case of external flow over a flat plate, this U is equal to .
 Based on the above, we can identify the following scales for the boundary layer
variables:

Prof P.C.Swain Page 3

 Advance Fluid Mechanics 2016

Variable Dimensional scale Non-dimensional scale


The symbol describes a value much smaller than 1.
 Now we analyse equations 4 - .6, and look at the order of magnitude of each individual
term

Eq 6 - the continuity equation

One general rule of incompressible fluid mechanics is that we are not allowed to drop any
term from the continuity equation.

 From the scales of boundary layer variables, the derivative is of

the order 1.
 The second term in the continuity equation should also be of
the order 1.The reason being has to be of the order because
becomes at its maximum.

Eq 4 - x direction momentum equation

 Inertia terms are of the order 1.

 is of the order 1
 is of the order .

However after multiplication with 1/Re, the sum of the two second order derivatives should
produce at least one term which is of the same order of magnitude as the inertia terms. This
is possible only if the Reynolds number (Re) is of the order of .

 It follows from that will not exceed the order of 1 so as to be in

balance with the remaining term.

 Finally, Eqs (4), (5) and (6) can be rewritten as

(7)

Prof P.C.Swain Page 4

 Advance Fluid Mechanics 2016

(8)

As a consequence of the order of magnitude analysis, can be dropped from

the x direction momentum equation, because on multiplication with it assumes the
smallest order of magnitude.

Eq 5 - y direction momentum equation.

 All the terms of this equation are of a smaller magnitude than those of Eq. (.4).
 This equation can only be balanced if is of the same order of magnitude as
other terms.
 Thus they momentum equation reduces to

(8)

 This means that the pressure across the boundary layer does not change.
The pressure is impressed on the boundary layer, and its value is determined by
hydrodynamic considerations.
 This also implies that the pressure p is only a function of x. The pressure forces on a
body are solely determined by the inviscid flow outside the boundary layer.
 The application of Eq. (28.4) at the outer edge of boundary layer gives

(9)

In dimensional form, this can be written as

Prof P.C.Swain Page 5

 Advance Fluid Mechanics 2016

(10)

On integrating Eq ( 28.8b) the well known Bernoulli's equation is obtained

(11)
a constant

 Finally, it can be said that by the order of magnitude analysis, the Navier-Stokes
equations are simplified into equations given below.

(12)

(13)

(14)


 These are known as Prandtl's boundary-layer equations.

The available boundary conditions are:

Solid surface

or (15)

Outer edge of boundary-layer

Prof P.C.Swain Page 6

 Advance Fluid Mechanics 2016

or (16)

 The unknown pressure p in the x-momentum equation can be determined from

Bernoulli's Eq. (28.9), if the inviscid velocity distribution U(x) is also known.

We solve the Prandtl boundary layer equations for and with U obtained from
the outer inviscid flow analysis. The equations are solved by commencing at the leading edge of
the body and moving downstream to the desired location

 it allows the no-slip boundary condition to be satisfied which constitutes a significant

improvement over the potential flow analysis while solving real fluid flow problems.
 The Prandtl boundary layer equations are thus a simplification of the Navier-Stokes
equations.

Boundary Layer Coordinates

 The boundary layer equations derived are in Cartesian coordinates.

 The Velocity components u and v represent x and y direction velocities respectively.
 For objects with small curvature, these equations can be used with -
 x coordinate : streamwise direction
 y coordinate : normal component
 They are called Boundary Layer Coordinates.

Application of Boundary Layer Theory

 The Boundary-Layer Theory is not valid beyond the point of separation.

 At the point of separation, boundary layer thickness becomes quite large for the thin
layer approximation to be valid.
 It is important to note that boundary layer theory can be used to locate the point of
seperation itself.
 In applying the boundary layer theory although U is the free-stream velocity at the outer
edge of the boundary layer, it is interpreted as the fluid velocity at the wall calculated
from inviscid flow considerations ( known as Potential Wall Velocity)
 Mathematically, application of the boundary - layer theory converts the character of
governing Navier-Stroke equations from elliptic to parabolic
 This allows the marching in flow direction, as the solution at any location is independent
of the conditions farther downstream.

Blasius Flow Over A Flat Plate

 The classical problem considered by H. Blasius was

1. Two-dimensional, steady, incompressible flow over a flat plate at zero angle of
incidence with respect to the uniform stream of velocity .

Prof P.C.Swain Page 7

 Advance Fluid Mechanics 2016

2. The fluid extends to infinity in all directions from the plate.

The physical problem is already illustrated in Fig. 1

 Blasius wanted to determine

(a) the velocity field solely within the boundary layer,
(b) the boundary layer thickness ,
(c) the shear stress distribution on the plate, and
(d) the drag force on the plate.
 The Prandtl boundary layer equations in the case under consideration are

(15)

The boundary conditions are

(16)

 Note that the substitution of the term in the original boundary layer momentum

equation in terms of the free stream velocity produces which is equal to zero.
 Hence the governing Eq. (15) does not contain any pressure-gradient term.
 However, the characteristic parameters of this problem are that
is,
 This relation has five variables .
 It involves two dimensions, length and time.
 Thus it can be reduced to a dimensionless relation in terms of (5-2) =3 quantities
( Buckingham Pi Theorem)
 Thus a similarity variables can be used to find the solution
 Such flow fields are called self-similar flow field .

Law of Similarity for Boundary Layer Flows

 It states that the u component of velocity with two velocity profiles of u(x,y) at

Prof P.C.Swain Page 8

 Advance Fluid Mechanics 2016

different x locations differ only by scale factors in u and y .

 Therefore, the velocity profiles u(x,y) at all values of x can be made congruent
if they are plotted in coordinates which have been made dimensionless with
reference to the scale factors.
 The local free stream velocity U(x) at section x is an obvious scale factor for u,
because the dimensionless u(x) varies between zero and unity with y at all
sections.
 The scale factor for y , denoted by g(x) , is proportional to the local boundary
layer thickness so that y itself varies between zero and unity.
 Velocity at two arbitrary x locations, namely x1 and x2 should satisfy the
equation

(17)

 Now, for Blasius flow, it is possible to identify g(x) with the boundary layers
thickness δ we know

Thus in terms of x we get

i.e.,

(18)

where
or more precisely,

Prof P.C.Swain Page 9

 Advance Fluid Mechanics 2016

(19)

The stream function can now be obtained in terms of the velocity components as

Or

(20)

where D is a constant. Also and the constant of integration is zero if

the stream function at the solid surface is set equal to zero.

Now, the velocity components and their derivatives are:

or

(21)

(22)

Prof P.C.Swain Page 10

 Advance Fluid Mechanics 2016

(23)

(24)

 Substituting (28.2) into (28.15), we have

or,

(25)
where

and

This is known as Blasius Equation .

Prof P.C.Swain Page 11

 Advance Fluid Mechanics 2016

 The boundary conditions as in Eg. (28.16), in combination with Eg. (28.21a) and
(28.21b) become

at , therefore

at therefore (26)

Equation (22) is a third order nonlinear differential equation .

 Blasius obtained the solution of this equation in the form of series expansion through
analytical techniques
 We shall not discuss this technique. However, we shall discuss a numerical technique to
solve the aforesaid equation which can be understood rather easily.
 Note that the equation for does not contain .
 Boundary conditions at and merge into the
condition . This is the key feature of similarity solution.
 We can rewrite Eq. (28.22) as three first order differential equations in the following way

(27)

(28)

(29)

 Let us next consider the boundary conditions.

1. The condition remains valid.

2. The condition means that .
3. The condition gives us .

Note that the equations for f and G have initial values. However, the value for H(0) is not
known. Hence, we do not have a usual initial-value problem.

Shooting Technique

We handle this problem as an initial-value problem by choosing values of and solving by

Prof P.C.Swain Page 12

 Advance Fluid Mechanics 2016

numerical methods , and .

In general, the condition will not be satisfied for the function arising from the
numerical solution.
We then choose other initial values of so that eventually we find an which results
in .
This method is called the shooting technique .

 In Eq. (28.24), the primes refer to differentiation wrt. the similarity variable . The
integration steps following Runge-Kutta method are given below.

(30)

(31)

(32)

 One moves from to . A fourth order accuracy is preserved if h is

constant along the integration path, that is, for all values of n . The values
of k, l and m are as follows.
 For generality let the system of governing equations be

Prof P.C.Swain Page 13

 Advance Fluid Mechanics 2016

In a similar way K3, l3, m3 and k4, l4, m4 mare calculated following standard formulae for the
Runge-Kutta integration. For example, K3 is given by

The functions F1, F2and F3 are G, H

, - f H / 2 respectively. Then at a distance from the wall, we have

(33)

(34)

(35)

(36)

 As it has been mentioned earlier is unknown. It must be determined

such that the condition is satisfied.

The condition at infinity is usually approximated at a finite value of (around ). The

process of obtaining accurately involves iteration and may be calculated using the procedure
described below.

 For this purpose, consider Fig. 28.2(a) where the solutions of versus for two
different values of are plotted.
The values of are estimated from the curves and are plotted in Fig. 28.2(b).
 The value of now can be calculated by finding the value at which the line 1-
2 crosses the line By using similar triangles, it can be said

that . By solving this, we get .

 Next we repeat the same calculation as above by using and the better of the two

initial values of . Thus we get another improved value . This process may

continue, that is, we use and as a pair of values to find more improved

Prof P.C.Swain Page 14

 Advance Fluid Mechanics 2016

values for , and so forth. The better guess for H (0) can also be obtained by using
the Newton Raphson Method. It should be always kept in mind that for each value
of , the curve versus is to be examined to get the proper value of
.
 The functions and are plotted in Fig. 28.3.The velocity
components, u and v inside the boundary layer can be computed from Eqs (28.21a) and
(28.21b) respectively.
 A sample computer program in FORTRAN follows in order to explain the solution
procedure in greater detail. The program uses Runge Kutta integration together with the
Newton Raphson method

Download the program

Fig 2 Correcting the initial guess for H(O)

Prof P.C.Swain Page 15

 Advance Fluid Mechanics 2016

Fig 3 f, G and H distribution in the boundary layer

 Measurements to test the accuracy of theoretical results were carried out by many
scientists. In his experiments, J. Nikuradse, found excellent agreement with the
theoretical results with respect to velocity distribution within the boundary layer
of a stream of air on a flat plate.
 In the next slide we'll see some values of the velocity profile shape
and in tabular format.

Prof P.C.Swain Page 16

 Advanced Fluid Mechanics 2016

Lecture Note 2

N – S Equations
Navier-Strokes Equation

 Generalized equations of motion of a real flow named after the inventors CLMH Navier
and GG Stokes are derived from the Newton's second law
 Newton's second law states that the product of mass and acceleration is equal to sum
of the external forces acting on a body.
 External forces are of two kinds-
 one acts throughout the mass of the body ----- body force ( gravitational
force, electromagnetic force)
 another acts on the boundary----- surface force (pressure and frictional
force).

Objective - We shall consider a differential fluid element in the flow field (Fig.1). Evaluate the
surface forces acting on the boundary of the rectangular parallelepiped shown below.

Fig.1 Definition of the components of stress and their locations in a differential fluid element

Prof P.C.Swain Page 1

 Advanced Fluid Mechanics 2016
 Let the body force per unit mass be

(1)

and surface force per unit volume be

(2)

 Consider surface force on the surface AEHD, per unit area,

[Here second subscript x denotes that the surface force is evaluated for the surface whose
outward normal is the x axis]

 Surface force on the surface BFGC per unit area is

 Net force on the body due to imbalance of surface forces on the above two surfaces is

(3)
(since area of faces AEHD and BFGC is dydz)

 Total force on the body due to net surface forces on all six surfaces is

(4)

 And hence, the resultant surface force dF, per unit volume, is

(5)
(since Volume= dx dy dz)

 The quantities , and are vectors which can be resolved into normal stresses
denoted by and shearing stresses denoted by as

(6)

Prof P.C.Swain Page 2

 Advanced Fluid Mechanics 2016

The stress system has nine scalar quantities. These nine quantities form a stress tensor.

Nine Scalar Quantities of Stress System - Stress Tensor

The set of nine components of stress tensor can be described as

(7)

 The stress tensor is symmetric,

 This means that two shearing stresses with subscripts which differ only in their sequence
are equal. For example

 Considering the equation of motion for instantaneous rotation of the fluid element (Fig.
24.1) about y axis, we can write

where =dxdydz is the volume of the element, is the angular acceleration

is the moment of inertia of the element about y-axis

 Since is proportional to fifth power of the linear dimensions and is proportional

to the third power of the linear dimensions, the left hand side of the above equation and
the second term on the right hand side vanishes faster than the first term on the right hand
side on contracting the element to a point.
 Hence, the result is

From the similar considerations about other two remaining axes, we can write

Prof P.C.Swain Page 3

 Advanced Fluid Mechanics 2016

which has already been observed in Eqs (24.2a), (24.2b) and (24.2c) earlier.

 Invoking these conditions into Eq. (24.12), the stress tensor becomes

(8)

 Combining Eqs (24.10), (24.11) and (24.13), the resultant surface force per unit volume
becomes

(9)

 As per the velocity field,

(10)

By Newton's law of motion applied to the differential element, we can write

or,

Substituting Eqs (24.15), (24.14) and (24.6) into the above expression, we obtain

(11)

Prof P.C.Swain Page 4

 Advanced Fluid Mechanics 2016

(12)

(13)

Since

Similarly others follow.

 So we can express , and in terms of field derivatives,

(14)

(15)

(16)

 These differential equations are known as Navier-Stokes equations.

 At this juncture, discuss the equation of continuity as well, which has a general
(conservative) form

(17)

 In case of incompressible flow ρ = constant. Therefore, equation of continuity for

incompressible flow becomes

Prof P.C.Swain Page 5

 Advanced Fluid Mechanics 2016

(18)

 Invoking Eq. (24.19) into Eqs (24.17a), (24.17b) and (24.17c), we get

Similarly others follow

Thus,

(19)

(20)

(21)

Vector Notation & derivation in Cylindrical Coordinates - Navier-Stokes equation

 Using, vector notation to write Navier-Stokes and continuity equations for

incompressible flow we have

(22)

And

(23)

 we have four unknown quantities, u, v, w and p ,

 we also have four equations, - equations of motion in three directions and the
continuity equation.
 In principle, these equations are solvable but to date generalized solution is not
available due to the complex nature of the set of these equations.

Prof P.C.Swain Page 6

 Advanced Fluid Mechanics 2016
 The highest order terms, which come from the viscous forces, are linear and of
second order
 The first order convective terms are non-linear and hence, the set is termed as quasi-
linear.
 Navier-Stokes equations in cylindrical coordinate (Fig. 24.2) are useful in solving
many problems. If , and denote the velocity components along the radial,
cross-radial and axial directions respectively, then for the case of incompressible
flow, Eqs (24.21) and (24.22) lead to the following system of equations:

FIG 2 Cylindrical polar coordinate and the velocity components

(24)

Prof P.C.Swain Page 7

 Advanced Fluid Mechanics 2016

(25)

(26)

(27)

A general way of deriving the Navier-Stokes equations from the basic laws of physics.

 Consider a general flow field as represented in Fig. 25.1.

 Imagine a closed control volume, within the flow field. The control volume is fixed
in space and the fluid is moving through it. The control volume occupies reasonably large
finite region of the flow field.
 A control surface , A0 is defined as the surface which bounds the volume .
 According to Reynolds transport theorem, "The rate of change of momentum for
a system equals the sum of the rate of change of momentum inside the control
volume and the rate of efflux of momentum across the control surface".
 The rate of change of momentum for a system (in our case, the control volume boundary
and the system boundary are same) is equal to the net external force acting on it.

Now, we shall transform these statements into equation by accounting for each term,

Prof P.C.Swain Page 8

 Advanced Fluid Mechanics 2016
FIG 25.1 Finite control volume fixed in space with the fluid moving through it

 Rate of change of momentum inside the control volume

(28)
(since t is independent of space variable)

 Rate of efflux of momentum through control surface

(29)

 Surface force acting on the control volume

``

(30)

 Body force acting on the control volume

(31)

in Eq. (25.4) is the body force per unit mass.

Prof P.C.Swain Page 9

 Advanced Fluid Mechanics 2016
 Finally, we get,

or

or,

(32)
or

We know that is the general form of mass conservation equation (popularly

known as the continuity equation), valid for both compressible and incompressible flows.

 Invoking this relationship in Eq. (25.5), we obtain

(33)
or

 Equation (25.6) is referred to as Cauchy's equation of motion . In this equation, is

the stress tensor,
 After having substituted we get

(34)

(35)
From Stokes's hypothesis we get,

Prof P.C.Swain Page 10

 Advanced Fluid Mechanics 2016
Invoking above two relationships into Eq.( 25.6) we get

(36)

This is the most general form of Navier-Stokes equation.

Exact Solutions Of Navier-Stokes Equations

Consider a class of flow termed as parallel flow in which only one velocity term is
nontrivial and all the fluid particles move in one direction only.

 We choose to be the direction along which all fluid particles travel ,

i.e. . Invoking this in continuity equation, we get

which means

 Now. Navier-Stokes equations for incompressible flow become

So, we obtain

which means

Prof P.C.Swain Page 11

 Advanced Fluid Mechanics 2016

(37)
and

Parallel Flow in a Straight Channel

Consider steady flow between two infinitely broad parallel plates as shown in Fig. 25.2.

Flow is independent of any variation in z direction, hence, z dependence is gotten rid of and Eq.
(25.11) becomes

FIG 25.2 Parallel flow in a straight channel

(38)

The boundary conditions are at y = b, u = 0; and y = -b, u = O.

 From Eq. (25.12), we can write

or

 Applying the boundary conditions, the constants are evaluated as

and

Prof P.C.Swain Page 12

 Advanced Fluid Mechanics 2016
So, the solution is

(39)

which implies that the velocity profile is parabolic.

Average Velocity and Maximum Velocity

 To establish the relationship between the maximum velocity and average velocity in the
channel, we analyze as follows

At y = 0, ; this yields

(40)

On the other hand, the average velocity,

or

(41)
Finally,

(42)
So, or

 The shearing stress at the wall for the parallel flow in a channel can be determined from
the velocity gradient as

Prof P.C.Swain Page 13

 Advanced Fluid Mechanics 2016
Since the upper plate is a "minus y surface", a negative stress acts in the positive x direction, i.e.
to the right.

 The local friction coefficient, Cf is defined by

(43)

where is the Reynolds number of flow based on average velocity and the channel
height (2b).

 Experiments show that Eq. (25.14d) is valid in the laminar regime of the channel flow.
 The maximum Reynolds number value corresponding to fully developed laminar flow,
for which a stable motion will persist, is 2300.
 In a reasonably careful experiment, laminar flow can be observed up to even Re =
10,000.
 But the value below which the flow will always remain laminar, i.e. the critical value of
Re is 2300.

Prof P.C.Swain Page 14

 Advanced Fluid Mechanics 2016

References:

1. Wand D.J., and Harleman D.R. (91964) “Fluid Dynamics”, Addison Wesley.
2. Schlichting, H.: (1976) “Boundary Layer theory”, International Text –
Butterworth
3. Lamb, H. (1945) “Hydrodynamics”, International Text – Butterworth
4. Lamb, H.R. (1945) “Hydrodynamics”, Rover Publications
5. Rouse, H. (1957), “Advanced Fluid Mechanics”, John Wiley & Sons, N
York
6. White, F.M. (1980) “Viscous Fluid Flow”, McGraw Hill Pub. Co, N York
7. Yalin, M.S.(1971), “Theory of Hydraulic Models”, McMillan Co., 1971.
8. Mohanty A.K. (1994), “Fluid Mechanics”, Prentice Hall of India, N Delhi

Prof P.C.Swain