Professional Documents
Culture Documents
Modeling and Analysis of Modern Fluid Problems Zhang Download PDF Chapter
Modeling and Analysis of Modern Fluid Problems Zhang Download PDF Chapter
problems Zhang
Visit to download the full and correct content document:
https://ebookmass.com/product/modeling-and-analysis-of-modern-fluid-problems-zha
ng/
Mathematics in Science and Engineering
Liancun Zheng
School of Mathematics and Physics,
University of Science and Technology Beijing,
Beijing, China
Xinxin Zhang
School of Energy and Environmental Engineering,
University of Science and Technology Beijing,
Beijing, China
Series Editor
Goong Chen
Academic Press is an imprint of Elsevier
125 London Wall, London EC2Y 5AS, United Kingdom
525 B Street, Suite 1800, San Diego, CA 92101-4495, United States
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
Copyright © 2017 Elsevier Inc. All rights reserved.
Practitioners and researchers must always rely on their own experience and knowledge
in evaluating and using any information, methods, compounds, or experiments
described herein. In using such information or methods they should be mindful of their
own safety and the safety of others, including parties for whom they have a professional
responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or
editors, assume any liability for any injury and/or damage to persons or property as a
matter of products liability, negligence or otherwise, or from any use or operation of
any methods, products, instructions, or ideas contained in the material herein.
ISBN: 978-0-12-811753-8
ISSN: 0076-5392
xi
xii
Liancun Zheng
School of Mathematics and Physics,
University of Science and Technology Beijing,
Beijing, China
Xinxin Zhang
School of Energy and Environmental Engineering,
University of Science and Technology Beijing,
Beijing, China
Chapter 1
Introduction
ABSTRACT
Nature is full of many nonlinear and random phenomena. The study of nonlinear
phenomena is a matter of natural sciences, engineering, and even social economic
problems. Many scientists and engineers spend a lot of time and effort to study
mathematical modeling and solving methods for these problems. This chapter presents
an introduction to modeling, the development of analytical methods for modern fluid
problems, and the outline of this book.
when x is the continuous point, the series converges to f(x); when x is the
discontinuity point, the series converges to
1
½ f ðx Þ þ f ðxþ Þ: (1.5)
2
There is a relationship between coefficients a0, a1, b1, ., and the function
of f(x)
8
>
>
>
> Z l
>
> 1 np
>
> a ¼ f ðxÞcos xdx ðn ¼ 0; 1; 2; .Þ;
< n l l
l
Z (1.6)
>
> l
>
> bn ¼
1 np
f ðxÞsin xdx ðn ¼ 1; 2; 3; .Þ:
>
>
>
> l l l
:
The coefficients determined by Eq. (1.6) are known as the Fourier coefficient
of the function f(x). The triangle series defined by the Fourier coefficient is
called the Fourier series.
It can be seen that the essence of the Taylor series and the Fourier series is
to select a suitable set of base functions (BFs)
40 ðxÞ; 41 ðxÞ; 42 ðxÞ; .; 4n ðxÞ; .; (1.7)
to express the unknown function f(x) as:
f ðxÞ ¼ a0 40 ðxÞ þ a1 41 ðxÞ þ a2 42 ðxÞ þ /; (1.8)
The coefficients are determined by the selected basis functions and the
given function f(x) itself.
When the BF is selected by 1, x, x2, ., xn,., we obtain the Taylor series.
If the BF is selected by 1; cos xl; sin xl; cos 2xl; sin 2xl; .; we obtain the
Fourier series.
It is obvious that the expressions of the different approximate series will be
obtained when we select different sets of BFs. Therefore, the essence of the
analytical approach to deal with the nonlinear problem is based on the
following ideas:
1. Select the appropriate set of linear independent basis functions
f4i ðxÞgN
i¼0 ˛C½a; b;
2. Express unknown functions 4(x) as the combination of the BFs
a þ bx
; (1.11)
c þ dx
The formula is bounded when x approaches infinity. When we approximate
f(x) using the first three coefficients of terms of the Taylor series, we can obtain
the following expression:
1 þ ð7=4Þx 1 5 25 3 125 4
¼ 1 þ x x2 þ x x þ /:
1 þ ð5=4Þx 2 8 32 128
In view of this formula, when x ¼ N, it has a value of 1.4. This result is
better than any approximation results obtained previously. Similarly, we can
compute a new approximate result as:
1 þ ð13=4Þx þ ð41=16Þx2 41
/ ¼ 1:413793103:
1 þ ð11=4Þx þ ð29=16Þx2 29
pffiffiffi
This new result is close to 2 z 1:414213562. Furthermore, if we continue
to use this method to calculate the approximation, we find that the more
coefficients are used, the better convergence result are obtained, which are
shown as:
In view of Eq. (1.16), multiplying Eq. (1.14) by QM(x), we can obtain the
following linear systems of coefficients:
a0 ¼ p0 ;
a1 þ a0 q1 ¼ p1 ;
a2 þ a1 q1 þ a 0 q2 ¼ p2 ;
« «
(1.17)
aL þ aL1 q1 þ / þ a0 qL ¼ pL ;
aLþ1 þ aL q 1 þ / þ aLMþ1 qM ¼ 0;
« «
aLþM þ aLþM1 q1 þ / þ a L qM ¼ 0;
and we may define
an h 0 if n < 0 and qj h 0 if j > M; qj h 0: (1.18)
Definition Padé approximants given here are different from the definition
of classical approximants in several ways. First, the expression of symbols is
the different in the classical definition:
½M;L ¼ ½L=M: (1.19)
However, some authors use the notation:
½L;M ¼ ½L=M: (1.20)
The formula employed here is used (Eq. 1.12) to avoid confusion. By
convention, L denotes the degree of numerator and M denotes the degree of
denominator and the following formula:
L þ M ¼ N; L M ¼ J; (1.21)
to express the sum and difference of those degrees of the numerator and the
denominator.
A important point in mathematics is that in the standardized condition, the
new definition is clearly different from the classical definition in the
normalization condition (Eq. 1.15). Frobenius (1881) and Padé (1892) required
only that QM(x) s 0, as illustrated in the following example:
AðxÞ ¼ 1 þ x2 þ /; (1.22)
For L ¼ M ¼ 1, it is easy to see that
P1 ðxÞ ¼ Q1 ðxÞ ¼ x; P1 ðxÞ=Q1 ðxÞ ¼ 1; (1.23)
It satisfies
QM ðxÞAðxÞ PL ðxÞ ¼ Oðx Nþ1 Þ; (1.24)
8 Modeling and Analysis of Modern Fluid Problems
However, it does not satisfy Eq. (1.14). In fact, for this series, the
expression [1/1] does not exist in the new definition.
dnþ1 ðεÞ
ðaÞ lim dn ðεÞ ¼ 0; ðbÞ lim ¼ 0 ðn ¼ 1; 2; /Þ: (1.36)
ε/N ε/N dn ðεÞ
Introduction Chapter j 1 11
X
N
uðx;εÞ ¼ u0 ðxÞ þ dn ðεÞun ðxÞ þ ZN ðx;εÞ; (1.37)
n¼1
3. Substitute the expanded formula into the original problem and compare the
coefficients of dn(ε); one can obtain the recursive equation and boundary
conditions as
L0 ðui Þ ¼ Hi ½u0 ; u1 ; /; uiþ1 ; x; B0; j ðui Þ ¼ Fj ½u0 ; u1 ; /; ui1 ; x; (1.38)
method, average method, and so on, but it is difficult to choose the form for
multiple scales variables. It is necessary to have some practical knowledge and
experience when using the method of multiple scales.
While pursuing a doctorate at the University of Science and Technology
Beijing (USTB), Professor Zheng’s doctoral student, Yan Zhang, proposed a
novel technique, i.e., the embedding-parameters perturbation method. The
basic idea of this method is by introducing special parameters transformation
for both independent and dependent variables, we can embed special small
parameters into nonlinear differential equations, which then can be solved by a
standard perturbation method. Thereafter, the solutions to the original
differential equations can be obtained by substituting the asymptotic series
solutions into the reversed transformation. Moreover, this method can be
applied to solve other nonlinear problems.
equations and partial differential equations. However, there are some limita-
tions to the Adomian decomposition method, i.e., the asymptotic series given
by the method usually contain polynomials that are a combination of the basis
functions of power series for the approximation, because the convergence
region of the power series is small. Thus, in some case the uniform effective
solution of the whole region usually cannot be obtained.
In addition, the Adomian decomposition method does not allow the selection
of different basis functions, and the method cannot adapt the convergence region
and the convergence rate of asymptotic solutions.
changing from initial solution u0(t) to the solution u(t) of Eq. (1.46). In
topology, this continuous variation is called homotopy.
In brief, HAM is based on the topology homotopy concept, but it is
different from the traditional one. By introducing a nonzero auxiliary
parameter h and a nonzero auxiliary function H, one constructs a new
homotopy:
e q;Z; HÞ ¼ ð1 qÞL½fðt; q;Z; HÞ u0 ðtÞ qZHðtÞA½fðt; q;Z; HÞ;
Hðf;
(1.51)
This formula is more general than Eq. (1.48), because Eq. (1.48) is only a
special case of Eq. (1.51) when Z ¼ 1 and H(t) ¼ 1, i.e.,
e q; 1; 1Þ
Hðf; qÞ ¼ Hðf; (1.52)
Analogously, when q increases from 0 to 1, the function f(t; q, Z,H)
changes its value from the initial value u0(t) to the solution u(t) of the original
nonlinear problem u(t). The solution to equation
e
H½fðt; q; Z; HÞ ¼ 0: (1.53)
depends not only on the embed parameter q but also on auxiliary parameter Z
and auxiliary function H(t). When q ¼ 1, the solution still depends on the
auxiliary parameter Z and auxiliary function H(t).
The convergence radius of the obtained series approximate solution
depends on the two auxiliary parameters; the convergence rate and the
convergence region can be adjusted by the auxiliary parameter. HAM is
completely free from the assumption of small parameters, which can overcome
the limitation of the perturbation method:
1. The nonlinear problem is effective even if the problem is not contained in
any small (or large) parameter.
2. The convergence region and the convergence rate of the asymptotic series
solution can be conveniently adjusted.
3. It is free to choose different initial solutions to approximate the nonlinear
problem.
HAM has limitations, including the introduction of auxiliary operators,
auxiliary parameters, and auxiliary functions. So far, there is no rigorous
mathematical theory to guide the selection of auxiliary operators, auxiliary
parameters, or auxiliary functions.
Fourier integral transforms, etc. Integral transformations can change the dif-
ferential equations into simple algebraic equations. The primitive function of a
linear differential equation can be changed to corresponding algebraic oper-
ations. Therefore, integral transformations have been widely used.
The differential transformation method (DTM) was first proposed by
Jakui Zhao (1988) to solve differential equations arising in circuits. With this
method, Zhao successfully solved the linear ordinary differential equation of
the circuit and some initial value problems of nonlinear ordinary differential
equations.
DTM is based on the Taylor series expansion of an unknown function, but
it is different from the traditional method of Taylor series. The Taylor series
method needs to do differential operations for unknown functions. However,
DTM does not need to do differential operations where the computing formula
of differential transformation can be obtained directly according to the
unknown function by an algebraic operation, which can be an iterative formula
of the Taylor expansion coefficient. Compared with the traditional Taylor
expansion method, DTM greatly reduces the amount of calculations and
makes the Taylor expansion possible for some complex equations of unknown
functions.
Assume a function w(t) has kth-order derivatives for variable t; the dif-
ferential transform of function w(t) is defined as (Zhao, 1988)
dk qðtÞwðtÞ
WðkÞ ¼ MðkÞ ; k ¼ 0; 1; 2; 3; .; (1.54)
dtk t¼t0
where w(t) and W(k) are called the original function and differential transform
function, respectively. The inverse differential transform function of W(k) is
defined as
1 X N
WðkÞ
wðtÞ ¼ ðt t0 Þk ; (1.55)
qðtÞ 0 MðkÞk!
where w(t,x) and W(k,h) are the original function and differential transform
functions, respectively. The inverse differential transform function of W(k,h) is
defined as
1 XN
Wðk; hÞ
wðt; xÞ ¼ ðt t0 Þk ðx x0 Þh : (1.57)
qðtÞpðxÞ 0 MðkÞNðhÞh!k!
In the same way, we can define the differential transform of three or more
variables. The solutions obtained by DTM are in form of Taylor series, which
has obvious advantages relative to other analytical methods. Since the DTM
was put forward, scholars have given it much attention. Many types of
commonly used functions of differential transformation formulas have been
derived and applications have become more widely known (Chen and Ho,
1999). This method has gradually been applied to many linear and nonlinear
ordinary differential equations, partial differential equations, FDEs, integral
equations, etc. (Arikoglu, 2005; Arikoglu and Ozkol, 2009; Ayaz, 2004).
Although DTM has the advantages of high accuracy and a smaller amount
of calculation, this method also has some limitations. DTM seems hard to
solve for some differential equations on an unbounded domain because the
results obtained by DTM are usually valid in a small region. This is because
series solutions obtained by DTM are divergent when the variables of the
problems go to infinity. Moreover, the power series in isolation are not useful
for handling boundary value problems on an unbounded domain (Abazari and
Borhanifar, 2010; Chen and Ho, 1996).
Boyd (1997) proved that under normal conditions, the power series
solution of a single form cannot effectively express an unbounded region or
even part of the bounded region in a boundary value problem. The reason is
Introduction Chapter j 1 19
1970s, Professor Mandelbrot first proposed “fractal” theory, and pointed out
that nature and many science and technology problems have a lot of fractal
dimensions, that there is a self-similar phenomenon, and that fractional
Brownian motion and fractional calculus exist between the whole and the part.
As the dynamic basis of fractal geometry and fractal dimension, fractional
calculus theory and FDE theory have developed rapidly and have become a hot
research topic in the world.
Fractional calculus provides a good tool to describe physical memory and
heredity. Fractional order calculus has been applied to many fields such as
flabby, oscillation, stochastic diffusion theory and wave propagation, biolog-
ical materials, control and robotics, viscoelastic dynamics, and quantum me-
chanics. Those applications have also accelerated the development of the
theory of fractional calculus. Fractional order calculus theory and the theory of
chaos and dissipative structure is considered as the current theory of nonlinear
science (Li et al., 2008; Zhu, 2006).
(1.61)
Definition 1.3.2 (Caputo fractional order derivative)
8
>
> Z t
>
> 1
>
< Gðm aÞ ðt sÞma1 f m ðsÞds; m 1 < a < m
a
a Dt f ðtÞ ¼ : (1.62)
C a
> dm
>
>
>
: dtm f ðtÞ;
> a ¼ m˛N
(1.65)
RN
where G($) is Gamma function GðzÞ ¼ 0 et tz1 dt, Re(z) > 0, and in the
definition, integral lower limit a ¼ 0.
Some main properties of fractional derivatives are:
Property 1.3.1 0 Dat ðCf ðtÞÞ ¼ C$0 Dat f ðtÞ, a ˛ R, C are constants.
Property 1.3.2 0 Dεt ð f1 ðtÞ þ f2 ðtÞÞ ¼ 0 Dat f1 ðtÞ þ 0 Dat f2 ðtÞ, a ˛ R.
b ab
Property 1.3.3 0 Da
t 0 Dt ð f ðtÞÞ ¼ 0 Dt f ðtÞ, a > 0, b > 0.
Particular, 0 Dat 0 Da
t f ðtÞ ¼ f ðtÞ, a > 0,
b u Gð1 þ uÞ uþb
0 Dt t ¼ t ; b > 0; u > 1; (1.66)
Gð1 þ u þ bÞ
Gð1 þ uÞ ua
a u
0 Dt t ¼ t ; a > 0; u > 1: (1.67)
Gð1 þ u aÞ
Z þN
1 1
f ðxÞ ¼ F fFðkÞ; xg ¼ eikx FðkÞdk; x ˛ R; (1.69)
2p N
where F(k) is called the Fourier transformation of function f(x), and f(x) is
called the inverse Fourier transformation of F(k).
The Fourier transform formula of the ath-order differential operator is
Ff0 Dat f ðxÞ; kg ¼ ðiwÞa FðkÞ: (1.70)
The convolution of the Fourier transformation is defined as
Z þN Z þN
f ðxÞ gðxÞ ¼ f ðεÞgðx εÞdε ¼ f ðx εÞgðεÞdε; (1.71)
N N
Z cþiN
1
f ðtÞ ¼ L1 fFð pÞ; tg ¼ ept Fð pÞdt; C ¼ Rð pÞ > C0 ; (1.74)
2pi ciN
MISSION TO TANGIER.
In 1844 Mr. Hay went to England on leave, and visited also
Stockholm and Copenhagen. At this latter capital he met the ‘fair girl’
who was to be his future wife, as Leila had predicted. Whilst in
Stockholm, he was presented to King Oscar by our Minister, Mr.
Cartwright, and in the course of conversation with His Majesty about
Morocco, pointed out the advisability of abolishing the old
Convention between Morocco and Sweden, and Morocco and
Denmark, which stipulated that $25,000 (£5,000) should be paid
annually to the Sultan, in order that vessels under the flags of these
two nations should pass the Straits unmolested by Moorish cruisers;
these cruisers having virtually ceased to exist, though the
Convention remained in force.
A rupture of relations between France and Morocco was at this
time imminent, and Mr. Hay’s father, then Political Agent at Tangier,
had been sent, with the knowledge of the French Government, to the
city of Marákesh on a mission to endeavour to induce the Sultan to
accept the French demands. On hearing of this expedition Mr. Hay
wrote to Lord Aberdeen, who was then Secretary for Foreign Affairs,
to offer his services temporarily in Morocco. This offer was accepted.
That Mr. Hay, while at Constantinople, had gained the kindly
opinion of Sir Stratford and Lady Canning may be gathered from the
following letter written to him after his departure from Constantinople,
when Lady Canning learnt that he had been sent to Tangier. The
note was accompanied by the gift of a beautiful cushion in Turkish
embroidery.
You must not leave Constantinople, my dear Mr. Hay, without some little
memento from me to remind you in future days of our life spent together on the
Bosphorus in which, though it may have had some cloudy moments, I hope the
bright ones have preponderated and will alone be remembered by you. We shall
miss you sadly; for your labours have not been thrown away on Sir Stratford, and
you have helped to keep us all in good humour with our neighbours, and for all this
I thank you much. Let us hear of you often, and believe that we shall feel interest
in all that concerns you.
Yours very sincerely,
E. C. Canning.
’Tis a sad thing that all folks in Europe, my masters in Downing Street may not
be excepted, have hardly any just conception of the difficulties of my position. It
would take a volume—not small—to relate the bother and the tricks and bad faith
with which I have had to contend—and as to going fast, as Mr. Bulwer has
everlastingly urged, who among mortal men can make Moors go fast, nay, nor
hardly move at all—in the straight path of honour and sound policy? . . .
Alas! I know not what to think. I had hoped the French would have waited until
my report reached Tangier or myself arrived there and told them all. So they are
now preparing to cast fire and the sword on this unhappy country of ignorant
barbarians.
The Moors are mere children, vain children; obstinate, through a shocking
bigotry and ignorance scarcely credible. They have, I believe, had at least two
collisions with the French on their frontier; but all their acts of folly were, I am
certain, without authority. Alas, again, for the poor Sultan; he cannot manage his
own people! If the war do burst forth here, when shall it end? There will be an
internal revolution forthwith, I am almost sure! And drivellers in pomposity and self-
sufficiency would ever publish that all was well.
The elder Mr. Hay did not long survive the effects of the journey,
with all its worry and vexation; but succumbed shortly after his
return, to low fever and other complaints. During his illness, which
lasted several months, Mr. John Hay was directed by Lord Aberdeen
to take charge of political affairs in Morocco, whilst Mr. H. Murray, the
Consul, conducted the consular duties.
The crisis was one of considerable importance. In addition to the
internal difficulties of Morocco, questions with foreign Powers
embarrassed the Sultan’s Government. Denmark and Sweden had
sent squadrons in this year to Moorish waters, demanding the
abrogation of the treaty referred to in Mr. Hay’s audience of King
Oscar.
The Spanish Government had also a question pending with
Morocco regarding the neutral ground and frontier of Ceuta; and, for
the settlement of this question, Sir Henry Bulwer, then H.B.M.’s
Minister at Madrid, had been appointed special Plenipotentiary.
In the following letter to his late chief at Constantinople Mr. Hay
gives an account of the state of affairs which he found on his arrival
at Tangier.