Professional Documents
Culture Documents
(Download PDF) Techniques of Functional Analysis For Differential and Integral Equations 1St Edition Paul Sacks Full Chapter PDF
(Download PDF) Techniques of Functional Analysis For Differential and Integral Equations 1St Edition Paul Sacks Full Chapter PDF
https://ebookmass.com/product/advanced-differential-
equations-1st-edition-youssef-raffoul/
https://ebookmass.com/product/schaums-outline-of-differential-
equations-fifth-edition-bronson/
https://ebookmass.com/product/transforms-and-partial-
differential-equations-t-veerarajan/
https://ebookmass.com/product/proper-orthogonal-decomposition-
methods-for-partial-differential-equations-chen/
Boundary value problems for systems of differential,
difference and fractional equations : positive
solutions 1st Edition Johnny Henderson
https://ebookmass.com/product/boundary-value-problems-for-
systems-of-differential-difference-and-fractional-equations-
positive-solutions-1st-edition-johnny-henderson/
https://ebookmass.com/product/mechanics-and-physics-of-
structured-media-asymptotic-and-integral-equations-methods-of-
leonid-filshtinsky-1st-edition-igor-andrianov/
https://ebookmass.com/product/integral-equations-for-real-life-
multiscale-electromagnetic-problems-electromagnetic-waves-
francesca-vipiana/
https://ebookmass.com/product/linear-algebra-and-partial-
differential-equations-t-veerarajan/
Techniques of Functional
Analysis for Differential and
Integral Equations
Mathematics in Science and
Engineering
Techniques of
Functional Analysis for
Differential and
Integral Equations
Paul Sacks
Department of Mathematics, Iowa State University,
Ames, IA, United States
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
No part of this publication may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, recording, or any information storage and
retrieval system, without permission in writing from the publisher. Details on how to seek
permission, further information about the Publisher’s permissions policies and our arrangements
with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency,
can be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the
Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience
broaden our understanding, changes in research methods, professional practices, or medical
treatment may become necessary.
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: 978-0-12-811426-1
Preface ix
2. Vector Spaces 29
2.1 Axioms of a Vector Space 29
2.2 linear Independence and Bases 31
2.3 linear Transformations of a Vector Space 32
2.4 Exercises 33
3. Metric Spaces 35
3.1 Axioms of a Metric Space 35
3.2 Topological Concepts 37
3.3 Functions on Metric Spaces and Continuity 40
3.4 Compactness and Optimization 42
3.5 Contraction Mapping Theorem 45
3.6 Exercises 48
vi Contents
4. Banach Spaces 51
4.1 Axioms of a Normed linear Space 51
4.2 Infinite Series 53
4.3 linear Operators and Functionals 54
4.4 Contraction Mappings in a Banach Space 56
4.5 Exercises 57
5. Hilbert Spaces 59
6. Distribution Spaces 75
6.1 The Space of Test Functions 76
6.2 T he Space of Distributions 76
6.3 Algebra and Calculus With Distributi ons 80
6.3.1 Multiplication of Distributions 80
6.3.2 Convergence of Distributions 81
6.3.3 Derivative of a Distribution 83
6.4 Convolution and Distributions 88
6.5 Exercises 92
7. Fourier Analysis 95
7.1 Fourier Series in One Space Dimension 95
7.2 Alternative Forms of Fourier Series 100
7.3 More Abou t C onver gen ce of Fourier Series 101
7.4 The Fourier Transform on ]RN 104
7.5 Further Properties of the Fourier Transform 107
7.6 Fourier Series of Distributions 111
7.7 Fourier Transforms of Distributions 114
7.8 Exercises 119
Appendix 295
A.1 Lebesgue Measure and the Lebesgue Integral 295
A.2 Inequalities 298
A.3 Integration by Parts 301
A.4 Spherical Coordinates in JRN 302
Bibliography 305
Index 307
Preface
ix
x Preface
I would like to thank past and present friends and colleagues for many
helpful discussions and suggestions about the teaching of Applied Mathematics
at this level and points of detail in earlier drafts of this book: Tuncay Aktosun,
Jim Evans, Scott Hansen, Fritz Keinert, Howard Levine, Gary Lieberman,
Hailiang Liu, Tasos Matzavinos, Hien Nguyen, John Schotland, Mike Smiley,
Pablo Stinga and Jue Yan. All mistakes and shortcomings are of course my own
doing.
Paul Sacks
October 28, 2016
Chapter 1
where t0 is a given point in (a, b) and γ0 , . . . , γn−1 are given constants. Thus
we are prescribing the value of the solution and its derivatives up through order
n − 1 at the point t0 . The problem of solving Eq. (1.1.2) together with the initial
conditions (1.1.7) is called an initial value problem (IVP). It is a very important
fact that under fairly unrestrictive hypotheses a unique solution exists. In stating
conditions on f , we regard it as a function f = f (t, y1 , . . . , yn ) defined on some
domain in Rn+1 .
Some Basic Discussion of Differential and Integral Equations Chapter | 1 3
where up is any particular solution of Eq. (1.1.3), and u1 , . . . , un are any n linearly
independent solutions of the corresponding homogeneous equation Lu = 0. Any
such set of functions u1 , . . . , un is also called a fundamental set for Lu = 0.
Example 1.3. If Lu = u + u then by direct substitution we see that u1 (t) =
sin t, u2 (t) = cos t are solutions, and they are clearly linearly independent. Thus
{sin t, cos t} is a fundamental set for Lu = 0 and u(t) = c1 sin t + c2 cos t is
the general solution of Lu = 0. For the inhomogeneous ODE u + u = et one
may check that up (t) = 12 et is a particular solution, so the general solution is
u(t) = c1 sin t + c2 cos t + 12 et . 2
define the so-called integrating factor ρ(t) = eP(t) where P is any function
satisfying P = p. Multiplying the equation through by ρ we then get the
equivalent equation
(ρu) = ρq (1.1.15)
so if we pick Q such that Q = ρq, the general solution may be given as
Q(t) + C
u(t) = (1.1.16)
ρ(t)
● For the linear homogeneous constant coefficient ODE
n
Lu = aj u(j) = 0 (1.1.17)
j=0
if we look for solutions in the form u(t) = eλt then by direct substitution
we find that u is a solution provided λ is a root of the corresponding
characteristic polynomial
n
P(λ) = aj λj (1.1.18)
j=0
for some constants a0 , . . . , an . In this case we look for solutions in the form
u(t) = (t−t0 )λ with λ to be found. Substituting into Eq. (1.1.19) we will find
again an nth order polynomial whose roots determine the possible values of
λ. The interested reader may refer to any standard undergraduate level ODE
book for the additional considerations which arise in the case of complex or
repeated roots.
Tu(x) = K(x, y)u(y) dy (1.2.20)
Here the function K is called the kernel of the integral operator T, which is
linear since Eq. (1.1.6) obviously holds.
A class of associated integral equations is then
λu(x) − K(x, y)u(y) dy = f (x) x ∈ (1.2.21)
for some scalar λ and given function f in some appropriate class. If λ = 0
then Eq. (1.2.21) is said to be a first kind integral equation, otherwise it is
second kind. Let us consider some simple examples which may be studied by
elementary means.
Example 1.6. Let = (0, 1) ⊂ R and K(x, y) ≡ 1. The corresponding first
kind integral equation is therefore
1
− u(y) dy = f (x) 0 < x < 1 (1.2.22)
0
For simplicity here we will assume that f is a continuous function. The left-
hand side is independent of x, thus a solution can exist only if f (x) is a constant
function. When f is constant, on the other hand, infinitely many solutions will
exist, since we just need to find any u with the given definite integral.
For the corresponding second kind equation,
1
λu(x) − u(y) dy = f (x) (1.2.23)
0
a solution, if one exists, must have the specific form u(x) = (f (x) + C)/λ
for some constant C. Substituting into the equation then gives, after obvious
algebra, that
1
C(λ − 1) = f (y) dy (1.2.24)
0
Thus, for any continuous function f and λ = 0, 1, there exists a unique solution
of the integral equation, namely
1
f (x) f (y) dy
u(x) = + 0 (1.2.25)
λ λ(λ − 1)
In the remaining case that λ = 1, it is immediate from Eq. (1.2.24) that a solution
1
can exist only if 0 f (y) dy = 0, in which case u(x) = f (x) + C is a solution for
any choice of C. 2
This very simple example already exhibits features which turn out to be
common to a much larger class of integral equations of this general type.
These are
Some Basic Discussion of Differential and Integral Equations Chapter | 1 7
● The first kind integral equation will require much more restrictive conditions
on f in order for a solution to exist.
● For most λ = 0 the second kind integral equation has a unique solution for
any f .
● There may exist a few exceptional values of λ for which either existence or
uniqueness fails in the corresponding second kind equation.
All of these points will be elaborated and made precise in Chapter 12.
Example 1.7. Let = (0, 1) and
x
Tu(x) = u(y) dy (1.2.26)
0
corresponding to the kernel
1 y<x
K(x, y) = (1.2.27)
0 x≤y
The corresponding integral equation may then be written as
x
λu(x) − u(y) dy = f (x) (1.2.28)
0
This is the prototype of an integral operator of so-called Volterra type, see
Definition 1.1 below.
In the first kind case, λ = 0, we see that f (0) = 0 is a necessary condition
for solvability, in which case the solution is u(x) = −f (x), provided that f is
differentiable in some suitable sense. For λ = 0 we note that differentiation of
Eq. (1.2.28) with respect to x gives
1 f (x)
u − u= (1.2.29)
λ λ
This is an ODE of the type (1.1.14), and so may be solved by the method given
there. The result, after some obvious algebraic manipulation, is
ex/λ 1 x (x−y)/λ
u(x) = f (0) + e f (y) dy (1.2.30)
λ λ 0
Note, however, that by an integration by parts, this formula is seen to be
equivalent to
x
f (x) 1
u(x) = + 2 e(x−y)/λ f (y) dy (1.2.31)
λ λ 0
Concerning the two simple integral equations just discussed, there are again
some features which will turn out to be generally true.
● For the first kind equation, there are fewer restrictions on f needed for
solvability in the Volterra case (1.2.28) than in the non-Volterra case
(1.2.23).
● There are no exceptional values λ = 0 in the Volterra case, that is, a unique
solution exists for every λ = 0 and every continuous f .
Finally let us mention some of the more important ways in which integral
operators, or the corresponding integral equations, are classified:
Definition 1.1. The kernel K(x, y) is called
● symmetric if K(x, y) = K(y, x)
● Volterra type if N = 1 and K(x, y) = 0 for x > y or x < y
● convolution type if K(x,y) = F(x − y) for some function F
● Hilbert-Schmidt type if × |K(x, y)|2 dxdy < ∞
● singular if K(x, y) is unbounded on ×
Important examples of integral operators, some of which will receive much
more attention later in the book, are the Fourier transform
1
Tu(x) = e−ix·y u(y) dy (1.2.32)
(2π )N/2 RN
the Laplace transform
∞
Tu(x) = e−xy u(y) dy (1.2.33)
0
the Hilbert transform
∞
1 u(y)
Tu(x) = dy (1.2.34)
π −∞ x−y
and the Abel operator
x u(y)
Tu(x) = √ dy (1.2.35)
0 x−y
for some function f . This is then the general solution of the given PDE, which
we note contains an arbitrary function f . 2
Example 1.9. Next consider, again for N = 2, m = 1, the PDE
∂u ∂u
a +b =0 (1.3.44)
∂x1 ∂x2
where a, b are fixed constants, at least one of which is not zero. The equation
amounts precisely to the condition that u has directional derivative 0 in the
direction θ = a, b , so u is constant along any line parallel to θ. This in turn
leads to the conclusion that u(x1 , x2 ) = f (ax2 − bx1 ) for some arbitrary function
f , which at least for the moment would seem to need to be differentiable. 2
10 Techniques of Functional Analysis for Differential and Integral Equations
xux + uy = 1 (1.3.52)
The characteristics in this case are the curves x = set , y = t for fixed s, or
x = sey in nonparametric form. Note here that the solution is defined throughout
the x, y plane even though nothing in the preceding discussion guarantees that.
Since h has not been otherwise prescribed we may also regard Eq. (1.3.55)
as the general solution of Eq. (1.3.52), again containing one arbitrary
function. 2
The attentive reader may already realize that this procedure cannot work
in all cases, as is made clear by the following consideration: if c ≡ 0 and is
itself a characteristic curve, then the solution on would have to simultaneously
be equal to the given function h and to be constant, so that no solution can
exist except possibly in the case that h is a constant function. From another,
more general, point of view we must eliminate the parameters s, t by inverting
the relations x = x(s, t), y = y(s, t) to obtain s, t in terms of x, y, at least
near . According to the inverse function theorem this should require that the
Jacobian matrix
∂x ∂y
a(f (s), g(s)) b(f (s), g(s))
∂t ∂t = (1.3.56)
∂x
∂s
∂y
∂s
f (s) g (s)
t=0
direction at every point. We say that is noncharacteristic for the PDE (1.3.45)
when this condition holds.
The following precise theorem can be established, see, for example, Chapter
1 of [19], or Chapter 3 of [11]. The proof amounts to showing that the method
of constructing a solution just described can be made rigorous under the stated
assumptions.
Theorem 1.2. Let ⊂ R2 be a continuously differentiable curve, which is
noncharacteristic for Eq. (1.3.45), h a continuously differentiable function on
, and let a, b, c be continuously differentiable functions in a neighborhood of
. Then there exists a unique continuously differentiable function u(x, y) defined
in a neighborhood of which is a solution of Eq. (1.3.45).
The method of characteristics is capable of a considerable amount of
generalization, in particular to first order PDEs in any number of independent
variables, and to fully nonlinear first PDEs, see the references just mentioned
previously.
where A, B, C are real constants, not all zero. Consider introducing new coordi-
nates ξ , η by means of a linear change of variable
ξ = αx + βy η = γ x + δy (1.3.58)
with similar expressions for uxy and uyy . Substituting into Eq. (1.3.57) the
resulting PDE is
auξ ξ + buξ η + cuηη = 0 (1.3.61)
where
a = α 2 A + αβB + β 2 C (1.3.62)
b = 2αγ A + (αδ + βγ )B + 2βδC (1.3.63)
c = γ 2 A + γ δB + δ 2 C (1.3.64)
We now seek to make special choices of α, β, γ , δ to achieve as simple a form
as possible for the transformed PDE (1.3.61).
Suppose first that B2 − 4AC > 0, so that there exist two real and distinct
roots r1 , r2 of Ar2 + Br + C = 0. If α, β, γ , δ are chosen so that
α γ
= r1 = r2 (1.3.65)
β δ
then a = c = 0, and αδ − βγ = 0, so that the transformed PDE is simply, after
division by a constant, uξ η = 0. The general solution of this second order PDE
is easily obtained: uξ must be a function of ξ alone, so integrating with respect
to ξ and observing that the “constant of integration” could be any function of η,
we get
u(ξ , η) = F(ξ ) + G(η) (1.3.66)
for any differentiable functions F, G. Finally reverting to the original coordinate
system, the result is
u(x, y) = F(αx + βy) + G(γ x + δy) (1.3.67)
an expression for the general solution containing two arbitrary functions.
The lines αx + βy = C, γ x + δy = C are called the characteristics for
Eq. (1.3.57). Characteristics are an important concept for this and some more
general second order PDEs, but they don’t play as central a role as in the first
order case.
Example 1.11. For the PDE
uxx − uyy = 0 (1.3.68)
we have − 4AC > 0 and the roots r satisfy − 1 = 0. We may then choose,
B2 r2
for example, α = β = γ = 1, δ = −1, to get the general solution
u(x, y) = F(x + y) + G(x − y) (1.3.69)
2
Next assume that B2 − 4AC = 0. If either of A or C is 0, then so is B, in
which case the PDE already has the form uξ ξ = 0 or uηη = 0, say the first of
these without loss of generality. Otherwise, choose
14 Techniques of Functional Analysis for Differential and Integral Equations
B
α=− β=1 γ =1 δ=0 (1.3.70)
2A
to obtain a = b = 0, c = A, so that the transformed PDE in all cases may be
taken to be uξ ξ = 0.
Finally, if B2 − 4AC < 0 then A = 0 must hold, and we may choose
2A −B
α=√ β=√ γ =0 δ=1 (1.3.71)
4AC − B 2 4AC − B2
in which case the transformed equation is
uξ ξ + uηη = 0 (1.3.72)
We have therefore established that any PDE of the type (1.3.57) can be
transformed, by means of a linear change of variables, to one of the three simple
types,
uξ η = 0 uξ ξ = 0 uξ ξ + uηη = 0 (1.3.73)
each of which then leads to a prototype for a certain larger class of PDEs. If we
allow lower order terms
Auxx + Buxy + Cuyy + Dux + Euy + Fu = 0 (1.3.74)
then after the transformation (1.3.58) it is clear that the lower order terms remain
as lower order terms. Thus any PDE of the type (1.3.74) is, up to a change of
coordinates, one of the three types (1.3.73), up to lower order terms, and only
the value of the discriminant B2 − 4AC needs to be known to determine which
of the three types is obtained.
The preceding discussion motivates the following classification: The PDE
(1.3.74) is said to be:
● hyperbolic if B2 − 4AC > 0
● parabolic if B2 − 4AC = 0
● elliptic if B2 − 4AC < 0
The terminology comes from an obvious analogy with conic sections, that is, the
solution set of Ax2 + Bxy + Cy2 + Dx + Ey + F = 0 is respectively a hyperbola,
parabola, or ellipse (or a degenerate case) according as B2 − 4AC is positive,
zero, or negative.
We can also allow the coefficients A, B, . . . , G to be variable functions of x, y,
and in this case the classification is done pointwise, so the type can change. An
important example of this phenomenon is the so-called Tricomi equation (see,
e.g., Chapter 12 of [14])
which is hyperbolic for x > 0 and elliptic for x < 0. One might refer to the
equation as being parabolic for x = 0 but generally speaking we do not do this,
Some Basic Discussion of Differential and Integral Equations Chapter | 1 15
since it is not really meaningful to speak of a PDE being satisfied in a set without
interior points.
The preceding discussion is special to the case of N = 2 independent
variables—in the case of N ≥ 3 there is no such complete classification. As
we will see there are still PDEs referred to as being hyperbolic, parabolic, or
elliptic, but there are others which are not of any of these types, although these
tend to be of less physical importance.
Wave Equation
For the hyperbolic case we consider the wave equation
utt − c2 uxx = 0 (1.3.76)
where c > 0 is a constant. Here we have changed the name of the variable y
to t, following the usual convention of regarding u = u(x, t) as depending on a
“space” variable x and “time” variable t. This PDE arises in the simplest model
of wave propagation in one space dimension, where u represents, for example,
the displacement of a vibrating medium from its equilibrium position, and c is
the wave speed.
Following the procedure outlined at the beginning of this section, an
appropriate change of coordinates is ξ = x + ct, η = x − ct, and we obtain
the expression, also known as d’Alembert’s formula, for the general solution,
u(x, t) = F(x + ct) + G(x − ct) (1.3.77)
for arbitrary twice differentiable functions F, G. The general solution may be
viewed as the superposition of two waves of fixed shape, moving to the right
and to the left with speed c.
The IVP for the wave equation consists in solving Eq. (1.3.76) for x ∈ R and
t > 0 subject to the side conditions
u(x, 0) = f (x) ut (x, 0) = g(x) x∈R (1.3.78)
where f , g represent the initial displacement and initial velocity of the vibrating
medium. This problem may be completely and explicitly solved by means of
d’Alembert’s formula. Setting t = 0 and using the prescribed side conditions,
we must have
F(x) + G(x) = f (x) c(F (x) − G (x)) = g(x) x∈R (1.3.79)
16 Techniques of Functional Analysis for Differential and Integral Equations
x
Integrating the second relation gives F(x) − G(x) = 1c 0 g(s) ds + C for some
constant C, and combining with the first relation yields
1 1 x
F(x) = f (x) + g(s) ds + C
2 c 0
(1.3.80)
1 1 x
G(x) = f (x) − g(s) ds − C
2 c 0
Substituting into Eq. (1.3.77) and doing some obvious simplification we obtain
1 1 x+ct
u(x, t) = (f (x + ct) + f (x − ct)) + g(s) ds (1.3.81)
2 2c x−ct
We remark that a general solution formula like Eq. (1.3.77) can be given for
any PDE which is exactly transformable to uξ η = 0, that is to say, any hyperbolic
PDE of the form (1.3.57), but once lower order terms are allowed such a simple
solution method is no longer available. For example, the so-called Klein-Gordon
equation utt − uxx + u = 0 may be transformed to uξ η + 4u = 0 which
unfortunately cannot be solved in so transparent a form. Thus the d’Alembert
solution method, while very useful when applicable, is limited in its scope.
Heat Equation
Another elementary method, which may be used in a wide variety of situations,
is the separation of variables technique. We illustrate with the case of the initial
and boundary value problem
ut = uxx 0<x<1 t>0 (1.3.82)
u(0, t) = u(1, t) = 0 t>0 (1.3.83)
u(x, 0) = f (x) 0<x<1 (1.3.84)
Here Eq. (1.3.82) is the heat equation, a parabolic equation modeling, for
example, the temperature in a one-dimensional medium u = u(x, t) as a function
of location x and time t, Eq. (1.3.83) are the boundary conditions, stating that
the temperature is held at temperature zero at the two boundary points x = 0
and x = 1 for all t, and Eq. (1.3.84) represents the initial condition, that is, that
the initial temperature distribution is given by the prescribed function f (x).
We begin by ignoring the initial condition and otherwise looking for special
solutions of the form u(x, t) = φ(t)ψ(x). Obviously u = 0 is such a solution,
but cannot be of any help in eventually solving the full stated problem, so we
insist that neither of φ or ψ is the zero function. Inserting into Eq. (1.3.82) we
obtain immediately that
φ (t)ψ(x) = φ(t)ψ (x) (1.3.85)
must hold, or equivalently
φ (t) ψ (x)
= (1.3.86)
φ(t) ψ(x)
Some Basic Discussion of Differential and Integral Equations Chapter | 1 17
Since the left side depends on t alone and the right side on x alone, it must be that
both sides are equal to a common constant which we denote by −λ (without yet
at this point ruling out the possibility that λ itself is negative or even complex).
We have therefore obtained ODEs for φ and ψ
φ (t) + λφ(t) = 0 ψ (x) + λψ(x) = 0 (1.3.87)
linked via the separation constant λ. Next, from the boundary condition (1.3.83)
we get φ(t)ψ(0) = φ(t)ψ(1) = 0, and since φ is nonzero we must have ψ(0) =
ψ(1) = 0.
The ODE and side conditions for ψ, namely
ψ (x) + λψ(x) = 0 0 < x < 1 ψ(0) = ψ(1) = 0 (1.3.88)
is the simplest example of a so-called Sturm-Liouville problem, a topic which
will be studied in detail in Chapter 13, but this particular case can be handled
by elementary considerations. We emphasize that our goal is to find nonzero
solutions of Eq. (1.3.88), along with the values of λ these correspond to, and as
we will see, only certain values of λ will be possible.
Considering first the case that λ > 0, the general solution of the ODE is
√ √
ψ(x) = c1 sin λx + c2 cos λx (1.3.89)
c1 + c2 = 0 c1 ek + c2 e−k = 0 (1.3.90)
we see that the unique solution is c1 = c2 = 0 for any k > 0. Likewise we can
check that ψ = 0 is the only possible solution for k = 0 and for nonreal k.
For each allowed value of λ we obviously have the corresponding function
φ(t) = e−λt , so that
uk (x, t) = e−k
2π 2t
sin kπx k = 1, 2, . . . (1.3.91)
represents, aside from multiplicative constants, all possible product solutions of
Eqs. (1.3.82), (1.3.83).
To complete the solution of the initial and boundary value problem, we
observe that any sum ∞ k=1 ck uk (x, t) is also a solution of Eqs. (1.3.82), (1.3.83)
as long as ck → 0 sufficiently rapidly, and we try to choose the coefficients ck
to achieve the initial condition (1.3.84). This amounts to the requirement that
18 Techniques of Functional Analysis for Differential and Integral Equations
∞
f (x) = ck sin kπx (1.3.92)
k=1
hold. For any f for which such a sine series representation is valid, we then have
the solution of the given PDE problem
∞
ck e−k
2π 2t
u(x, t) = sin kπ x (1.3.93)
k=1
The question then becomes to characterize this set of f ’s in some more straight-
forward way, and this is done, among many other things, within the theory
of Fourier series, which will be discussed in Chapter 7. Roughly speaking the
conclusion will be that essentially any reasonable function can be represented
this way, but there are many aspects to this, including elaboration of the precise
sense in which the series converges. One other fact concerning this series which
we can easily anticipate at this point, is a formula for the coefficient ck : If we
assume that Eq. (1.3.92) holds, we can multiply both sides by sin mπ x for some
integer m and integrate with respect to x over (0, 1), to obtain
1 1
cm
f (x) sin mπx dx = cm sin2 mπx dx = (1.3.94)
0 0 2
1
since 0 sin kπx sin mπ x dx = 0 for k = m. Thus, if f is representable by a sine
series, there is only one possibility for the kth coefficient, namely
1
ck = 2 f (x) sin kπx dx (1.3.95)
0
Laplace Equation
Finally we discuss a model problem of elliptic type,
uxx + uyy = 0 x2 + y2 < 1 (1.3.96)
u(x, y) = f (x, y) x2 + y2 = 1 (1.3.97)
where f is a given function. The PDE in Eq. (1.3.96) is known as Laplace’s
∂2 ∂2
equation, and is commonly written as u = 0 where = ∂x 2 + ∂y2 is the
Laplace operator, or Laplacian. A function satisfying Laplace’s equation in some
set is said to be a harmonic function on that set, thus we are solving the bound-
ary value problem of finding a harmonic function in the unit disk x2 + y2 < 1
subject to a prescribed boundary condition on the boundary of the disk.
One should immediately recognize that it would be natural here to make use
of polar coordinates (r, θ ), where according to the usual calculus notations,
y
r = x2 + y2 tan θ = x = r cos θ y = r sin θ (1.3.98)
x
and we regard u = u(r, θ ) and f = f (θ ).
Some Basic Discussion of Differential and Integral Equations Chapter | 1 19
∂u sin θ ∂u
= cos θ − (1.3.101)
∂r r ∂θ
∂u
and similar expressions for ∂y and the second derivatives. The end result is
1 1
uxx + uyy = urr + ur + 2 uθθ = 0 (1.3.102)
r r
We may now try separation of variables, looking for solutions in the special
product form u(r, θ ) = R(r)(θ ). Substituting into Eq. (1.3.102) and dividing
by R gives
R (r) R (r) (θ )
r2 +r =− (1.3.103)
R(r) R(r) (θ )
so both sides must be equal to a common constant λ. Therefore R and must
be nonzero solutions of
Next it is necessary to recognize that there are two “hidden” side conditions
which we must make use of. The first of these is that must be 2π periodic,
since otherwise it would not be possible to express the solution u in terms of
the original variables x, y in an unambiguous way. We can make this explicit by
requiring
and here we encounter the second hidden condition: the solution R should not be
singular at the origin, since otherwise the PDE would not be satisfied throughout
the unit disk. Thus we should choose c2 = 0 in each case, leaving R(r) = rk , k =
0, 1, . . . .
Summarizing, we have found all possible product solutions R(r)(θ ) of
Eq. (1.3.96), and they are
This is another problem in the theory of Fourier series, which is very similar to
that associated with Eq. (1.3.92), and again will be studied in detail in Chapter 7.
Exact formulas for the coefficients in terms of f may be given, as in Eq. (1.3.95),
see Exercise 1.19.
N
∂ 2u
u = (1.3.112)
k=1
∂xk2
Some Basic Discussion of Differential and Integral Equations Chapter | 1 21
● The PDE
u = h x ∈ (1.3.113)
is Poisson’s equation, or Laplace’s equation in the special case that h = 0.
It is regarded as being of elliptic type, by analogy with the N = 2 case
discussed in the previous section, or on account of a more general definition
of ellipticity which will be given in Chapter 8. The most common type of
side conditions associated with this PDE are
– Dirichlet, or first kind, boundary conditions
u(x) = g(x) x ∈ ∂ (1.3.114)
– Neumann, or second kind, boundary conditions
∂u
(x) = g(x) x ∈ ∂ (1.3.115)
∂n
∂n (x) = (∇u · n)(x) is the directional derivative in the direction of
where ∂u
the outward normal direction n(x) for x ∈ ∂.
– Robin, or third kind, boundary conditions
∂u
(x) + σ (x)u(x) = g(x) x ∈ ∂ (1.3.116)
∂n
for some given function σ .
● The PDE
u + λu = h x ∈ (1.3.117)
where λ is some constant, is the Helmholtz equation, also of elliptic type.
The three types of boundary condition mentioned for the Poisson equation
may also be imposed in this case.
● The PDE
ut = u x∈ t>0 (1.3.118)
is the heat equation and is of parabolic type. Here u = u(x, t), where
x is regarded as a spatial variable and t a time variable. By convention,
the Laplacian acts only with respect to the N spatial variables x1 , . . . , xN .
Appropriate side conditions for determining a solution of the heat equation
are an initial condition
u(x, 0) = f (x) x ∈ (1.3.119)
and boundary conditions of the Dirichlet, Neumann, or Robin type men-
tioned above. The only needed modification is that the functions involved
may be allowed to depend on t, for example, the Dirichlet boundary
condition for the heat equation is
u(x, t) = g(x, t) x ∈ ∂ t > 0 (1.3.120)
and similarly for the other two types.
22 Techniques of Functional Analysis for Differential and Integral Equations
● The PDE
utt = u x∈ t>0 (1.3.121)
is the wave equation and is of hyperbolic type. Since it is second order in t
it is natural that there be two initial conditions, usually given as
u(x, 0) = f (x) ut (x, 0) = g(x) x∈ (1.3.122)
Suitable boundary conditions for the wave equation are precisely the same
as for the heat equation.
● Finally, the PDE
iut = u x ∈ RN t>0 (1.3.123)
is the Schrödinger equation. Even when N = 1 it does not fall under the
classification
√ scheme of Section 1.3.2 because of the complex coefficient
i = −1. It is nevertheless one of the fundamental PDEs of mathematical
physics, and we will have some things to say about it in later chapters.
The spatial domain here is taken to be all of RN rather than a subset
because this is by far the most common situation and the only one which
will arise in this book. Since there is no spatial boundary, the only needed
side condition is an initial condition for u, u(x, 0) = f (x), as in the heat
equation case.
given until such choices are made. In fact, even the existence and uniqueness
requirements, which may seem more clear cut, may also turn out to require
much clarification in terms of what the exact meaning of “solution” is.
A problem which is not well-posed is called ill-posed. A classical problem
in which ill-posedness can be easily recognized is Hadamard’s example:
uxx + uyy = 0 −∞<x<∞ y>0 (1.4.125)
u(x, 0) = 0 uy (x, 0) = g(x) −∞<x<∞ (1.4.126)
Note that it is not of one of the standard types mentioned above.
If g(x) = α sin kx for some α, k > 0 then a corresponding solution is
sin kx ky
u(x, y) = α e (1.4.127)
k
This is known to be the unique solution, but notice that a change in α (i.e., of the
data g) of size implies a corresponding change in the solution for, say, y = 1 of
size ek . Since k can be arbitrarily large, it follows that the problem is ill-posed,
that is, small changes in the data do not necessarily lead to small changes in the
solution.
Note that in this example if we change the PDE from uxx + uyy = 0 to
uxx − uyy = 0 then (aside from the name of a variable) we have precisely the
problem (1.3.76), (1.3.78), which from the explicit solution (1.3.81) may be seen
to be well-posed under any reasonable interpretation. This serves to emphasize
that some care must be taken in recognizing what are the “correct” side
conditions for a given PDE. Other interesting examples of ill-posed problems
are given in Exercises 1.23 and 1.26, see also [26].
1.5 EXERCISES
1.1. Find a fundamental set and the general solution of u + u + u = 0.
1.2. Let L = aD2 + bD + c (a = 0) be a constant coefficient second order
linear differential operator, and let p(λ) = aλ2 + bλ + c be the associated
characteristic polynomial. If λ1 , λ2 are the roots of p, show that we can
express the operator L as L = a(D − λ1 )(D − λ2 ). Use this factorization
to obtain the general solution of Lu = 0 in the case of repeated roots,
λ1 = λ2 .
√
1.3. Show that the solution of the IVP y = 3 y, y(0) = 0 is not unique.
(Hint: y(t) = 0 is one solution, find another one.) Why doesn’t this
contradict the assertion in Theorem 1.1 about unique solvability of
the IVP?
1.4. Solve the IVP for the Cauchy-Euler equation
1.12. The method of characteristics developed in Section 1.3.1 for the linear
PDE (1.3.45) can be easily extended to the so-called semilinear equation
a(x, y)ux + b(x, y)uy = c(x, y, u) (1.5.128)
We simply replace Eq. (1.3.47) by
d
u(x(t), y(t)) = c(x(t), y(t), u(x(t), y(t))) (1.5.129)
dt
which is still an ODE along a characteristic. With this in mind, solve
1
ux + xuy + u2 = 0 u(0, y) = y>0 (1.5.130)
y
Vector Spaces
We will be working frequently with function spaces which are themselves
special cases of more abstract spaces. Most such spaces which are of interest
to us have both linear structure and metric structure. This means that given any
two elements of the space it is meaningful to speak of (i) a linear combination
of the elements, and (ii) the distance between the two elements. These two kinds
of concepts are abstracted in the definitions of vector space and metric space. In
this chapter we focus on the first of these aspects.
This notion of basis will be inadequate for later purposes, and will be replaced
by one which allows infinite sums, but this cannot be done until a meaning of
convergence is available. The notion of basis in Definition 2.4 is called a Hamel
basis if a distinction is necessary.
Definition 2.5. We say that dim (X), the dimension of X, is m if there exist
m linearly independent vectors in X but any collection of m + 1 elements of
X is linearly dependent. If there exists m linearly independent vectors for any
positive integer m, then we say dim (X) = ∞.
Proposition 2.2. The elements {x1 , x2 , . . . , xm } form a basis for
span ({x1 , x2 , . . . , xm }) if and only if they are linearly independent.
2.4 EXERCISES
2.1. Using only the vector space axioms, show that the zero element in [V3]
is unique.
2.2. Prove Propositions 2.2 and 2.3.
2.3. Show that the intersection of any family of subspaces of a vector space
is also a subspace. Is the same true for the union of subspaces?
2.4. Show that Mm×n , the set of m × n matrices, with the usual definitions of
addition and scalar multiplication, is a vector space of dimension mn.
Show that the subset of symmetric matrices n × n matrices forms a
subspace of Mn×n . What is its dimension?
2.5. Under what conditions on a measurable set E ⊂ RN and p ∈ (0, ∞] will
it be true that C(E) is a subspace of Lp (E)? Under what conditions is
Lp (E) a subspace of Lq (E)?
34 Techniques of Functional Analysis for Differential and Integral Equations
2.6. Let uj (t) = tλj where λ1 , . . . , λn are arbitrary unequal real numbers.
Show that {u1 . . . un } are linearlyindependent functions on any interval
n
(a, b) ⊂ R. (Suggestion: If j=1 αj t
λj ≡ 0, divide by tλ1 and
differentiate.)
2.7. A side condition for a differential equation is homogeneous if when-
ever two functions satisfy the side condition then so does any linear
combination of the two functions. For example, the Dirichlet type
boundary condition u α = 0 for x ∈ ∂ is homogeneous. Now let
Lu = |α|≤m aα (x)D u denote any linear differential operator. Show
that the set of functions satisfying Lu = 0 and any homogeneous side
conditions is a vector space.
2.8. Consider the differential equation u + u = 0 on the interval (0, π ).
What is the dimension of the vector space of solutions which satisfy
the homogeneous boundary conditions (a) u(0) = u(π ), and (b) u(0) =
u(π ) = 0. Repeat the question if the interval (0, π ) is replaced by (0, 1)
and (0, 2π ).
2.9. Let Df = f for any differentiable function f on R. For any N ≥ 0 show
that D : PN → PN is linear and find its null space and range.
2.10. If X and Y are vector spaces, then the Cartesian product of X and Y, is
defined as the set of ordered pairs
X × Y = {(x, y) : x ∈ X, y ∈ Y} (2.4.8)
Addition and scalar multiplication on X × Y are defined in the natural
way,
(x, y) + ( y) = (x +
x, x, y +
y) λ(x, y) = (λx, λy) (2.4.9)
(a) Show that X × Y is a vector space.
(b) Show that R × R is isomorphic to R2 .
2.11. If X, Y are vector spaces of the same finite dimension, show X and Y are
isomorphic.
2.12. Show that Lp (0, 1) and Lp (a, b) are isomorphic, for any a, b ∈ R and
p ∈ (0, ∞].
Chapter 3
Metric Spaces
and
d∞ (x, y) = max (|x1 − y1 |, |x2 − y2 |, . . . , |xn − yn |) (3.1.3)
Techniques of Functional Analysis for Differential and Integral Equations
http://dx.doi.org/10.1016/B978-0-12-811426-1.00003-9
© 2017 Elsevier Inc. All rights reserved. 35
36 Techniques of Functional Analysis for Differential and Integral Equations
1. That is, E is compact in RN . Compactness is discussed in more detail below, and we avoid using
the term until then.
Another random document with
no related content on Scribd:
appropriate powers of the mind are weakened, while its moral faculty
is at the same time, and by the same means, greatly deteriorated.
“Lalande, if not the most profound and original, was certainly the
most learned astronomer of France, and the principal benefactor of
the science to which he was so passionately devoted. He was
remarkable for the most egregious vanity, and for the broadest
eccentricities of character, and almost equally eminent for the most
noble virtues of the heart. By a very singular perversion of intellect,
he became a professed atheist, about the commencement of the
revolution; pronounced, in the year 1793, in the Pantheon, a
discourse against the existence of a God, with the red cap upon his
head; and displayed, on this subject, the most absolute insanity,
during the rest of his life. This monstrous infatuation betrayed him
into the most whimsical acts of extravagance, and particularly into
the publication of a Dictionary of Atheists, in which he enregistered
not only many of “the illustrious dead,” but a great number of his
cotemporaries, and among these, some of the principal dignitaries of
the empire.
This very Mr. Lalande, in the preface to the third edition of his
inestimable work entitled Astronomie, published at Paris so late as
the year 1792, shews, that astronomy furnishes most powerful
proofs of the being of a God. Yet this same man, in one year after,
when in his “dotage,” with a mind enfeebled by age, and corrupted
by the delusions of the new philosophy of his countrymen, became
an object of “derision,” and of “mockery,” even among Frenchmen;
for his absurdities, and his endeavours to set himself up as a
champion of atheism! Is it necessary to furnish the rational part of
mankind with a more striking, and at the same time a more
lamentable proof, of the deleterious effects produced by those
illusions, which, under the assumed name of “Philosophy,” have
been conjured up by some modern Theorists and Political
Speculators? Certainly, it is not. The instance, here adduced, may
stand as a monument of the folly and depravity of the Philosophy of
the Gallican School.
335. “If,” (says a late anonymous writer,) “from the advantages of
sound learning to the state, we turn to its influence on the characters
of individuals, we will find its effects to be no less striking. We will
find, that although, without much learning, man may become useful
and respectable, yet that he cannot, without it, become polished,
enlightened and great; he cannot ascend to that grade in the scale of
his Creator’s works, to which his powers are intended to exalt him. If
to this rule, a Franklin, a Rittenhouse, and a Washington present
exceptions, they are to regarded as mere exceptions, and therefore
do not amount to an infraction of the rule. They were prodigies;
which necessarily implies a departure from, and an ascendency over
common principles.” See an Account of Dickinson College, Carlisle,
in the Port Folio, for March, 1811; supposed to be written by
Professor Cooper.
Still, adds Dr. Rush, “It is true Dr. Sydenham did not adopt, or
follow, the errors of the schools in which he had been educated; but,
by knowing them thoroughly, he was able, more easily, to examine
and refute them.” Here, then, is an admission, that even an intimate
knowledge of such errors is eminently useful, by enabling a man of a
sound and cultivated mind to refute them: for, the refutation of
existing errors, affords a most important aid to the advancement of
true science.
336b. The three great Universities of England and Ireland enjoy the right, in
addition to many other important privileges, of sending, each, two members to
represent them in parliament, Would to heaven! that there were something like a
representation of the interests of learning and science, in the legislative bodies of
our own country.
338. The Greek and Latin are called by way of pre-eminence, the
learned languages. Baron Bielfeld enumerates the advantages
resulting from a knowledge of the former; among which he notices
that important one, of its enabling us more readily and clearly to
comprehend the meaning of that almost boundless list of terms in
the arts and sciences, used in modern languages and styled
technical, which are either altogether Grecian, or derived from that
language. He then makes this remark: “From all that has been said,
it is apparent how much utility attends the study of the Greek tongue;
and how much reason the English have, for applying themselves to
it, from their early youth.” “But,” observes this learned and
discriminating writer, “that which has given the Latin an advantage
over the Greek itself, that has rendered it indispensable to every
man of letters, and has made it the basis of erudition, is, that during
the middle age, and in general in all modern times, the learned of all
Europe have made it their common and universal language; so that
the Latin forms, if we may use the expression, the natural language
of the sciences.” Elem. of Univ. Erud.
339. Although Mr. T. Cooper (before quoted) admits, that the “strict
adherence to the syllogistic mode of reasoning,” that which he calls
“playing tricks with syllogisms,” together with “the legerdemain of
words without ideas,” was carried much too far by some late
metaphysical writers of eminence; yet he is of opinion, that “in
modern times, this invention of Aristotle is abandoned more than it
deserves to be: For,” continues Mr. Cooper, “no man can so skilfully
analyse the argument of another, as one who is well acquainted with
the rules of scholastic logic, and accustomed to apply them. Good
reasoners there are and will be, who know nothing of these rules, but
better reasoners who do.”
342. In the year 1789, Dr. Rittenhouse translated from the German
of Mr. Lessing, director of the theatre at Hamburg, a tragedy called
Lucia Sampson; which translation was printed; in the same year, by
Mr. Charles Cist, of Philadelphia. In the preface to it, the translator
says:—“This translation was attempted at the request of a friend;
and the many virtuous sentiments and excellent lessons of morality it
contains, will apologize for its being offered to the public. To young
ladies it may afford useful instruction, and will, from the nature of the
distress, be particularly useful to them: an elegant writer well
acquainted with the human heart, has observed, that the affection of
a father to his daughter unites extreme sensibility with the utmost
delicacy; and this sentiment is, no doubt, in a great degree
reciprocal.”
says; “Who could imagine that Locke was fond of romances; that
Newton once studied astrology; that Dr. Clarke valued himself for his
agility, and frequently amused himself, in a private room of his
house, in leaping over the tables and chairs; and that our author
himself (Mr Pope) was a great epicure.”
349. “Astronomy, like the Christian religion, if you will allow me the
comparison,” said our Philosopher, “has a much greater influence on
our knowledge in general, and perhaps on our manners too, than is
commonly imagined. Though but few men are its particular votaries,
yet the light it affords is universally diffused among us; and it is
difficult for us to divest ourselves of its influence so far, as to frame
any competent idea of what would be our situation without it.” See
Ritt. Orat.
354. Ibid.
355. This quotation and the other passages, before which inverted
commas are placed in the margin, in the two last paragraphs of the
text, are extracted from Dr. Rittenhouse’s Oration.
356. Dr. Thomas Newton, Bishop of Bristol.
358.
“Let others creep by timid steps, and slow,
On plain experience, lay foundations low;
By common sense, to common knowledge bred;
And lost to nature’s cause through nature led:
All-seeing in thy mists, we want no guide,
Mother of Arrogance and source of pride!
We nobly take the high priori road,
And reason downward, till we doubt of God.”
Pope’s Dunciad, b. IV. l. 455.
Mr. Pope had put the above poetical lines into the mouth of one of
his Dunces, when addressing himself to the goddess Dullness. And
as the great Dr. Samuel Clarke had previously endeavoured to shew,
[358a]
that the Being of a God may be demonstrated by arguments
deduced â priori, the Doctor conceived himself to be struck at,
among those “better reasoners” alluded to, in the note above
mentioned.
358a. In his work entitled, “A Discourse concerning the Being and Attributes of
a God, the Obligations of Natural Religion, and the truth and certainty of the
Christian Revelation; in answer to Mr. Hobbes, Spinoza, the Author of the Oracles
of Reason, and other deniers of natural and revealed Religion.”
360. Ibid.
361. “Other systems of Philosophy have ever found it necessary to
conceal their weakness and inconsistency, under the veil of
unintelligible terms and phrases, to which no two mortals, perhaps,
ever affixed the same meaning. But the philosophy of Newton
disdains to make use of such subterfuges; it is not reduced to the
necessity of using them, because it pretends not to be of nature’s
privy council, or to have access to her most inscrutable mysteries;
but, to attend carefully to her works, to discover the immediate
causes of visible effects, to trace those causes to others more
general and simple, advancing by slow and sure steps towards the
Great First Cause of all things.” Ritt. Orat.
366. “It belongs to monarchies,” says Dr. Rush, “to limit the
business of government to a privileged order of men.” See Eulog.
367. See Ritt. Orat. before the Am. Philos. Soc. in 1775.
369. Mr. Pope was not singular in the opinion here expressed: one
of the most illustrious legislators and best practical statesmen the
world has ever known, appears to have entertained the same
sentiment, when he penned the following passages: they are
extracted from the Frame of Government originally designed by
William Penn, for Pennsylvania: published in the year 1682.
375. Ibid.