Complete Download Variational Calculus with Engineering Applications Constantin Udriste PDF All Chapters
Complete Download Variational Calculus with Engineering Applications Constantin Udriste PDF All Chapters
https://ebookmass.com
https://ebookmass.com/product/variational-
calculus-with-engineering-applications-constantin-
udriste/
https://ebookmass.com/product/calculus-with-applications-11th-edition-
ebook-pdf/
testbankdeal.com
https://ebookmass.com/product/microeconomics-theory-and-applications-
with-calculus-5th-global-edition-jeffrey-m-perloff/
testbankdeal.com
https://ebookmass.com/product/engineering-practice-with-oilfield-and-
drilling-applications-1st-edition-scott-d-sudhoff/
testbankdeal.com
Fractional Calculus and Fractional Processes with
Applications to Financial Economics: Theory and
Application 1st Edition Fabozzi
https://ebookmass.com/product/fractional-calculus-and-fractional-
processes-with-applications-to-financial-economics-theory-and-
application-1st-edition-fabozzi/
testbankdeal.com
https://ebookmass.com/product/calculus-with-calcchat-calcview-12th-
edition-larson/
testbankdeal.com
https://ebookmass.com/product/calculus-and-its-applications-11th-
edition-ebook-pdf-version/
testbankdeal.com
https://ebookmass.com/product/mixed-phase-clouds-observations-and-
modeling-constantin-andronache/
testbankdeal.com
https://ebookmass.com/product/introduction-to-the-variational-
formulation-in-mechanics-edgardo-o-taroco/
testbankdeal.com
Variational Calculus with Engineering Applications
Variational Calculus with Engineering
Applications
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
except as permitted by law. Advice on how to obtain permission to reuse material from this title is
available at http://www.wiley.com/go/permissions.
The right of Constantin Udriste and Ionel Tevy to be identified as the authors of this work has been
asserted in accordance with law.
Registered Office(s)
John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
Editorial Office(s)
9600 Garsington Road, Oxford, OX4 2DQ, UK
The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
For details of our global editorial offices, customer services, and more information about Wiley products
visit us at www.wiley.com.
Wiley also publishes its books in a variety of electronic formats and by print-on-demand. Some content
that appears in standard print versions of this book may not be available in other formats.
A catalogue record for this book is available from the Library of Congress
Set in 9.5/12.5pt STIXTwoText by Integra Software Services Pvt. Ltd, Pondicherry, India
v
Contents
Preface ix
2 Variational Principles 23
2.1 Problems with Natural Conditions at the Boundary 23
2.2 Sufficiency by the Legendre-Jacobi Test 27
2.3 Unitemporal Lagrangian Dynamics 30
2.3.1 Null Lagrangians 31
2.3.2 Invexity Test 32
2.4 Lavrentiev Phenomenon 33
2.5 Unitemporal Hamiltonian Dynamics 35
2.6 Particular Euler–Lagrange ODEs 37
2.7 Multitemporal Lagrangian Dynamics 38
2.7.1 The Case of Multiple Integral Functionals 38
2.7.2 Invexity Test 40
2.7.3 The Case of Path-Independent Curvilinear Integral Functionals 41
2.7.4 Invexity Test 44
2.8 Multitemporal Hamiltonian Dynamics 45
2.9 Particular Euler–Lagrange PDEs 47
2.10 Maple Application Topics 48
vi Contents
4 Variational Integrators 79
4.1 Discrete Single-time Lagrangian Dynamics 79
4.2 Discrete Hamilton’s Equations 84
4.3 Numeric Newton’s Aerodynamic Problem 87
4.4 Discrete Multi-time Lagrangian Dynamics 88
4.5 Numerical Study of the Vibrating String Motion 92
4.5.1 Initial Conditions for Infinite String 94
4.5.2 Finite String, Fixed at the Ends 95
4.5.3 Monomial (Soliton) Solutions 96
4.5.4 More About Recurrence Relations 100
4.5.5 Solution by Maple via Eigenvalues 101
4.5.6 Solution by Maple via Matrix Techniques 102
4.6 Numerical Study of the Vibrating Membrane Motion 104
4.6.1 Monomial (Soliton) Solutions 105
4.6.2 Initial and Boundary Conditions 108
4.7 Linearization of Nonlinear ODEs and PDEs 109
4.8 Von Neumann Analysis of Linearized Discrete Tzitzeica PDE 113
4.8.1 Von Neumann Analysis of Dual Variational Integrator Equation 115
4.8.2 Von Neumann Analysis of Linearized Discrete Tzitzeica Equation 116
4.9 Maple Application Topics 119
Bibliography 203
Index 209
Visit https://ebookmass.com
now to explore a rich
collection of eBooks and enjoy
exciting offers!
ix
Preface
The Variational Calculus with Engineering Applications was and is being taught to
fourth-year engineering students, in the Faculty of Applied Sciences, Mathematics - Infor-
matics Department, from the University Politehnica of Bucharest, by Prof. Emeritus Dr.
Constantin Udriste. Certain topics are taught at other faculties of our university, espe-
cially as master’s or doctoral courses, being present in the papers that can be published
in journals now categorized as “ISI”.
The chapters were structured according to the importance, accessibility, and impact of
the theoretical notions able to outline a future specialist based on mathematical optimiza-
tion tools. The probing and intermediate variants lasted for a number of sixteen years,
leading to the selection of the most important manageable notions and reaching maturity
through this variant that we decided to publish at Wiley.
Now the topics of the book include seven chapters: Extrema of Differentiable Function-
als; Variational Principles; Optimal Models Based on Energies; Variational Integrators;
Miscellaneous Topics; Extremals with Nonholonomic Constraints; Problems: Free and
Constrained Extremals. To cover modern problem-solving methods, each chapter includes
Maple application topics. The scientific authority we have as professors in technical uni-
versity has allowed us to impose our point of view on the subject and on the types of
reasoning that deserve to be offered to readers of variational calculus texts. Everything
has been structured for the benefit of functional optimization that describes significant
physical or engineering phenomena. To unlock the students’ minds we preferred a simpli-
fied language, specific to applied mathematics, sending readers of pure mathematics to the
bibliography. Long experience as mathematics teachers in technical universities has pre-
pared us to fluently expose ideas undressed by excessive formalizations, with great impact
on the students’ understanding of the natural phenomena encountered in engineering and
economics as well as their dressing in mathematical clothing. In this sense, we preferred
mathematical modeling as a scientific support for intelligent presentation of the real world
and we avoided the hint of abstract notions specific to pure mathematics.
Topics that distinguish this book from existing books include: Maple application top-
ics, invexity tests, case of path-independent curvilinear integral functionals, variational
integrators, discretization of Hamilton ODEs, numerical study of vibrating string motion,
numerical study of vibrating membrane motion, linearization of nonlinear ODEs and
x Preface
PDEs, geometric dynamics and wind, extremals with nonholonomic constraints, free and
constrained extremals. In fact, this is the first book that studies the curvilinear integral
functionals at the level of student courses. The novelty issues were finalized following
repeated discussions with the professors from our faculty and from Faculty of Mathematics
in West University of Timisoara, especially with those interested in functional optimiza-
tion, to whom we bring thanks and thoughts of appreciation. We are open to any comments
or criticisms which bring didactic benefits for variational calculus theories.
The mottos are stanzas from two poems by Tudor Arghezi (1880–1967), a Romanian
writer, known for his contribution to the development of Romanian lyric poetry under
the influence of Baudelaireanism. His poetic work, of exemplary originality, represents
another significant age of Romanian literature.
Born in Bucharest of peasant stock, Tudor Arghezi was awarded Romania’s National
Poetry Prize in 1946 and the State Prize for Poetry in 1956.
The authors thank Associate Prof. Dr. Oana-Maria Pastae, “Constantin Brancusi”
University of Tg-Jiu, for the English improvement of the manuscript. We are indebted to
Prof. Dr. Dumitru Opris from the West University of Timisoara for the solutions offered to
some discretization problems.
Motto:
“I wrote them with my nail on the plaster
On a wall of empty cracks,
In the dark, in my solitude,
Unaided by the bull lion vulture
of Luke, Mark and John.”
Tudor Arghezi – Flowers of Mildew
Variational calculus aims to generalize the constructions of classical analysis, to solve extrema
problems for functionals. In this introductory chapter, we will study the problem of differen-
tiable functionals defined on various classes of functions. The news refers to path integral
functionals. For details, see [4, 6, 7, 8, 14, 18, 21, 30, 58].
The first term on the right-hand side represents the differential of the function 𝑓 at the
point 𝑥, a linear form of argument growth, the vector ℎ; the second term is the deviation
from the linear approximation and is small in relation to ℎ, in the sense that
𝑟(𝑥, ℎ)
lim = 0.
ℎ→0 ‖ℎ‖
Let 𝐔, 𝐕 be two normed vector spaces. The previous definition can be extended imme-
diately to the case of functions 𝑓 ∶ 𝐷 ⊂ 𝐔 → 𝐕. If 𝐔 is a space of functions, and 𝐕 = R,
then instead of function we use the term functional.
Variational Calculus with Engineering Applications, First Edition. Constantin Udriste and Ionel Tevy.
© 2023 John Wiley & Sons Ltd. Published 2023 by John Wiley & Sons Ltd.
2 1 Extrema of Differentiable Functionals
𝑟(𝑥, ℎ)
20 lim = 0.
ℎ→0 ‖ℎ‖
The linear continuous operator 𝐝𝑓(𝑥, ⋅) is called the derivative of the function 𝑓 at given
point 𝑥.
For a given nonzero vector ℎ ∈ 𝐔 and 𝑡 > 0, the vector
𝑓(𝑥 + 𝑡ℎ) − 𝑓(𝑥)
𝛿(𝑥, ℎ) = lim ∈ 𝐕,
𝑡→0 𝑡
if the limit exists, is called the variation or derivative of the function 𝑓 at the point 𝑥 along
the direction ℎ.
A differentiable function of real variables has a derivative along any direction. The prop-
erty is also kept for functions between normed vector spaces. Specifically, the next one
takes place:
𝑟(𝑥, 𝑡ℎ)
= 𝐝𝑓(𝑥, ℎ) + lim ‖ℎ‖ = 𝐝𝑓(𝑥, ℎ).
𝑡ℎ→0 ‖𝑡ℎ‖
Let 𝑓 be a functional. The simplest examples of functionals are given by formulas: (i)
evaluation functional (“application of a function on the value at a point”), 𝑓 ↦ 𝑓(𝑥0 ),
1.1 Differentiable Functionals 3
where 𝑥0 is a fixed point; (ii) definite integration (functional defined by definite integral);
(iii) numerical quadrature defined by definite integration:
𝑏
𝐼(𝑓) = 𝑎0 𝑓(𝑥0 ) + 𝑎1 𝑓(𝑥1 ) + ... + 𝑎𝑛 𝑓(𝑥𝑛 ) = ∫ 𝑓(𝑥)𝑑𝑥, ∀𝑓 ∈ 𝑃𝑛 ,
𝑎
where 𝑃𝑛 means the set of all polynomials of degree at most 𝑛; and (iv) distributions in
analysis (as linear functionals defined on spaces of test functions). We will continue to
deal with definite integral type functionals, as the reasoning is more favorable to us.
defined on the space 𝐶 0 [𝑎, 𝑏] of the continuous functions on the segment [𝑎, 𝑏], endowed
with the norm of uniform convergence. The Lagrangian of the functional (i.e., the function
𝐿(𝑡, 𝑥)) is presumed continuous and with continuous partial derivatives of the first order in
the domain 𝑎 ≤ 𝑡 ≤ 𝑏, −∞ < 𝑥 < +∞.
Let us determine the variation of the functional 𝐽(𝑥(⋅)) when the argument 𝑥(𝑡) increases
with ℎ(𝑡):
𝑏
△𝐽(𝑥(⋅)) = 𝐽(𝑥 + ℎ) − 𝐽(𝑥) = ∫ [𝐿(𝑡, 𝑥(𝑡) + ℎ(𝑡)) − 𝐿(𝑡, 𝑥(𝑡))]𝑑𝑡.
𝑎
𝜕𝐿
𝐿(𝑡, 𝑥 + ℎ) − 𝐿(𝑡, 𝑥) = ℎ + 𝑟(𝑡, 𝑥, ℎ),
𝜕𝑥
where
𝑟(𝑡, 𝑥, ℎ)
lim = 0.
ℎ→0 ℎ
Therefore
𝑏
𝜕𝐿(𝑡, 𝑥(𝑡))
𝛿𝐽(𝑥, ℎ) = ∫ ℎ(𝑡)𝑑𝑡.
𝑎 𝜕𝑥
In this way, according to the previous proposition and the derivation formula for integrals
with parameters, we find
𝑏
𝑑𝐽 | 𝑑 |
𝛿𝐽(𝑥, ℎ) = (𝑥 + 𝜀ℎ)||| = ∫ 𝐿(𝑡, 𝑥(𝑡) + 𝜀ℎ(𝑡))𝑑𝑡 |||
𝑑𝜀 |𝜀=0 𝑑𝜀 𝑎 |𝜀=0
𝑏
𝜕𝐿(𝑡, 𝑥(𝑡))
=∫ ℎ(𝑡)𝑑𝑡. (1)
𝑎 𝜕𝑥
4 1 Extrema of Differentiable Functionals
defined on the space 𝐶 1 [𝑎, 𝑏] of functions with a continuous derivative on the segment [𝑎, 𝑏],
endowed with the norm of uniform convergence of derivatives. The Lagrangian 𝐿(𝑡, 𝑢, 𝑣) of the
functional is supposed to have first-order continuous partial derivatives. To write the integral
functional, we use in fact the pullback form 𝐿(𝑡, 𝑥(𝑡), 𝑥 ′ (𝑡)) of the Lagrangian.
Applying the derivation formula of the integrals with parameter we obtain
𝑑𝐽 |
𝛿𝐽(𝑥, ℎ) = (𝑥 + 𝜀ℎ)|||
𝑑𝜀 |𝜀=0
𝑏
𝑑 |
= ∫ 𝐿(𝑡, 𝑥(𝑡) + 𝜀ℎ(𝑡), 𝑥 ′ (𝑡) + 𝜀ℎ′ (𝑡))𝑑𝑡 |||
𝑑𝜀 𝑎 |𝜀=0
𝑏
𝜕𝐿 𝜕𝐿
=∫ [ (𝑡, 𝑥(𝑡), 𝑥 ′ (𝑡)) ℎ(𝑡) + ′ (𝑡, 𝑥(𝑡), 𝑥 ′ (𝑡)) ℎ′ (𝑡)] 𝑑𝑡. (2)
𝑎 𝜕𝑥 𝜕𝑥
Analogously, we can extend the result to the functional
𝑏
𝐽(𝑥(⋅)) = ∫ 𝐿(𝑡, 𝑥(𝑡), 𝑥 (1) (𝑡), ..., 𝑥 (𝑚) (𝑡)) 𝑑𝑡,
𝑎
defined on the space 𝐶 𝑚 [𝑎, 𝑏] of functions with continuous derivatives up to the order 𝑚
inclusive, on the segment [𝑎, 𝑏]. In this way we have
𝑏
𝜕𝐿 𝜕𝐿 (1) 𝜕𝐿 (𝑚)
𝛿𝐽(𝑥, ℎ) = ∫ [ ℎ(𝑡) + ℎ (𝑡) + ... + ℎ (𝑡)] 𝑑𝑡. (3)
𝑎 𝜕𝑥 𝜕𝑥 (1) 𝜕𝑥 (𝑚)
Example 1.4: Let us consider a functional which depends on several function variables, for
example
𝑏
𝐽(𝑥1 (⋅), ..., 𝑥 𝑛 (⋅)) = ∫ 𝐿(𝑡, 𝑥 1 (𝑡), ..., 𝑥 𝑛 (𝑡), 𝑥̇ 1 (𝑡), ..., 𝑥̇ 𝑛 (𝑡)) 𝑑𝑡,
𝑎
defined on the space (𝐶 1 [𝑎, 𝑏])𝑛 , whose elements are vector functions 𝑥(𝑡) =
(𝑥 1 (𝑡), ..., 𝑥 𝑛 (𝑡)), 𝑡 ∈ [𝑎, 𝑏], with the norm of uniform convergence of derivatives
‖𝑥‖ = max [|𝑥 1 (𝑡)|, ..., |𝑥𝑛 (𝑡)|, |𝑥̇ 1 (𝑡)|, ..., |𝑥̇ 𝑛 (𝑡)| ]
𝑡∈[𝑎,𝑏]
Now, let us consider functionals whose arguments are functions of several real variables.
Example 1.5: Let Ω ⊂ R2 be a compact domain. As an example we will take, first, the
double integral functional
𝜕𝑤 𝜕𝑤
𝑤𝑥 = , 𝑤𝑦 = .
𝜕𝑥 𝜕𝑦
The functional is defined on the space 𝐶 1 (Ω) of all functions with continuous partial deriva-
tives; the norm of the space 𝐶 1 (Ω) is given by
‖𝑤‖ = max [ |𝑤(𝑥, 𝑦)|, |𝑤𝑥 (𝑥, 𝑦)|, |𝑤𝑦 (𝑥, 𝑦)| ].
(𝑥,𝑦)∈Ω
Supposing that the Lagrangian 𝐿 has continuous partial derivatives, the variation of the
functional 𝐽, as the argument 𝑤 grows with ℎ(𝑥, 𝑦), is
𝑑𝐽 |
𝛿𝐽(𝑤, ℎ) = (𝑤 + 𝜀ℎ)|||
𝑑𝜀 |𝜀=0
𝑑 |
= ∫ ∫ 𝐿(𝑥, 𝑦, 𝑤 + 𝜀ℎ, 𝑤𝑥 + 𝜀ℎ𝑥 , 𝑤𝑦 + 𝜀ℎ𝑦 ) 𝑑𝑥𝑑𝑦 |||
𝑑𝜀 Ω
|𝜀=0
𝜕𝐿 𝜕𝐿 𝜕𝐿
=∫ ∫ ( ℎ+ ℎ + ℎ ) 𝑑𝑥 𝑑𝑦. (5)
Ω 𝜕𝑤 𝜕𝑤𝑥 𝑥 𝜕𝑤𝑦 𝑦
where Γ is a piecewise 𝐶 1 curve which joins two fixed points 𝐴, 𝐵 in a compact domain
Ω ⊂ R2 .
Suppose the argument 𝑤(𝑥, 𝑦) grows with ℎ(𝑥, 𝑦). Then the variation of the functional is
𝐽(𝑥0 ) ≤ 𝐽(𝑥), ∀𝑥 ∈ 𝑉 ∩ 𝐷.
If the point 𝑥0 has a neighborhood 𝑉 on which the opposite inequality takes place
𝐽(𝑥0 ) ≥ 𝐽(𝑥), ∀𝑥 ∈ 𝑉 ∩ 𝐷,
we say that 𝑥0 is a point of local maximum for the functional 𝐽. The minimum and
maximum points are called relative extrema points.
In classical analysis, the extrema points of a differentiable function are among the critical
points, that is, among the points that cancels first-order derivatives. A similar property
occurs in the case of functional ones on normed vector spaces of functions, only in this
case the extrema points are found between the extremals (critical points).
Proposition 1.2: If the function 𝑥 is an extremum point for the functional 𝐽, interior point
of the set 𝐷 and if 𝐽 is differentiable at this point, then 𝛿𝐽(𝑥, ℎ) = 0 for any growth ℎ.
Proof. Let ℎ be a growth of the argument 𝑥 (function); since 𝑥 is an interior point of the set
𝐷, the function 𝐽(𝑥 + 𝜀ℎ) of real variable 𝜀 is defined on hole interval [−1, 1]. This function
has an extremum point at 𝜀 = 0 and is derivable at this point. Then its derivative must
vanish at 𝜀 = 0. It follows
𝑑𝐽 |
𝛿𝐽(𝑥, ℎ) = (𝑥 + 𝑡ℎ)||| = 0.
𝑑𝑡 |𝑡=0
Any point 𝑥 at which the variation 𝛿𝐽(𝑥, ℎ) of the functional 𝐽 is canceled identically
with respect to ℎ is called either the stationary point or critical point or extremal of the
functional.
Hence, for establishing the extrema points of a functional, the variation 𝛿𝐽(𝑥, ℎ) must be
expressed, determine the critical points (those at which the variation is canceled identically
with respect to ℎ) and then choose from these the minimum points or the maximum points.
Commentary The variation 𝛿𝐽(𝑥, ℎ) of the functional 𝐽 is a continuous linear functional
on the normed vector space 𝐔 on which is defined 𝐽. The set 𝐔∗ of all continuous linear
functionals on 𝐔 is called dual space. Thus the evaluation of the variation of a functional
imposes the existence of some representation theorems for the dual space.
For example, for any linear functional 𝑇 ∶ R𝑛 → R there exists a vector 𝑎 = (𝑎1 , ..., 𝑎𝑛 ) ∈
R such that
𝑛
Also, any continuous and linear functional 𝑇 ∶ 𝐶 0 [𝑎, 𝑏] → R, on the space of continuous
functions, can be written in the form
𝑏
𝑇(ℎ) = ∫ ℎ(𝑡) 𝑑𝑔𝑇 (𝑡),
𝑎
where 𝑔𝑇 is a function with bounded variation and continuous at the right (F. Riesz’s
Theorem).
To determine the extrema points, we need the identical cancelation of the first variation.
For that we use
Proof. Suppose there exists a point 𝑡0 ∈ [𝑎, 𝑏] at which 𝑥(𝑡0 ) ≠ 0. Let 𝑥(𝑡0 ) > 0. Then
there exists a neighborhood 𝑉 = (𝑡0 − 𝜀, 𝑡0 + 𝜀) of 𝑡0 where 𝑥(𝑡) > 0. Let us consider
a continuous function ℎ, strictly positive on the neighborhood 𝑉 and null outside it (for
example, ℎ(𝑡) = (𝑡 − 𝑡0 + 𝜀)(𝑡0 + 𝜀 − 𝑡) for 𝑡 ∈ 𝑉). Then
𝑏 𝑡0 +𝜀
∫ 𝑥(𝑡)ℎ(𝑡) 𝑑𝑡 = ∫ 𝑥(𝑡)ℎ(𝑡) 𝑑𝑡 > 0,
𝑎 𝑡0 −𝜀
which contradicts the hypothesis and concludes the proof of the lemma.
Remark 1.1: 1) For the set of continuous real functions 𝐶 0 [𝑎, 𝑏], the integral ⟨𝑥, 𝑦⟩ =
𝑏
∫ 𝑥(𝑡)ℎ(𝑡) 𝑑𝑡 is a scalar product. Consequently,
𝑎
𝑏
⟨𝑥, 𝑦⟩ = ∫ 𝑥(𝑡)ℎ(𝑡) 𝑑𝑡 = 0, ∀ℎ(𝑡)
𝑎
implies 𝑥(𝑡) = 0.
2) The set of functions ℎ you need for checking the equality in the lemma can often
be reduced. As can be seen from the proof of the fundamental lemma, we can impose
derivability conditions of any order to the functions ℎ and cancelation at the end of the
interval etc. It is essential that for any 𝑡 ∈ [𝑎, 𝑏] and any 𝜀 > 0 to exist, in the set of
test functions, a strictly positive function on the interval (𝑡0 − 𝜀, 𝑡0 + 𝜀) and null function
otherwise.
3) In a similar wording, the fundamental lemma of the variational calculus remains in
force for functions of several variables.
Visit https://ebookmass.com
now to explore a rich
collection of eBooks and enjoy
exciting offers!
8 1 Extrema of Differentiable Functionals
4) We can also see that in previous constructions it is sufficient that the functional 𝐽 be
defined on a linear variety (affine subspace) or at least on a convex set: with two points 𝑥
and 𝑥 + ℎ to be in the domain of 𝐽 all points of the form 𝑥 + 𝑡ℎ for any 𝑡 ∈ R or at least
𝑡 ∈ [−1, 1] too. Specifically, we can impose restrictions (conditions) on extrema points and
variations ℎ, but remaining in a linear variety.
Problem 1.2.1: a) Let 𝑥 be a 𝐶 𝑘 function on the interval [𝑎, 𝑏]. If for any ℎ ∈ 𝐶 𝑘 [𝑎, 𝑏],
null at 𝑎 and 𝑏, together with its first 𝑘 − 1 derivatives, the equality
𝑏
∫ 𝑥(𝑡) ℎ(𝑘) (𝑡) 𝑑𝑡 = 0
𝑎
𝑏 𝑡0 +𝜀
= (−1)𝑘 ∫ 𝑥 (𝑘) (𝑡)ℎ(𝑡) 𝑑𝑡 = (−1)𝑘 ∫ 𝑥(𝑘) (𝑡)ℎ(𝑡) 𝑑𝑡 ≠ 0,
𝑎 𝑡0 −𝜀
Obviously, if 𝑔 = ℎ(𝑘) , the property (P) results by integrating by parts as many times as
necessary.
Conversely, let us denote
𝑡 𝑠𝑘 𝑠2 𝑡
1
ℎ(𝑡) = ∫ ∫ ... ∫ 𝑔(𝑠1 ) 𝑑𝑠1 ...𝑑𝑠𝑘 = ∫ 𝑔(𝑠)(𝑡 − 𝑠)𝑘−1 𝑑𝑠.
𝑎 𝑎 𝑎
𝑛! 𝑎
Then ℎ = 𝑔 and ℎ(𝑎) = ℎ′ (𝑎) = ...ℎ(𝑘−1) (𝑎) = 0, and the equalities in the property (𝑃), in
(𝑘)
which we put 𝑔 = ℎ(𝑘) and integrate by parts, give us successively ℎ(𝑘−1) (𝑏) = ... = ℎ′ (𝑏) =
ℎ(𝑏) = 0.
1.2 Extrema of Differentiable Functionals 9
and determine the coefficients 𝑐0 , 𝑐1 , ...𝑐𝑘−1 so that the function 𝑔 verifies the equalities of
the property (𝑃).
The orthogonalization properties in a Euclidean vector space ensure the possibility of
this determination. Then we can take ℎ(𝑘) = 𝑔 and get
𝑏 𝑏
0=∫ 𝑥(𝑡) 𝑔(𝑡) 𝑑𝑡 = ∫ (𝑔(𝑡) + (𝑐0 + 𝑐1 𝑡 + ... + 𝑐𝑘−1 𝑡𝑘−1 )) 𝑔(𝑡) 𝑑𝑡
𝑎 𝑎
𝑏 𝑏 𝑏
=∫ 𝑔2 (𝑡) 𝑑𝑡 + ∫ (𝑐0 + 𝑐1 𝑡 + ... + 𝑐𝑘−1 𝑡 𝑘−1 ) 𝑔(𝑡) 𝑑𝑡 = ∫ 𝑔2 (𝑡) 𝑑𝑡.
𝑎 𝑎 𝑎
Remark 1.2: The previous result illustrates on a concrete situation the equality 𝑈 ⊥⊥ =
𝑈, for any closed subspace of a Hilbert space.
Let us examine the detection of critical points (extremals) for the functionals whose
variations were determined in the previous section.
the variation is given by formula (1). Then, applying the fundamental lemma we obtain
𝜕𝐿
(𝑡, 𝑥(𝑡)) = 0. (7)
𝜕𝑥
Equation (7) can be solved, according to the implicit function theorem, obtaining solutions
𝑥 = 𝑥(𝑡), around the points (𝑡0 , 𝑥0 ) at which
𝜕𝐿 𝜕2 𝐿
(𝑡0 , 𝑥0 ) = 0, (𝑡 , 𝑥 ) ≠ 0.
𝜕𝑥 𝜕𝑥2 0 0
Example 1.9: A special place, generic in the variational calculus, has the functional
𝑏
𝐽(𝑥(⋅)) = ∫ 𝐿(𝑡, 𝑥(𝑡), 𝑥 ′ (𝑡)) 𝑑𝑡,
𝑎