Deterministic and Stochastic Modeling

in Computational Electromagnetics
Dragan Poljak
Visit to download the full and correct content document:
https://ebookmass.com/product/deterministic-and-stochastic-modeling-in-computation
al-electromagnetics-dragan-poljak/
Deterministic and Stochastic Modeling
in Computational Electromagnetics
 IEEE Press
445 Hoes Lane
Piscataway, NJ 08854
IEEE Press Editorial Board
Sarah Spurgeon, Editor in Chief

Jón Atli Benediktsson Behzad Razavi Jeffrey Reed

Anjan Bose Jim Lyke Diomidis Spinellis
James Duncan Hai Li Adam Drobot
Amin Moeness Brian Johnson Tom Robertazzi
Desineni Subbaram Naidu Ahmet Murat Tekalp
Deterministic and Stochastic Modeling
in Computational Electromagnetics

Integral and Differential Equation Approaches

Dragan Poljak and Anna Šušnjara

University of Split
Croatia
Copyright © 2024 by The Institute of Electrical and Electronics Engineers, Inc. All rights reserved.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey.

Published simultaneously in Canada.

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, scanning, or otherwise,
except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without
either the prior written permission of the Publisher, or authorization through payment of the
appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers,
MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests
to the Publisher for permission should be addressed to the Permissions Department, John Wiley &
Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online
at http://www.wiley.com/go/permission.

Trademarks: Wiley and the Wiley logo are trademarks or registered trademarks of John Wiley &
Sons, Inc. and/or its affiliates in the United States and other countries and may not be used
without written permission. All other trademarks are the property of their respective owners.
John Wiley & Sons, Inc. is not associated with any product or vendor mentioned in this book.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best
efforts in preparing this book, they make no representations or warranties with respect to the
accuracy or completeness of the contents of this book and specifically disclaim any implied
warranties of merchantability or fitness for a particular purpose. No warranty may be created or
extended by sales representatives or written sales materials. The advice and strategies contained
herein may not be suitable for your situation. You should consult with a professional where
appropriate. Further, readers should be aware that websites listed in this work may have changed
or disappeared between when this work was written and when it is read. Neither the publisher
nor authors shall be liable for any loss of profit or any other commercial damages, including
but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please
contact our Customer Care Department within the United States at (800) 762-2974, outside the
United States at (317) 572-3993 or fax (317) 572-4002.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in
print may not be available in electronic formats. For more information about Wiley products,
visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data

Names: Poljak, D. (Dragan), author. | Šušnjara, Anna, author.
Title: Deterministic and stochastic modeling in computational
electromagnetics : integral and differential equation approaches /
Dragan Poljak, Anna Šušnjara.
Description: Hoboken, New Jersey : Wiley, [2024] | Includes index.
Identifiers: LCCN 2023037537 (print) | LCCN 2023037538 (ebook) | ISBN
9781119989240 (hardback) | ISBN 9781119989257 (adobe pdf) | ISBN
9781119989264 (epub)
Subjects: LCSH: Electromagnetism–Mathematical models. | Stochastic models.
Classification: LCC QC760 .P625 2024 (print) | LCC QC760 (ebook) | DDC
537.01/515–dc23/eng/20231017
LC record available at https://lccn.loc.gov/2023037537
LC ebook record available at https://lccn.loc.gov/2023037538

Cover Image and Design: Wiley

Set in 9.5/12.5pt STIXTwoText by Straive, Pondicherry, India

To our beloved ones…
 vii

Contents

About the Authors xv

Preface xvii

Part I Some Fundamental Principles in Field Theory 1

1 Least Action Principle in Electromagnetics 3

1.1 Hamilton Principle 4
1.2 Newton’s Equation of Motion from Lagrangian 7
1.3 Noether’s Theorem and Conservation Laws 8
1.4 Equation of Continuity from Lagrangian 12
1.5 Lorentz Force from Gauge Invariance 16
References 18

2 Fundamental Equations of Engineering Electromagnetics 21

2.1 Derivation of Two-Canonical Maxwell’s Equation 21
2.2 Derivation of Two-Dynamical Maxwell’s Equation 22
2.3 Integral Form of Maxwell’s Equations, Continuity Equations, and
Lorentz Force 25
2.4 Phasor Form of Maxwell’s Equations 27
2.5 Continuity (Interface) Conditions 29
2.6 Poynting Theorem 30
2.7 Electromagnetic Wave Equations 32
2.8 Plane Wave Propagation 35
2.9 Hertz Dipole as a Simple Radiation Source 37
2.9.1 Determination of the Q-Factor 40
2.10 Wire Antennas of Finite Length 41
2.10.1 Dipole Antennas 42
2.10.2 Pocklington Integro-Differential Equation for Straight Thin Wire 43
References 45
viii Contents

3 Variational Methods in Electromagnetics 47

3.1 Analytical Methods 47
3.1.1 Capacity of Insulated Charged Sphere 47
3.1.2 Spherical Grounding Resistance 49
3.2 Variational Basis for Numerical Methods 51
3.2.1 Poisson’s Equation 51
3.2.2 Scalar Potential Integral Equation (SPIE) 52
3.2.3 Correlation Between Variational Principle and Weighted Residual
(Galerkin) Approach 53
3.2.4 Ritz Method 53
References 54

4 Outline of Numerical Methods 57

4.1 Variational Basis for Numerical Methods 60
4.2 The Finite Element Method 61
4.2.1 Basic Concepts of FEM – One-Dimensional FEM 62
4.2.2 Two-Dimensional FEM 66
4.2.3 Three-Dimensional FEM 73
4.3 The Boundary Element Method 77
4.3.1 Constant Boundary Elements 80
4.3.2 Linear and Quadratic Elements 83
4.3.3 Quadratic Elements 84
4.3.4 Numerical Solution of Integral Equations Over Unknown Sources 85
References 86

Part II Deterministic Modeling 87

5 Wire Configurations – Frequency Domain Analysis 89

5.1 Single Wire in the Presence of a Lossy Half-Space 89
5.1.1 Horizontal Dipole Above a Homogeneous Lossy Half-Space 89
5.1.1.1 Integro-differential Equation Formulation 90
5.1.1.2 Numerical Solution of the Pocklington Equation 91
5.1.1.3 Computational Example 95
5.1.2 Horizontal Dipole Buried in a Homogeneous Lossy Half-Space 96
5.1.2.1 Pocklington Integro-differential Equation Formulation 97
5.1.2.2 Numerical Solution of the Pocklington Equation 98
5.1.2.3 Computational Example 99
5.2 Horizontal Dipole Above a Multi-layered Lossy Half-Space 100
5.2.1 Integral Equation Formulation 101
 Contents ix

5.2.2 Radiated Field 106

5.2.3 Numerical Results 108
5.3 Wire Array Above a Multilayer 125
5.3.1 Formulation 129
5.3.2 Numerical Procedures 130
5.3.3 Computational Examples 133
5.4 Wires of Arbitrary Shape Radiating Over a Layered Medium 150
5.4.1 Curved Single Wire in Free Space 154
5.4.2 Curved Single Wire in the Presence of a Lossy Half-space 155
5.4.3 Multiple Curved Wires 156
5.4.3.1 Numerical Solution Procedures 160
5.4.3.2 Computational Examples 162
5.4.4 Electromagnetic Field Coupling to Arbitrarily Shaped Aboveground
Wires 166
5.4.4.1 Formulation via a Set of Coupled Integro-differential Equations 167
5.4.4.2 Numerical Solution of Coupled Pocklington Equations 169
5.4.4.3 Computational Example 170
5.4.5 Buried Wires of Arbitrary Shape 176
5.4.5.1 Formulation 176
5.4.5.2 Numerical Procedure 181
5.4.5.3 Computational Examples 182
5.5 Complex Power of Arbitrarily Shaped Thin Wire Radiating Above
a Lossy Half-Space 186
5.5.1 Theoretical Background 189
5.5.2 Numerical Results 193
References 202

6 Wire Configurations – Time Domain Analysis 207

6.1 Single Wire Above a Lossy Ground 208
6.1.1 Case of Perfectly Conducting (PEC) Ground and Dielectric
Half-Space 213
6.1.2 Modified Reflection Coefficient for the Case of an Imperfect
Ground 214
6.2 Numerical Solution of Hallen Equation via the Galerkin–Bubnov
Indirect Boundary Element Method (GB-IBEM) 222
6.2.1 Computational Examples 226
6.3 Application to Ground-Penetrating Radar 228
6.3.1 Transient Field due to Dipole Radiation Reflected from the Air–Earth
Interface 231
x Contents

6.3.1.1 Numerical Evaluation Procedure 232

6.3.1.2 Numerical Results 234
6.3.2 Transient Field Transmitted into a Lossy Ground Due to Dipole
Radiation 235
6.3.2.1 Numerical Evaluation of the Transmitted Field 241
6.3.2.2 Numerical Results 242
6.4 Simplified Calculation of Specific Absorption in Human Tissue 246
6.4.1 Calculation of Specific Absorption 247
6.4.2 Numerical Results 249
6.5 Time Domain Energy Measures 255
6.6 Time Domain Analysis of Multiple Straight Wires above a Half-Space by
Means of Various Time Domain Measures 260
6.6.1 Theoretical Background 261
6.6.1.1 Time Domain Energy Measures and Power Measure 263
6.6.1.2 Root Mean Square Value of Current Distribution 263
6.6.2 Numerical Results 264
6.6.2.1 Configuration 1 265
6.6.2.2 Configuration 2 267
6.6.2.3 Configuration 3 270
6.6.2.4 Configuration 4 270
6.6.2.5 Configuration 5 272
6.6.2.6 Configuration 6 277
References 280

7 Bioelectromagnetics – Exposure of Humans in GHz Frequency

Range 285
7.1 Assessment of Sab in a Planar Single Layer Tissue 286
7.1.1 Analysis of Dipole Antenna in Front of Planar Interface 287
7.1.2 Calculation of Absorbed Power Density 290
7.1.3 Computational Examples 291
7.2 Assessment of Transmitted Power Density in a Single Layer Tissue 295
7.2.1 Formulation 296
7.2.2 Results for Current Distribution 300
7.2.2.1 Results for Transmitted Field, VPD, and TPD 301
7.2.2.2 Different Distance from the Interface 302
7.2.2.3 Different Antenna Length 309
7.2.2.4 Different Frequencies 316
7.3 Assessment of Sab in a Multilayer Tissue Model 318
7.3.1 Theoretical Background 318
 Contents xi

7.3.2 Results 320

7.4 Assessment of Transmitted Power Density in the Planar Multilayer
Tissue Model 325
7.4.1 Formulation 325
7.4.2 Results 327
7.4.2.1 Two-Layer Model 327
7.4.2.2 Three-Layer Model 329
7.4.2.3 Skin Depth and Saturation Depth 333
References 337

8 Multiphysics Phenomena 339

8.1 Electromagnetic-Thermal Modeling of Human Exposure to HF
Radiation 340
8.1.1 Electromagnetic Dosimetry 340
8.1.2 Thermal Dosimetry 342
8.1.3 Computational Examples 346
8.2 Magnetohydrodynamics (MHD) Models for Plasma
Confinement 348
8.2.1 The Grad-Shafranov Equation 348
8.2.1.1 Analytical Solution 350
8.2.1.2 Analytical Results 353
8.2.1.3 Solution by the Finite Difference Method (FDM) 354
8.2.1.4 Solution by the Finite Element Method (FEM) 356
8.2.1.5 Computational Examples 358
8.2.2 Transport Phenomena Modeling 359
8.2.2.1 Transport Equations 361
8.2.2.2 Current Diffusion Equation and Equilibrium in Tokamaks 361
8.2.2.3 FEM Solution of CDE 364
8.2.2.4 Analytical Solution Procedure 367
8.2.2.5 Numerical Results 368
8.3 Modeling of the Schrodinger Equation 370
8.3.1 Derivation of the Schrodinger Equation 371
8.3.2 Analytical Solution of the Schrodinger Equation 372
8.3.3 FDM Solution of the Schrodinger Equation 373
8.3.4 FEM Solution of the Schrodinger Equation 374
8.3.5 Neural Network Approach to the Solution of the Schrodinger
Equation 375
References 381
xii Contents

Part III Stochastic Modeling 385

9 Methods for Stochastic Analysis 387

9.1 Uncertainty Quantification Framework 388
9.1.1 Uncertainty Quantification (UQ) of Model Input Parameters 388
9.1.2 Uncertainty Propagation (UP) 389
9.1.3 Monte Carlo Method 391
9.2 Stochastic Collocation Method 393
9.2.1 Computation of Stochastic Moments 393
9.2.2 Interpolation Approaches 394
9.2.3 Collocation Points Selection 396
9.2.4 Multidimensional Stochastic Problems 397
9.2.4.1 Tensor Product 397
9.2.4.2 Sparse Grids 398
9.2.4.3 Stroud’s Cubature Rules 399
9.3 Sensitivity Analysis 402
9.3.1 “One-at-a-Time” (OAT) Approach 402
9.3.2 ANalysis Of VAriance (ANOVA)-Based Method 403
References 404

10 Stochastic–Deterministic Electromagnetic Dosimetry 407

10.1 Internal Stochastic Dosimetry for a Simple Body Model Exposed to
Low-Frequency Field 408
10.2 Internal Stochastic Dosimetry for a Simple Body Model Exposed to
Electromagnetic Pulse 413
10.3 Internal Stochastic Dosimetry for a Realistic Three-Compartment
Human Head Exposed to High-Frequency Plane Wave 417
10.4 Incident Field Stochastic Dosimetry for Base Station Antenna
Radiation 423
References 430

11 Stochastic–Deterministic Thermal Dosimetry 433

11.1 Stochastic Sensitivity Analysis of Bioheat Transfer Equation 434
11.2 Stochastic Thermal Dosimetry for Homogeneous Human Brain 437
11.3 Stochastic Thermal Dosimetry for Three-Compartment Human
Head 447
11.4 Stochastic Thermal Dosimetry below 6 GHz for 5G Mobile
Communication Systems 450
References 457
 Contents xiii

12 Stochastic–Deterministic Modeling in Biomedical Applications of

Electromagnetic Fields 459
12.1 Transcranial Magnetic Stimulation 460
12.2 Transcranial Electric Stimulation 466
12.2.1 Cylinder Representation of Human Head 467
12.2.2 A Three-Compartment Human Head Model 469
12.2.3 A Nine-Compartment Human Head Model 472
12.3 Neuron’s Action Potential Dynamics 481
12.4 Radiation Efficiency of Implantable Antennas 488
References 498

13 Stochastic–Deterministic Modeling of Wire Configurations in Frequency

and Time Domain 503
13.1 Ground-Penetrating Radar 503
13.1.1 The Transient Current Induced Along the GPR Antenna 504
13.1.2 The Transient Field Transmitted into a Lossy Soil 508
13.2 Grounding Systems 515
13.2.1 Test Case #1: Soil And Lighting Pulse Parameters are Random
Variables 518
13.2.2 Test Case #2: Soil and Electrode Parameters are Random Variables 519
13.2.3 Test Case #3: Soil, Electrode, and Lighting Pulse Parameters are Random
Variables 520
13.3 Air Traffic Control Systems 523
13.3.1 Runway Covered with Snow 526
13.3.2 Runway Covered with Vegetation 529
References 530

14 A Note on Stochastic Modeling of Plasma Physics Phenomena 535

14.1 Tokamak Current Diffusion Equation 535
References 543

Index 545
 xv

About the Authors

Dragan Poljak (Senior Member, IEEE) received the Ph.D. degree in electrical
engineering from the University of Split, Croatia, in 1996. He is currently a full
professor with the Department of Electronics and Computing, University of Split.
He is also involved in ITER Physics EUROfusion Collaboration and in the
Croatian Center for Excellence in Research for Technology Sciences. He has
published more than 160 journals and 250 conference papers and authored some
books, e.g. two by Wiley, Hoboken, NJ, USA, and one by Elsevier, St. Louis, MO,
USA. His research interests include computational electromagnetics (electromag-
netic compatibility, bioelectromagnetics, and plasma physics). From May 2013 to
June 2021, he was a member of the Board of the Croatian Science Foundation. He
is a member of the Editorial Board of Engineering Analysis with Boundary
Elements, Mathematical Problems in Engineering, and IET Science, Measurement
and Technology. He was awarded several prizes for his achievements, such as the
URSI Young Scientists Award in 1999, the National Prize for Science in 2004, the
Croatian Section of IEEE Annual Award in 2016, the Technical Achievement
Award of the IEEE EMC Society in 2019, and the George Green Medal from
the University of Mississippi in 2021. He is active in a few working groups of
the IEEE/International Committee on Electromagnetic Safety (ICES) Technology
Committee 95 SC6 EMF Dosimetry Modeling.

Anna Šušnjara received her PhD degree in electrical engineering from the
University of Split, Croatia, in 2021. She is currently a postdoc researcher at the
Faculty of Electrical Engineering, Mechanical Engineering and Naval Architec-
ture, University of Split. Her research interests include numerical modeling,
uncertainty quantification, and sensitivity analysis in computational electromag-
netics. Dr. Šušnjara is involved in ITER physics EUROfusion collaboration. From
2015 to 2021, she was a member of EUROfusion work package for code develop-
ment in European Transport Solver (ETS), while from 2021, she has been a
member of IFMIF-DONES project. Dr. Šušnjara is a member of IEEE and BIOEM
xvi About the Authors

societies. She currently serves as Vice President of IEEE EMC Croatian chapter. To
date, Dr. Šušnjara has (co)authored 19 journal and more than 40 conference
papers. She serves as a reviewer for seven journals and two conferences.
Dr. Šušnjara gave lectures about computational electromagnetics at several
European academic institutions and tutorials at international scientific
conferences. She was awarded several prizes for her achievements. In 2023, she
received the URSI Young Scientist Award at the 35th URSI GASS in Sapporo,
Japan. She was also awarded the National Prize for Science and the University
of Split Prize for Science in 2021 and 2022, respectively. In 2016, she received
the best poster paper award at BioEM conference in Ghent, Belgium, and spent
one month at Politecnico di Torino as SPI2016 Young Investigator Training
Program awardee.
 xvii

Preface

Most of the computational models used in engineering electromagnetics are

deterministic in nature, i.e. one deals with an exact set of input data in a sense
of either material properties or geometry. However, in many scenarios, there
are problems with uncertainty in the input data set as some system properties
are partly or entirely unknown. Therefore, a stochastic approach is required to
determine the relevant statistics about the given responses, thus providing the
assessment of the related confidence intervals in the set of numerical results
obtained as an output of a given deterministic model. Of particular interest are
nonintrusive stochastic approaches that could be easily coupled with widely used
well-established deterministic models, by efficiently postprocessing numerical
results arising from deterministic models.
The goal of this book is to demonstrate the efficiency of parallel use of determin-
istic and stochastic models featuring combination of well-established analytical/
numerical methods with stochastic analysis techniques. The nonintrusive stochas-
tic approach presented in the book can be readily incorporated into majority of
computational electromagnetics (CEM) models with little effort aiming to quickly
provide a more detailed insight into the relationship between the input parameters
and the output of interest.
A variety of examples throughout the book are presented to clearly demonstrate
the efficiency of deterministic-stochastic approaches in CEM models, and a refer-
ence list is given at the end of each chapter. The book provides computational
examples illustrating successful application of stochastic collocation (SC) tech-
nique in the areas of ground-penetrating radars (GPRs), grounding systems, radi-
ation from 5G systems, human exposure to electromagnetic fields, transcranial
magnetic stimulation (TMS), transcranial electric stimulation (TES), transient
analysis of buried wires, and design of instrumental landing system (ILS).
The book is divided into three parts. Part I outlines the fundamentals of
classical electromagnetics and basics of numerical modeling. Part II deals with
deterministic models pertaining to analysis of thin wires in both frequency and
xviii Preface

time domains, human exposure to electromagnetic fields in GHz frequency range,

and multiphysics phenomena such as plasma confinement in tokamak. Finally,
Part III is entirely devoted to stochastic modeling covering a detailed description
of SC method and sensitivity analysis. Part III also contains a number of applica-
tions arising from electromagnetic-thermal dosimetry, biomedical applications,
electromagnetic compatibility (EMC), GPRs, grounding systems, air traffic control
systems, and transport phenomena in tokamak.
The material given in this book is dominantly based on papers previously pub-
lished by the authors, suitably modified to present the results in a uniform design
and format. Additional related material is planned to be prepared aiming to com-
plete some details and extensions of the published work. An extensive reference
list of other related work is also included. The goal is to provide a reference on
the deterministic-stochastic modeling in different areas of CEM covering fre-
quency and time domain analyses of wire antennas and their various applications,
EMC, computational models of lines and cables, lightning, grounding systems,
GPRs, magnetohydrodynamics, bioelectromagnetics, and biomedical applications
of electromagnetic fields.
As the book covers multidisciplinary phenomena, such as electromagnetic-
thermal dosimetry, magnetohydrodynamics, and plasma physics, the authors
hope it could be of interest to multidisciplinary researchers, engineers, physicists,
and mathematicians.

The Authors
Split, Croatia, 2023
 1

Part I

Some Fundamental Principles in Field Theory

 3

Least Action Principle in Electromagnetics

Laws of nature are governed by following fundamental principles – the action

principle, locality, Lorentz invariance, and gauge invariance [1]. Hamilton’s prin-
ciple, or the least action principle, is originally developed for classical mechanics
stating that a particle, among all of the trajectories between fixed time instants t1
and t2, follows the path which minimizes the action. Action is defined as time inte-
gral of the difference between the kinetic energy and potential energy, respec-
tively. Thus, Hamilton’s principle somehow requires the time averages of the
kinetic energy and potential energy to distribute as equally as possible (equiparti-
tion) [2]. In classical mechanics, Hamilton’s principle and Newton’s second law
represent equivalent formulations.
An extension of Hamilton’s principle from classical mechanics to classical elec-
tromagnetics can be undertaken starting with the analysis of the motion of single
charged particle [3]. Next step is to construct a Lagrangian for the electromagnetic
field by extending the Lagrangian pertaining to classical mechanics. From the cor-
responding Lagrangians, featuring Noether’s theorem and gauge invariance, it is
possible to derive equation of continuity for the charge, Lorentz force, and
Maxwell’s equations, which can be found elsewhere, e.g. [2–5].
Generally, when a functional is extremal, Noether’s theorem yields the conser-
vation law. Thus, invariance of the system under a time translation results in the
energy conservation. It is also worth noting that space translation invariance
corresponds to the conservation of linear momentum, rotation invariance corre-
sponds to the conservation of angular momentum, while gauge invariance yields
the charge conservation [1, 2].
These derivations are recently reviewed in [6–8].
This chapter first deals with derivation of continuity equation and Lorentz
force, and a derivation of Maxwell’s equations from the electromagnetic field

Deterministic and Stochastic Modeling in Computational Electromagnetics: Integral and Differential

Equation Approaches, First Edition. Dragan Poljak and Anna Šušnjara.
© 2024 The Institute of Electrical and Electronics Engineers, Inc.
Published 2024 by John Wiley & Sons, Inc.
4 1 Least Action Principle in Electromagnetics

functional is carried out. Finally, a variational basis of numerical solution

methods in electromagnetics is discussed.

1.1 Hamilton Principle

Hamilton variational principle represents not only the basis of modern analytical
dynamics but also of universal physical laws, i.e. fundamental laws of classical
physics can be understood in terms of action.
This section first deals with Hamilton’s variational principle in mechanics,
Newton’s equation of motion, and Noether’s theorem. Then the variational
principle in electromagnetics is discussed.
For simplicity, a system with one degree of freedom represented by a generalized
coordinate q is considered together with related function of position, velocity, and
time L q, q, t where q denotes the time derivative.
The task is to determine how a point particle should move in this one-
dimensional space so that the time integral of L is minimized compared with
the integral over the conceivable paths between the same starting and end points,
as depicted in Fig. 1.1. The solution is given by stating q as a function of time
q = q(t).
To compare all paths having the same starting and end points, the variation of
function q is zero at both ends [1–4]
δq = 0 11
i.e. all alternatives start at instant t1 and arrive together at instant t2.
The minimum condition is then given by a functional F expressed in terms of the
integral [1–4]

q(t) Figure 1.1 The varied function q(t).

δq = 0
q2

δq = 0
q1 δq ≠ 0

t1 t2 t
 1.1 Hamilton Principle 5

t2

F= L q, q, t dt = min 12
t1

In classical mechanics, function L is referred to as Lagrangian and is expressed as

L = W kin − W pot 13

where Wkin and Wpot are the kinetic energy and potential energy, respectively.
According to the calculus of variation, the functional approaches minimum value
t2

F= W kin − W pot dt = min 14

t1

when its variation vanishes, i.e.:

δF = 0 15

which can also be written as

t2

δ Ldt = 0 16
t1

Therefore, function q(t) minimizes functional (1.2) or (1.4), respectively, and it

follows
t2

δF = δLdt 17
t1

For the simplest case given by L q, q, t the variation of function L is given by

∂L ∂L
δL = δq + δq 18
∂q ∂q

And by performing some further mathematical manipulation, one readily

obtains
t2
∂L d ∂L ∂L t2
δF = − δqdt + δq t1 =0 19
∂q dt ∂q ∂q
t1

As δq = 0 at the ends of the path, the second term at the right-hand side auto-
matically vanishes.
6 1 Least Action Principle in Electromagnetics

Furthermore, according to the fundamental lemma of variational calculus [4]

the first integral term at the right-hand side of (1.9) vanishes if the following con-
dition is satisfied

∂L d ∂L
− =0 1 10
∂q dt ∂q

It is worth noting that the second order differential Eq. (1.10) relates the position
q for the time t, and also determines the true path of the system when two end
positions and times are given. This equation is known as Lagrange–Euler equation
of motion [1, 2].
In the case of a single function of multiple variables, L qk , qk , ξ, t (1.10) becomes

∂L ∂L ∂L ∂ ∂L
− − =0 1 11
∂q ∂ξ ∂qξ ∂t ∂qt

Finally, if few functions of multiple independent variables L qk , qk , x, y, z, t are

considered, it follows
N
∂L ∂ ∂L ∂ ∂L ∂ ∂L ∂ ∂L
− − − =0
k=1
∂qk ∂x ∂ qk x ∂y ∂ qk y ∂z ∂ qk z ∂t ∂ qk t

1 12

which, for example, pertains to three-dimensional problems in electromagnetics.

Hamilton variational principle can be considered a general law not only for
particle dynamics but also for the dynamics of continuous materials.
An extension to three-dimensional problems in continuous materials, i.e. for
physical fields, the variational principle corresponding to Eq. (1.6) is given by

t2

δ Ld dVdt = 0 1 13
t1 V

where Ld is the so-called Lagrange density defined as

L = Ld dV 1 14
V

and has a unit of energy per volume.

It is worth noting that the variational principle is an invariant for coordinate
transformations [1].
 1.2 Newton’s Equation of Motion from Lagrangian 7

1.2 Newton’s Equation of Motion from Lagrangian

Lagrangian in classical mechanics for a particle with mass m with displacement r

dr
at time t L0 r , ,t is of the form
dt
2
dr 1 dr
L0 r , ,t = m − W pot r 1 15
dt 2 dt

dr
where is the particle velocity and W pot r stands for potential energy (not due
dt
to electromagnetic field).
The corresponding action F0 related to Lagrangian (1.15) is given by integral
t2

F0 = L0 dt 1 16
t1

Now varying the action (1.16)

δF 0 = 0 1 17
It can be written as:
∂L ∂W pot ∂L ∂W pot ∂L ∂W pot
= − , = − , = − 1 18
∂x ∂x ∂y ∂y ∂z ∂z
∂L ∂L ∂L
= mx, = my, = mz 1 19
∂x ∂y ∂z
and it follows:
∂W pot d
− − mx = 0
∂x dt
∂W pot d
− − my = 0 1 20
∂y dt
∂W pot d
− − mz = 0
∂z dt
which simply gives:

d2 mx ∂W pot
= −
dt 2 ∂x
d2 my ∂W pot
= − 1 21
dt 2 ∂y
d mz
2
∂W pot
= −
dt 2 ∂z
8 1 Least Action Principle in Electromagnetics

If mass m is regarded as constant quantity, right-hand side of (1.21) represents

the components of mechanical force:

d2 x
F mech,x = m
dt 2
d2 y
F mech,y = m 1 22
dt 2
d2 z
F mech,z = m
dt 2
As the mechanical force is defined as a negative gradient of the potential energy

F mech = − ∇W pot 1 23

one simply obtains a set of Newton’s equations of motion

d2 r
m = − ∇W pot r 1 24
dt 2
where the right-hand side of (1.24) represents a force acting on the particle.

1.3 Noether’s Theorem and Conservation Laws

There are physical quantities that do not change throughout the time development
of physical systems. These quantities are stated to be conserved under certain con-
ditions which are governed by conservation laws. It can be shown that conserva-
tion laws are a consequence of the symmetry properties of a physical system
(invariance properties of a system under a group of transformations [2, 9]). The
symmetry properties of the system and conservation laws are connected with
Noether’s theorem, e.g. [2, 9].
Let uk(x) (k = 1,2,…n) be a set of differentiable functions of the independent var-
iable x and let vk(x) be the first derivatives, i.e. it can be written as
duk
vk x = 1 25
dx
Now Lagrangian L is defined as a function of x, and n functions of uk and n func-
tions of vk.
L = L x, uk , vk 1 26
If one considers an infinitesimal transformation T
x x = x + δx 1 27
 1.3 Noether’s Theorem and Conservation Laws 9

Furthermore, under T, one consequently has:

uk x uk x = uk x + δuk x 1 28
vk x vk x = vk x + δvk x 1 29

Now assuming the functional

S = Ldx 1 30
Ω

to be invariant under T so that the transformation (1.30) which maps the interval Ω
into Ω does not change, it follows

δS = L dx − Ldx = 0 1 31
Ω Ω

Performing some mathematical manipulations, one obtains the following

expression

d ∂L ∂L d ∂L
− δuk − vk δx = Lδx + δuk − vk δx 1 32
dx ∂vk ∂uk dx ∂vk

Finally, Noether’s theorem states that if functional S is invariant under the infin-
itesimal one-parameter group of transformations T, then the set of n equations
d ∂L ∂L
− =0 1 33
dx ∂vk ∂uk
simply gives

d ∂L
Lδx + δuk − vk δx =0 1 34
dx ∂vk

i.e. it can be written as

∂L
Lδx + δuk − vk δx = const 1 35
∂vk

namely, the expression is independent of x variable.

On the other hand, Newton’s equation of motion can be expressed in terms of
Lagrangian being a function of the generalized coordinates qk (k = 1, 2, …, n) and
their derivatives, i.e.

d ∂L ∂L
− =0 1 36
dt ∂qk ∂q
10 1 Least Action Principle in Electromagnetics

which is equivalent to the request

t2

δ L dt = 0 1 37
t1

Furthermore, Hamilton’s principle (the least action principle), which states

that solutions qk of the system of mass points for which the integral in (C6) is
an extremum contrary to all functions qk

qk t = qk t + δqk t 1 38

If one compares (1.34) with (1.35) with δx = 0 and δuk = 0, it follows that these
statements are identical Namely, it can be written as

∂L
Lδt + δqk − qk δt = const 1 39
∂qk

i.e. the left-hand side is a constant of motion.

As any time translation is expressed as δt 0 and δqk = 0, the following expres-
sion is obtained

∂L
H= q − L = const 1 40
∂qk k

stating that Hamiltonian H which is equal to the total energy of the system (sum of
kinetic energy and potential energy) is constant of the motion.
In other words, time translation symmetry corresponds to the conservation of
energy. Therefore, Noether’s theorem states that for a system with time translation
symmetry, the energy of the system is conserved.
The energy conservation can be demonstrated by using Lagrangian and
Hamiltonian in classical mechanics for one-dimensional case.
Thus, Lagrangian is of the form
1 2
L= mx − W pot x 1 41
2
where
dx
x= 1 42
dt
while Hamiltonian of the system, according to (1.40), is then

∂L
H= x−L 1 43
∂x
 1.3 Noether’s Theorem and Conservation Laws 11

As from (1.42), one simply obtains

∂L
x = mx 1 44
∂x
and the Hamiltonian now can be written as

1 2
H= mx + W pot x 1 45
2
For a system with a time translation symmetry

t t = t + δt 1 46

the Hamiltonian of the system remains invariant.

To check if the Hamiltonian is a conserved quantity under a time translation, it
should be differentiated with respect to time, i.e. it follows

dH d ∂L d ∂L dL
= x−L = x − 1 47
dt dt ∂x dt ∂x dt

and, one has

d ∂L d ∂L ∂L dx d ∂L ∂L
x = x+ = x+ x 1 48
dt ∂x dt ∂x ∂x dt dt ∂x ∂x

dL ∂L dx ∂L dx ∂L ∂L ∂L ∂L
= + + = x+ x+ 1 49
dt ∂x dt ∂x dt ∂t ∂x ∂x ∂t
Now, inserting (1.48) and (1.49) into (1.47) yields

dH d ∂L ∂L ∂L ∂L ∂L
= x+ x− x+ x+ 1 50
dt dt ∂x ∂x ∂x ∂x ∂t

and one obtains

dH d ∂L ∂L ∂L
= x− x− 1 51
dt dt ∂x ∂x ∂t

which can be written:

dH d ∂L ∂L ∂L
= − x− 1 52
dt dt ∂x ∂x ∂t

Now, as Euler Lagrange equation vanishes

d ∂L ∂L
− =0 1 53
dt ∂x ∂x
12 1 Least Action Principle in Electromagnetics

and (1.52) becomes

dH ∂L
= − 1 54
dt ∂t
Now, considering Lagrangian (1.41) it is evident that its time derivative is zero,
as there is no explicit time dependence. Consequently, the total time derivative of
the Hamiltonian is also zero, and it can be concluded that Hamiltonian is con-
served under the time translation (1.46).
It is worth noting that space-time invariance corresponds to the conservation of
linear momentum, and rotation invariance corresponds to the conservation of
angular momentum.

1.4 Equation of Continuity from Lagrangian

A change in system which does not affect the action integral, or the equation of
motions, is referred to as invariance or symmetry [1].
Lagrangian L in classical mechanics is defined as difference between kinetic
energy (Wkin) and potential energy (Wpot) of the system
L = W kin − W pot 1 55

while the corresponding Euler–Lagrange equation of motion governing the con-

sidered trajectory of a particle is given by
∂L d ∂L
− =0 1 56
∂q dt ∂q
where q pertains to the position of the particle and q denotes its time derivative.
Furthermore, Lagrangian for particle of mass m with displacement r at time t
dr
L r, ,t is given by [3, 4]
dt
2
dr 1 dr 1
L r, ,t = m − W pot r = mr − W pot r 1 57
dt 2 dt 2

dr
where = r is the particle velocity and W pot r stands for potential energy (in
dt
the absence of electromagnetic field).
The corresponding action F related to Lagrangian (1.57) is given by integral
t2

F= Ldt 1 58
t1
 1.4 Equation of Continuity from Lagrangian 13

Now varying action (1.58)

δF = 0 1 59
yields Newton’s second law:
d2 r
m = − ∇W pot r 1 60
dt 2
where the right-hand side represents a force acting on the particle.
Now, according to the symmetry of the Lagrangian stemming from Noether’s
theorem [2, 3, 9], a total time derivative may be added to Lagrangian (1.57) without
changing the equation of motion [3].
Thus, choosing an arbitrary differentiable scalar function qΛ r , t , it can be
written as
dΛ
L =L+q 1 61
dt
where q stands for the electric charge.
Now, a new action F can be defined as follows
t2

F = L dt 1 62
t1

and using (1.61), it follows

t2 t2 t2
dΛ dΛ
F = L+q dt = Ldt + q dt 1 63
dt dt
t1 t1 t1

Furthermore, one obtains

F = F + q Λ r t2 , t2 − Λ r t1 , t1 1 64

As the end points of the interval are fixed

δr = 0 1 65
performing the variation of functional F’ simply yields
δF = δF 1 66
Therefore, the Lagrangian L and L are equivalent.
For the point particle, the charge density can be written as

ρ R, t = qδ R − r t 1 67

where δ [ ] is the Dirac delta function.

14 1 Least Action Principle in Electromagnetics

In the next step, the current density can be expressed as charge in motion, i.e.

dr t
J R, t = q δ R− r t 1 68
dt

Now, taking into account the total differential of scalar function Λ r , t

∂Λ ∂Λ ∂Λ ∂Λ ∂Λ
dΛ = dx + dy + dz + dt = ∇Λ d r + dt 1 69
∂x ∂y ∂z ∂t ∂t

total time derivative is obtained as

dΛ dr ∂Λ ∂Λ d r ∂Λ
=∇ Λ + = r0 + 1 70
dt dt ∂t ∂n dt ∂t
and the Lagrangian can be written as follows

∂Λ dr ∂Λ
L =L+q + 1 71
∂n dt ∂t

Integrating over volume V yields

ρ R, t dV = qδ R − r t dV 1 72
V V

and one simply has

q = ρ R, t dV 1 73
V

Finally, Lagrangian (1.61) takes the form

dr ∂Λ R, t
L =L+ ρ R, t ∇Λ R, t dV + ρ R, t dV 1 74
dt ∂t
V V

which, taking into account (1.68), becomes

∂Λ R, t
L =L+ J R, t ∇Λ R, t dV + ρ R, t dV 1 75
∂t
V V

If V tends to infinity and current is assumed to vanish at infinity after performing

integration by parts, it follows
 1.4 Equation of Continuity from Lagrangian 15

d
L = L − ∇ J R, t Λ R, t dV + ρ R, t Λ R, t dV
dt
V V
1 76
∂ρ R, t
− Λ R, t dV
∂t
V

which now can be written as

∂ρ R, t d
L = L− ∇ J R, t − Λ R, t dV + ρ R, t Λ R, t dV
∂t dt
V V

1 77
Now Lagrangian L can be equivalent to (1.57) if the second term of the right-
hand side is equal to zero.
According to the fundamental lemma of variational calculus, as Λ R, t is an
arbitrary space-dependent function, the second integral term from the right-hand
side vanishes identically if the following condition is satisfied

∂ρ R, t
∇ J R, t − =0 1 78
∂t
Expression (1.78) is a well-known equation of continuity relating to charge and
current density, respectively.
Therefore, one of the basic equations of classical electromagnetics is derived
from the gauge invariance of classical mechanics, i.e. from the symmetry property
of the Lagrangian.
Namely, the gauge invariance of the Lagrangian implies conservation of electric
charge.
It is worth noting that most of standard textbooks used for various courses in
electromagnetics start from Maxwell’s equations and the continuity equation is
usually derived from Maxwell’s equations (which are considered rigorous mathe-
matical expressions of previously discovered laws as a consequence of experi-
ments) via certain mathematical manipulations.
This chapter exploits the approach that goes the other way around by exploiting
the symmetry properties of the Lagrangian, similar to the approach presented in
[3, 6–8].
It is rather worth stressing that within the framework of such an approach, elec-
tromagnetic potentials are regarded as more fundamental entities than in the case
in which one considers corresponding equation of motions instead of Lagrangian.
Namely, in latter approach, starting from the electric and magnetic fields being
16 1 Least Action Principle in Electromagnetics

gauge invariant, potentials are no more than auxiliary functions, i.e. pure math-
ematical constructs with no physical meaning.

1.5 Lorentz Force from Gauge Invariance

In the next step, Lagrangian (1.75) is rewritten by introducing vector function

A = ∇Λ R, t 1 79

and scalar function

∂Λ R, t
φ= − 1 80
∂t
which now gives new Lagrangian

Li = L + J R, t A R, t dV − ρ R, t φ R, t dV 1 81
V V

Therefore, current density J R, t is coupled to the vector potential A R, t ,

while charge density is coupled to the scalar potential φ R, t [3, 4].

Therefore, integral terms in (1.81) describe interaction of the charged particle
with the electromagnetic field potentials.
Again, to obtain an equivalent Lagrangian, the total time derivative is added
dΛ
Li = Li + q 1 82
dt
and the obtained Lagrangian is
dΛ
Li = L + J AdV − ρφdV + q 1 83
dt
V V

which can be also written as

L =L+ J A dV − ρφ dV 1 84
V V

If the potentials are defined as follows [2, 3]

A = A + ∇Λ R, t 1 85
 1.5 Lorentz Force from Gauge Invariance 17

∂Λ R, t
φ = φ− 1 86
∂t
The vector and scalar potentials in (1.83) correspond to the additional terms
in (1.75).
Expressions (1.85, 1.86) are referred to as gauge transformations added to ensure
invariance of the Lagrangian under the addition of a total time derivative.
The corresponding equation of motions of the particle interacting with electro-
magnetic field can now be obtained from the least action principle.
The total action for the charged particle Fi can be written as
t2

Fi = Li dt 1 87
t1

Varying functional (1.87) and taking into account

dA ∂A dx ∂A dy ∂A dz ∂A ∂A
= + + + = v ∇ A+ 1 88
dt ∂x dt ∂y dt ∂z dt ∂t ∂t
leads to the following equation of motion [3]

d2 r ∂A dr
m = − ∇W pot r + q − ∇φ − +q × ∇×A 1 89
dt 2 ∂t dt

representing Newton’s second law with additional forces due to the existence of
the electromagnetic field.
Note that (1.89) is equivalent to expression

d2 r ∂A dr
m = − ∇W pot r + q − ∇φ − +q × ∇×A 1 90
dt 2 ∂t dt

The vector and scalar potentials are not unique and consequently not measur-
able physical quantities aiming as they are mathematical constructs and do not
represent physical fields [1, 3, 4].
However, the electric field and magnetic field can be defined in terms of poten-
tials A and φ as they are invariant under gauge transformations (1.85) and (1.86) as
follows [3]:

B=∇×A 1 91

∂A
E = − ∇φ − 1 92
∂t
where B denotes the magnetic field and E stands for the electric field.
18 1 Least Action Principle in Electromagnetics

It can be concluded that electric and magnetic fields are obtained from the equa-
tions of motion for charge particles and they are gauge invariant.
The total equation of motion can be finally written as follows

d2 r dr
m = − ∇W pot r + qE + q ×B 1 93
dt 2 dt
Equation (1.93) represents Newton’s second law of motion with the Lorentz
force included, i.e. it can be written as

d2 r
m = F mech + F EM 1 94
dt 2

where F mech is the force acting to the particle due to the potential energy

F mech = − ∇W pot 1 95

while the Lorentz force due to the interaction of particles with the electromagnetic
field is

F EM = q E + v × B 1 96

Therefore, (1.96) is the second fundamental equation of electromagnetics

derived from the gauge invariance starting from action in classical mechanics.
Namely, according to the gauge invariance, vector potential can be changed
without any effect on the charged particle. This can be referred to as gauge
transformation.
It is also worth noting that Lorentz force (1.96), being a basis of classical particle
electromagnetics, cannot be obtained from Maxwell’s equations for stationary
media, while (1.78, 1.91, 1.92) can be readily derived from Maxwell’s equation,
which is available elsewhere, e.g. in [4].

References
1 L. Suskind and A. Friedman, Special Relativity and Classical Field Theory, New
York, UK: Penguin, 2017.
2 D. E. Neuenschwander, Emmy Noether’s Wonderful Theorem, Baltimore, USA: The
Johns Hopkins University Press, 2011.
3 D. H. Cobe, “Derivation of Maxwell’s equations from the gauge invariance of
classical mechanics,” American Journal of Physics, vol. 48, no. 348, pp. 348–353, 1980.
4 D. Poljak, Advanced Modeling in Computational Electromagnetic Compatibility,
New Jersey, USA: John Wiley & Sons, Inc., 2007.
 References 19

5 C. Civelek and T. F. Bechteler, Lagrangian formulation of electromagnetic fields in

nondispersive medium by means of the extended Euler–Lagrange differential
equation, International Journal of Engineering Science, vol. 46, no. 12, pp. 1218–
1227, 2008.
6 D. Poljak, “Review of Least Action Principle in Electromagnetics, Part II: Derivation
of Continuity Equation and Lorentz Force, Maxwell’s Equations”, in 2022
International Conference on Software, Telecommunications and Computer Networks
(SoftCOM), Split, Croatia, SoftCOM, 2022.
7 D Poljak, “Review of Least Action Principle in Electromagnetics, Part II: Derivation
of Maxwell’s Equations,” SoftCOM, 2022.
8 D. Poljak, “Review of Least Action Principle in Electromagnetics, Part III:
Variational Basis for Numerical methods,” SoftCOM, 2022.
9 T. B. Mieling, “Noether’s Theorem Applied to Classical Electrodynamics,” Lecture
Notes, November 2017.
 21

Fundamental Equations of Engineering Electromagnetics

This chapter deals with a derivation of Maxwell’s equations from Hamilton’s

principle in electromagnetics. It is well known that electromagnetic fields can
essentially be treated as mechanical systems [1, 2]. As in Chapter 1, Hamilton’s
principle has been used to derive continuity equation and Lorentz force law, this
variational principle should also provide a single equation from which Maxwell’s
equation can be deduced. Namely, as Hamilton’s principle is applicable to contin-
uous media, it can also be used to describe the behavior of fields, i.e. it should be
possible to set up a suitable Lagrangian function of the field variables, such that
Hamilton’s principle yields the equations of motion of the field – Maxwell’s
equations.
Though there is no general rule for the derivation of the Lagrangian function,
the electromagnetic energy density is defined as a function of position in space
in terms of the field variables.
An intuitive approach would be to correlate the kinetic field energy with mag-
netic field energy and the potential energy with the electric field energy [1, 2].
Kinematical Maxwell’s equations are derived from gauge symmetry, while two
dynamical Maxwell’s equations are derived from functional of electromagnetic
energy.

2.1 Derivation of Two-Canonical Maxwell’s Equation

It has been shown in Section 1.5, featuring the gauge symmetry, that electric and
magnetic fields can be expressed in terms of the vector and scalar potentials
regardless of the fact that these potentials are not unique and thus not measurable.
Namely, these potentials do not represent physical quantities [3].

Deterministic and Stochastic Modeling in Computational Electromagnetics: Integral and Differential

Equation Approaches, First Edition. Dragan Poljak and Anna Šušnjara.
© 2024 The Institute of Electrical and Electronics Engineers, Inc.
Published 2024 by John Wiley & Sons, Inc.
22 2 Fundamental Equations of Engineering Electromagnetics

On the other hand, the electric and magnetic fields can be defined in terms of
these mathematical constructs which are invariant under gauge transformations
[3], i.e. it can be written as

B=∇×A 21
∂A
E = − ∇φ − 22
∂t
where A and φ stand for magnetic vector and electric scalar potential, respectively.
Now the two kinematical equations that electric and magnetic fields satisfy can
be obtained from Eqs (2.1) and (2.2) in which electric and magnetic fields are
expressed in terms of potentials.
Thus, taking the curl of (2.1), the right-hand side vanishes

∇ B=∇ ∇×A =0 23

and one obtains the Gauss’s law for the magnetic field in a differential form

∇ B=0 24
i.e. the fourth Maxwell’s equation states that no magnetic monopoles exist.
Furthermore, taking the curl of (2.2), the first term of the right-hand side
vanishes

∂A ∂
∇ × E = ∇ × − ∇φ − = − ∇×A 25
∂t ∂t

One obtains Faraday law in the differential form

∂B
∇×E= − 26
∂t
i.e. the first curl Maxwell’s equation.
So, the kinematical Maxwell’s equations are obtained from the electric and mag-
netic fields expressed in terms of vector and scalar potentials featuring the use of
gauge symmetry.

2.2 Derivation of Two-Dynamical Maxwell’s Equation

It has been shown in [3] that Lagrangian for the charged particle is given in
the form
1 2
L= mr − W pot + J A − ρ ϕ dV 27
2
V
 2.2 Derivation of Two-Dynamical Maxwell’s Equation 23

Now this difference between kinetic energy and potential energy can be written
by relating kinetic energy with magnetic field energy and potential energy with the
electric field energy, i.e. the new Lagrangian is given by [1, 2]

1
L= B H − D E + J A − ϕ ρ dV = Ld dV 28
2
V V

where Ld stands for Lagrangian density of the form

1
Ld = B H −E D + J A−ρ ϕ 29
2

which is quantity introduced in mathematical physics when fields are considered,

rather than particles [1, 2].
Therefore, total Lagrangian density is composed of a sum of Lagrangian for the
fields and sources, respectively.
Functional of electromagnetic energy containing source densities in the volume
of interest is of the form [4]

t2
1 1
F= B H− D E + J A − ϕ ρ dVdt 2 10
2 2
t1 V

Now, taking into account (2.1) and (2.2), the electric and magnetic fields are
expressed in terms of their potentials resulting in the following functional

t2 2
1 2 1 ∂A
F= ∇×A − ε ∇φ + + J A − φ ρ dVdt
2μ 2 ∂t
t1 V

2 11

And related Lagrangian density is given by

2
1 2 1 ∂A
Ld = ∇×A − ε ∇φ + + J A−φ ρ 2 12
2μ 2 ∂t

Furthermore, Lagrangian density (2.12) is requested to satisfy Lagrange–Euler

equations [4]

N
∂L ∂ ∂L ∂ ∂L ∂ ∂L ∂ ∂L
− − − − =0
k=1
∂qk ∂x ∂ qk x ∂y ∂ qk y ∂z ∂ qk z ∂t ∂ qk t

2 13
24 2 Fundamental Equations of Engineering Electromagnetics

Furthermore, the following expressions are obtained:

∂L
= −ρ 2 14
∂φ
∂ ∂L ∂ ∂φ ∂ ∂Dx
= −ε = εEx = 2 15
∂x ∂φ ∂x ∂x ∂x ∂x
∂
∂x
∂ ∂L ∂ ∂φ ∂ ∂Dy
= −ε = εEy = 2 16
∂y ∂φ ∂y ∂y ∂y ∂y
∂
∂y
∂ ∂L ∂ ∂φ ∂ ∂Dz
= −ε = εEz = 2 17
∂z ∂φ ∂z ∂z ∂z ∂z
∂
∂z

which simply gives

∂Dx ∂Dy ∂Dz

+ − =ρ 2 18
∂x ∂y ∂z

i.e. one obtains third divergence from Maxwell’s equation

∇ D=ρ 2 19

Proceeding with the application of Lagrange–Euler Eq. (2.13), one has:

∂L
= Jz 2 20
∂Az
∂ ∂L ∂ ∂φ ∂Az ∂Dz
= −ε + = − 2 21
∂t ∂Az ∂t ∂z ∂t ∂t
∂
∂t

∂ ∂L 1 ∂ ∂Az ∂Ax ∂H y
= − − = − 2 22
∂x ∂Az μ ∂x ∂x ∂z ∂x
∂
∂x

∂ ∂L 1 ∂ ∂Az ∂Ay ∂H x
= − − = 2 23
∂y ∂Az μ ∂y ∂y ∂z ∂y
∂
∂y
∂ ∂L
=0 2 24
∂z ∂Az
∂
∂z
 2.3 Integral Form of Maxwell’s Equations, Continuity Equations, and Lorentz Force 25

which can be written

∂Dz ∂H x ∂H y
Jz + + − =0 2 25
∂t ∂y ∂x
Performing similar mathematical operations for other vector potential compo-
nents yields, the following set of partial differential equations [4]
∂H y ∂H x ∂Dz
− = Jz + 2 26
∂x ∂y ∂t
∂H z ∂H y ∂Dx
− = Jx + 2 27
∂y ∂z ∂t
∂H x ∂H z ∂Dy
− = Jy + 2 28
∂z ∂x ∂t
which can be written in a vector form as second curl Maxwell’s equation

∂D
∇xH = J + 2 29
∂t
representing generalized Ampere’s law in the differential form.
Thus, Maxwell’s equations implicitly contain minimum energy principle. Con-
sequently, any equation derived from Maxwell’s equations satisfies this principle.
Similar analysis could be found elsewhere, e.g. in [4, 5].

2.3 Integral Form of Maxwell’s Equations, Continuity

Equations, and Lorentz Force

Integral form of Maxwell’s equations can be obtained from the differential form of
Maxwell’s equations by applying Stokes’ theorem and Gauss’s divergence
theorem.
Integral form of the Faraday law states that any change of magnetic flux density
B through any closed loop induces an electromotive force around the loop. Taking
the surface integration over (2.6) and applying the Stokes’ theorem yields:

∂B
Ed s = − dS 2 30
∂t
c S

where the line integral is taken around the loop and with dS = ndS.
The voltage induced by a varying flux has a polarity such that the induced cur-
rent in a closed path gives rise to a secondary magnetic flux opposing the change in
original time-varying magnetic flux.
Another random document with
no related content on Scribd:
4048 Davis H, S’t 38 July
A 27
Dec
12311 Delancey L D Art 2F
9
111 Aug
7013 Day W H
H 27
Sept
9073 Decker C Cav 7M
17
119 Aug
4608 Decker J P
C 3
Aug
7150 Demos B F 78 F
29
16 June
2497 Denhart W Cav
K 26
79 July
4422 Denior E, S’t
B 31
31 Sept
7514 Deming Joseph
D 1
15 Feb
12660 Denton E, Cor 65
B 16
Detreeman D, 44 June
2231 64
S’t E 20
16 Aug
5165 DePue J W
C 9
21 Apr
352 Deraus G W
B 2
89 June
2365 Drieks Henry
C 23
15 Jan
12547 Dilley A 65
E 28
May
1314 Dodson M, S’t Cav 3H 64
23
8187 Dock C “ 9H Sept
 8
July
3834 Dodd G W 21 F
23
July
4207 Dodson R B Cav 6B
29
July
2867 Dooley James Cav 16 L
4
May
1441 Doran W H 78 I
28
May
1103 Donen C 6 I
15
38 June
1727 Dowd J W
G 8
16 May
1342 Dowdy John
K 24
112 Oct
10143 Dowell J W
K 1
24 Oct
10496 Downer A
H 8
65 Jan
12436 Doyle P 65
H 11
112 Jan
12476 Doyle J
I 17
34 Aug
5053 Drake R R 64
H 8
24 Oct
10332 Dresser C
G 4
89 Sept
9678 Drum G
- 24
July
3123 Dudley J W 89 F
10
35 June
2666 Dumond P
E 29
9947 Dunn Alexander 75 Sept
A 28
30 Jan
12496 Dunsing A 65
C 21
30 Sept
9037 Dyer J C 64
D 17
53 Feb
12686 Drew E 65
D 20
26 Mar
209 Eadley Levi 64
H 28
100 Sept
8045 Easinbeck M
D 6
Easley W A, 21 Oct
10909
Cor G 14
Aug
5992 Eastman Wm 36 F
17
51 Aug
4962 Edwards C D
K 7
92 Sept
8084 Elliott Ed
B 7
26 Sept
9703 Ellis William
G 25
Sept
9734 Ellison W Cav 14 F
25
112 July
2249 Elslin James
E 24
22 Aug
4502 Emery J 64
K 1
Aug
4979 Emmerson J Cav 16 L
7
Sept
9717 Erb J 9C
25
12628 Ermains F Cav 14 Feb 65
 M 14
16 Mch
214 Errickson Cav 64
M 28
29 June
2211 Ench W
H 20
Nov
11727 Enrow W Cav 7M
1
Sept
2986 Evans J 9C
25
July
3373 Eydroner R 74 F
15
23 Aug
6268 Fagan O
G 20
June
2436 Fandish S Art 1A
25
21 June
2230 Farmer F
A 20
51 Aug
4991 Farnham C A
D 7
Ferguson L, 115 Oct
10740
Corpl K 14
June
2512 File R 11 K
26
14 Feb
12628 Fermer J 65
M 20
21 July
3854 Finch F M 64
G 24
Sept
10097 Fink J P 53 F
30
65 Oct
11541 Fish J
G 27
123 Sept
9845 Fisher S F
F 27
2129 Fitzgerald H Cav 16 I June
18
42 Sept
9992 Flanagan J
H 29
Aug
6972 Floyd A 7A
27
Oct
10881 Ford W J, S’t 17 I
13
112 Mch
161 Folk A P
G 26
93 June
2564 Forney D
G 27
16 Sept
8230 Foster A J Cav
M 8
112 Sept
7720 Foster B B
G 12
Foster E S, Jan
12473 9A 65
Corpl 17
14 April
531 Fowler John 64
D 13
120 Dec
12275 Frame W
E 17
April
12837 Francis J F 12 I 65
19
Aug
5933 Franklin H 81 F 64
17
16 April
432 Frass Louis
E 8
July
4031 Freeman D Cav 11 L
26
129 June
2080 French J 63
B 17
2210 Fritz P, S’t 38 June 64
 C 20
May
1055 Fremont James Cav 7B
13
112 April
497 Fuller Ira B
D 11
Sept
8114 Funk Wm 26 F
8
23 June
2021 Furlong H
B 15
20 Sept
9926 Gaines C
B 28
21 May
1347 Gallagher P
C 24
59 Apr
579 Garvin John, S’t
- 16
30 Mar
12801 Gerlock D 65
C 30
24 May
1340 German P 64
G 24
93 May
1416 Gibson H D
K 27
July
4201 Gibson L F 78 I
29
16 Aug
4485 Gichma J, S’t Cav
G 1
89 June
1652 Giles J V
H 5
112 Sept
7988 Giles S P
A 6
84 Aug
5144 Gillespie J W
H 9
May
1499 Gillgrease J Cav 16 I
30
1868 Gilmore J “ 16 June
E 12
14 Mar
12731 Gleason G M “ 65
A 4
73 June
1850 Glidwell F, Cor 64
K 11
51 June
2001 Goffinet P
D 15
89 Oct
10307 Goddard H
G 4
47 July
4203 Gooles H F, S’t
B 29
114 Apr
12847 Gordon I 65
B 25
Sept
7953 Gore F 36 I 64
5
15 Sept
7761 Gore N
C 4
Aug
6111 Garrig J 78 F
18
39 Jan
12461 Gott H 65
C 15
24 Sept
9403 Graber J 64
H 21
81 Sept
9312 Graber J F
D 20
21 June
2164 Grace W
D 19
41 Aug
6617 Graham M J 64
E 23
51 Oct
10998 Gravel J
C 16
2942 Greadley H 20 July
 A 6
Aug
4560 Greathouse J 6 I
2
Greaves 16 Apr
783
George K 28
79 Nov
12116 Green C
A 22
23 Oct
11155 Green John
H 19
Sept
7836 Green M 9C
4
July
3111 Greenwall B Cav 16 L
11
120 Nov
11778 Greer Geo, Cor
D 3
29 Oct
10594 Gress J
B 10
42 Apr
12834 Grimmins M A 65
H 17
79 July
4083 Griswold J P 64
E 27
66 June
2501 Grogan H
B 26
42 Oct
10466 Grower H
K 7
79 July
3730 Gulk P
B 21
72 Aug
5025 Guyen William
E 8
16 Aug
5961 Gonder H Cav
B 17
16 Aug
5074 Hageman Jas “
E 8
4094 Haggard E “ 16 July
K 27
89 Nov
11959 Haginis W
B 11
14 July
2825 Haines Theo Cav
M 3
16 Mar
63 Haks William
E 19
Oct
11572 Hall G H Cav 7B
27
41 Dec
12314 Hall H C, Cor
D 20
Aug
7194 Hall J L 9C
29
89 Dec
12223 Hall J L
G 4
103 Nov
11833 Hall Peter, Cor
D 5
22 Sept
10061 Haley C H
H 30
82 May
1241 Hallam Wm
H 20
21 June
2605 Hanna P
G 28
107 Mar
187 Hanna H, Cor
C 24
39 Oct
11188 Hansom D
E 19
65 Apr
318 Harken John
E 2
Aug
6684 Harlan J C 7L
24
6113 Harrell G 120 Aug
 K 19
112 June
2633 Harrington S M
A 29
79 Nov
11725 Harris E K
C 1
Oct
10447 Harris G W 9G
7
Harshman 84 Sept
8715
Peter H 14
16 June
2677 Hart George Cav
K 30
16 June
2202 Hart W
K 19
39 June
1980 Harney E
B 15
15 Oct
10606 Hathaway S Cav
B 10
15 Mar
12791 Hanch L 65
D 18
Sept
8608 Hawkins J W 79 I 64
12
Hayward W G, June
2326 16 I
Cor 22
Aug
5192 Hayworth F Cav 7 I
10
June
1852 Hegenberg W 24 F
11
77 Sept
8798 Helch S
K 15
31 Aug
6489 Hendson Geo B
C 22
23 July
1162 Henry Wm P
A 17
6035 Herdson Wm H 107 Aug
C 18
14 Sept
8428 Herrell Wm Cav
K 11
84 June
2365 Hess H
G 27
38 June
1906 Hester John
G 13
Sept
7865 Hicks Geo W 65 F
5
11 Sept
8303 Hicks H
G 10
85 May
1102 Hicks W
D 15
14 Nov
12070 Highland C Cav
C 17
24 Apr
725 Hilderbrand N
G 25
115 Sept
8830 Hill Aaron
C 15
36 Mar
67 Hill David, Cor
A 19
11 Sept
8721 Hill Henry
- 14
Aug
4489 Hill J Cav 9F
1
Feb
12683 Hinchcliff J 8B 65
20
112 Aug
6117 Hoen Peter 64
H 19
July
3825 Hoffman J Cav 7 I 64
23
11847 Hofman R 35 Nov
 C 5
98 June
2098 Hook Jas J, S’t
E 17
112 July
3255 Hoppock I
F 13
Sept
9880 Honeson A F 38 F
27
Sept
9214 Hormer J 38 F
19
86 Nov
12090 Horn T
A 18
Mar
89 Horseman W Cav 16 I
21
79 Aug
5812 Howard D N, S’t
E 16
Howard G S, 127 Nov
10782
Cor K 3
July
3211 Howell J W 78 F
12
100 Oct
11506 Hoye J
A 26
Aug
5741 Hude C 24 F
15
107 Aug
6035 Hudson W H
C 13
125 Sept
9962 Hughes D L
H 28
14 Mar
12755 Hulse A B 65
D 12
108 Oct
11140 Hungerford N 64
I 19
Aug
6085 Huntley R 89 F
18
 1136 Hulburt D 84 May
C 16
23 May
1162 Hurry W B
A 16
104 Aug
5019 Hutchins S
A 8
92 Aug
4583 Hustand B F, S’t
D 2
July
4091 Hyber John Cav 6A
27
July
3312 Iverson J S Cav 16 I
14
Jaccards S A, 29 July
4132
S’t E 28
51 June
2658 Jackson H
C 29
123 Oct
10287 Jackson M
F 4
Mar
12797 Janks J P Cav 3A 65
18
73 July
3686 Jarvis J 64
K 20
30 Aug
6733 Jenningsen G B
E 24
June
1845 Jenny E H, Cor 79 F
11
14 June
2135 Jewet F
A 18
June
1996 Johnson C W Cav 7F
15
Johnson 125 Sept
9458
Joseph K 21
1412 Johnson J S 7C May
 27
Johnson 100 Aug
5395
Samuel B 12
27 Sept
9827 Jones G W
E 27
117 Sept
8971 Jones J
E 16
41 Aug
4889 Jones P
G 6
112 Apr
644 Jones Thomas
E 12
June
2567 Jones Thos Cav 16 F
27
27 July
2990 Jones Wm
D 7
84 June
1764 Jordan B W
D 9
38 Sept
9153 Jordan M
C 18
July
2961 Joy B 16 I
6
90 June
2241 Joyce A
D 20
Oct
10513 Justice H Cav 7H
8
95 Nov
12052 Kane H
A 16
29 July
4308 Kappel H
H 30
Aug
4743 Keefe James P Art 2M
5
20 Sept
8348 Kelaze E
G 10
 18 Kell M R, Cor 49 Mar
D 7
Aug
7183 Kelly John 75 F
29
Aug
6795 Kelley William 94 I
25
38 Aug
5518 Kennedy M
C 13
Jan
12488 Kent J 14 F 65
19
96 Aug
5707 Kerbey John 64
H 15
22 Apr
396 Kiger John
E 6
42 Oct
10520 Kilkreath J
A 8
Mar
82 Kimball Jas Art 25 L
20
16 Mar
158 Kinkle John
G 25
82 Apr
696 Kinderman G
D 26
38 Sept
7807 Kingham J
G 4
65 Apr
685 Klinehaus D
G 23
113 Aug
4766 Kenigge A
C 5
Aug
4908 Knight J 9H
6
108 Nov
11891 Knoble P
E 7
4700 Koahl J Cav 16 Aug 64
 H 4
July
2754 Krail J “ 16 I
1
14 Feb
12685 Kreiger J 65
E 20
16 Apr
652 Kaiber John 64
D 20
June
1809 Keyser John 32 I
10
89 Sept
7927 Lacost J M
E 5
100 Aug
7299 Ladien J
H 23
38 Aug
7105 Lambert C
D 20
78 Oct
10419 Lamsden W H
A 6
59 Nov
12044 Lance V
D 16
14 Dec
12270 Langley G
K 12
Aug
5906 Lanner W A Cav 9E
16
93 May
1233 Law Henry
G 20
89 Sept
9635 Lawrene L G
G 24
125 Oct
10179 Lape J
A 1
98 Oct
10896 Leatherman M
E 14
115 Sept
8464 Leach W
B 11
4172 Lee A 112 July
B 28
16 Sept
8524 Lee P, Corp
A 12
May
963 Lee Thomas 8E
9
May
1297 Lee W E Cav 16 I
23
79 Oct
11258 Lewis Charles
A 21
Aug
6238 Lewis Thomas 2L
20
Oct
10148 Lickey J B, S’t 96 F
1
113 Sept
8295 Liday J
I 9
112 Aug
6295 Liken John, S’t
I 20
57 June
1685 Linday B
H 6
99 Sept
7768 Linderman H A
B 4
113 Aug
6414 Lindsay A
D 22
June
1818 Linebergh I, Cor 16 F
10
Oct
11449 Linwood J 79 F
25
Dec
12358 Lipsey D, Cor Cav 2C
30
112 Oct
10405 Lord L B, Corpl
B 6
11222 Lorsam C 89 Oct
 C 20
16 June
2268 London L Cav
D 21
35 May
1017 Lowry Frank
E 11
29 June
2342 Lusk John
B 23
23 May
1456 Lutz John
H 29
100 Sept
8196 Lyman J
D 8
38 Oct
11467 Lynch V, Corpl
C 26
14 Oct
10849 Mack J Cav
G 13
96 Aug
5390 Madden L
D 12
Maddock J W, 79 Oct
11358
S’t A 23
12 Aug
10982 Madrill A
A 12
38 July
3935 Malcolm J R
K 25
215 July
2868 Manning A
A 4
16 May
953 Manty P, S’t Cav
E 8
16 June
2050 Markman Wm Cav
K 16
Aug
6333 Marritt H Cav 16 L
21
Marshall A, 96 July
2762
Corpl C 2
8444 Martin A Cav 16 L May
28
July
4071 Martin I 9K
27
42 Mch
12757 Masman S 65
G 12
93 May
863 Mason Thos B 64
B 3
111 May
1428 Massey W H
D 28
12 April
746 Master Wm
A 26
Mathening A D, April
429 79 I
S’t 8
32 Mch
12744 Mathews F M 65
G 7
19 May
1061 Maxem H C 64
H 13
July
3280 Maxwell S Cav 8C
13
Sept
10319 May M H 89 I
29
104 July
3100 McCampbell D
B 10
Mch
56 McCleary Thos Cav 16 L
17
16 May
1315 McClusky Jas Cav
K 27
103 Aug
4850 McCray A
A 6
96 June
1617 McCready Wm
C 4
6513 McCreary J 119 Aug
 C 22
16 Aug
5724 McCone R Cav 64
K 15
13 July
3050 McCunne H
C 8
16 July
3470 McEntire L Cav
K 17
30 Aug
5283 McGee Wm
D 11
119 Oct
11623 McGivens J
A 28
89 Nov
11952 McLarens B
A 10
June
1634 McLaughlin B 90 I
5
McLing Benj, 23 July
3169
Cor E 11
93 Aug
4725 McMahon M
E 4
McMillan W B, 112 May
1337
Cor E 24
78 Sept
9763 McMiller W B
D 25
80 April
692 McShaw B
B 23
92 Sept
9710 McWorthy W M
G 25
19 July
3279 Mead G
H 14
Aug
4648 Medler H 38 I
3
51 Aug
6266 Mee William
C 20