*361*
*131*
*7MB*

*English*
*Pages 660*
*Year 2020*

- Author / Uploaded
- Balaji S

*Table of contents : Preface......Page 5Contents......Page 8About the Author......Page 14 1.1 Introduction......Page 15 1.3 Symbolic Representation of Vectors......Page 20 1.5 Subtraction of Vectors......Page 21 1.7 Multiplication of Vectors: Dot Product of Two Vectors......Page 22 1.8 Multiplication of Vectors—Cross—Product of Two Vectors......Page 23 1.9 Vector Components and Unit Vectors......Page 25 1.10 Triple Products......Page 31 1.11 Line, Surface and Volume Integration......Page 34 1.12 Flux......Page 37 1.13 Vector Differentiation: Gradient of a Scalar Function......Page 38 1.14 Vector Differentiation: Divergence of a Vector......Page 41 1.15 Vector Differentiation: Curl of a Vector......Page 46 1.16 Divergence Theorem......Page 54 1.17 Stoke’s Theorem......Page 61 1.18 The Gradient Theorem......Page 66 1.19 Others Coordinate Systems......Page 67 1.19.1 Spherical Polar Coordinates......Page 68 1.19.2 Cylindrical Coordinates......Page 73 1.20 Important Vector Identities......Page 76 1.21 Two and Three Dimensions......Page 78 Exercises......Page 82 2.1 Coulomb’s Law......Page 85 2.2 Electric Field Intensity......Page 87 2.3 Electric Field Intensity Due to a Group of Discrete Point Charges......Page 88 2.4 Continuous Charge Distributions......Page 89 2.5 A Note about Coulomb’s Law......Page 91 2.6 Calculating Electric Field E Using Coulomb’s Law......Page 92 2.6 Calculating Electric Field E Using Coulomb’s Law......Page 93 2.6 Calculating Electric Field E Using Coulomb’s Law......Page 94 2.6 Calculating Electric Field E Using Coulomb’s Law......Page 96 2.6 Calculating Electric Field E Using Coulomb’s Law......Page 98 2.6 Calculating Electric Field E Using Coulomb’s Law......Page 99 2.6 Calculating Electric Field E Using Coulomb’s Law......Page 101 2.6 Calculating Electric Field E Using Coulomb’s Law......Page 102 2.7 Solid Angle......Page 103 2.8 Gauss’s Law......Page 107 2.9 Sketches of Field Lines......Page 110 2.10 Curl of E......Page 112 2.11 Potential of Discrete and Continuous Charge Distributions......Page 115 2.12 Calculating Electric Field Using Gauss’s Law and Potential......Page 119 2.13 Electric Field Due to an Infinite Line Charge......Page 121 2.14 Electric Field Due to the Finite Line Charge......Page 127 2.15 Electric Field Along the Axis of a Uniformly Charged Circular Disc......Page 133 2.16 Electric Field Due to an Infinite Plane Sheet of Charge......Page 135 2.17 Electric Field of a Uniformly Charged Spherical Shell......Page 142 2.18 Comparison of Coulomb’s Law, Gauss’s Law and Potential Formulation......Page 150 2.19 Electric Field of a Dipole......Page 155 2.20 Calculation of Potential Using {\rm V} = \int\limits {{{\bf E}}\cdot{\hbox{d}}{\varvec l}}......Page 160 2.21 The Conservative Nature of Electric Field......Page 165 2.22 The Reference Point R in the Equation {\hbox{V}} = - \int\limits_{{\rm R}}^{{\rm P}} {{{\bf E}}\cdot {\hbox{d}}{\varvec l}}......Page 167 2.22 The Reference Point R in the Equation {\hbox{V}} = - \int\limits_{{\rm R}}^{{\rm P}} {{{\bf E}}\cdot {\hbox{d}}{\varvec l}}......Page 171 2.22 The Reference Point R in the Equation {\hbox{V}} = - \int\limits_{{\rm R}}^{{\rm P}} {{{\bf E}}\cdot {\hbox{d}}{\varvec l}}......Page 174 2.23 Poisson’s and Laplace’s Equation......Page 175 2.23 Poisson’s and Laplace’s Equation......Page 176 2.24 Conductors......Page 177 2.25 Boundary Conditions......Page 181 2.26 Uniqueness Theorem......Page 184 Exercises......Page 189 Exercises......Page 190 Exercises......Page 191 Exercises......Page 192 Exercises......Page 193 Exercises......Page 194 3.1 Work Done......Page 195 3.2 Energy in Electrostatic Fields......Page 197 3.3 Equipotential Surfaces......Page 201 3.4 A Note on Work Done......Page 203 3.4 A Note on Work Done......Page 206 3.5 Method of Images......Page 208 3.6 Point Charge Near a Grounded Conducting Sphere......Page 216 3.7 Laplace’s Equation—Separation of Variables......Page 220 3.8 Separation of Variables Laplace’s Equation in Cartesian Coordinates......Page 222 3.9 Potential Between Two Grounded Semi Infinite Parallel Electrodes Separated by a Plane Electrode Held by a Potential Vo......Page 223 3.10 Potential Between Two Grounded Conducting Electrodes Separated by Two Conducting Side Plates Maintained at Vo Potentials Vo and Vo......Page 228 3.11 Separation of Variables—Laplace’s Equation in Spherical Polar Coordinates......Page 231 3.11 Separation of Variables—Laplace’s Equation in Spherical Polar Coordinates......Page 232 3.11 Separation of Variables—Laplace’s Equation in Spherical Polar Coordinates......Page 233 3.11 Separation of Variables—Laplace’s Equation in Spherical Polar Coordinates......Page 236 3.11 Separation of Variables—Laplace’s Equation in Spherical Polar Coordinates......Page 239 3.12 Separation of Variables—Laplace’s Equation in Cylindrical Coordinates......Page 242 3.12 Separation of Variables—Laplace’s Equation in Cylindrical Coordinates......Page 244 3.12 Separation of Variables—Laplace’s Equation in Cylindrical Coordinates......Page 245 3.13 Summary......Page 246 3.15 Dielectric in an Electric Field......Page 247 3.16 Polar and Non-Polar Molecules......Page 248 3.17 Potential Produced by the Polarized Dielectric......Page 249 3.18 Bound Charges {{\varvec \upsigma}}_{{{\bf p}}} \,{{\bf and}}\,{{\varvec \uprho}}_{{{\bf p}}}......Page 252 3.19 Electric Displacement Vector and Gauss Law in Dielectrics......Page 255 3.20 Linear Dielectrics......Page 256 3.22 Boundary Conditions in the Presence of Dielectrics......Page 257 3.22 Boundary Conditions in the Presence of Dielectrics......Page 258 3.22 Boundary Conditions in the Presence of Dielectrics......Page 259 3.22 Boundary Conditions in the Presence of Dielectrics......Page 260 3.22 Boundary Conditions in the Presence of Dielectrics......Page 264 3.22 Boundary Conditions in the Presence of Dielectrics......Page 265 3.22 Boundary Conditions in the Presence of Dielectrics......Page 266 3.24 Principle of a Capacitor......Page 267 3.24.1 Capacity of a Parallel Plate Capacitor......Page 268 3.24.2 Capacity of a Parallel Plate Capacitor with Two Dielectrics......Page 269 3.25 Capacitance of a Spherical Capacitor......Page 270 3.26 Capacitance of a Cylindrical Capacitor......Page 272 3.27 Capacitors in Parallel and Series......Page 274 3.28 Energy Stored in a Capacitor......Page 276 3.28 Energy Stored in a Capacitor......Page 277 3.28 Energy Stored in a Capacitor......Page 278 3.28 Energy Stored in a Capacitor......Page 279 Exercises......Page 280 Exercises......Page 281 Exercises......Page 282 Exercises......Page 283 Exercises......Page 284 Exercises......Page 285 Exercises......Page 286 Exercises......Page 287 4.1 Introduction......Page 288 4.2 Lorentz Force Law......Page 289 4.3 Applications of Lorentz Force—Hall Effect......Page 290 4.5 Magnetic Force Between Two Current Elements......Page 293 4.6 Biot–Savart Law......Page 294 4.7 Current Distributions......Page 296 4.8 Magnetic Flux Density Due to a Steady Current in a Infinitely Long Straight Wire......Page 300 4.9 Ampere’s Circuital Law......Page 302 4.10 Equation of Continuity......Page 309 4.11 The Divergence of B......Page 310 4.12 Magnetic Monopoles......Page 312 4.13 Magnetic Vector Potential......Page 315 4.14 Magnetic Scalar Potential......Page 319 4.15 Comments on Magnetic Vector Potential A and Magnetic Scalar Potential Vm......Page 322 4.16 B of a Current-Carrying Infinitely Long Straight Conductor......Page 326 4.17 B of a Current-Carrying Finite Straight Conductor......Page 327 4.18 B Along the Axis of the Current-Carrying Circular Loop......Page 328 4.19 B Inside a Long Solenoid......Page 330 4.20 B of a Toroid......Page 335 4.21 Summary......Page 337 4.22 Magnetic Dipole......Page 339 4.23 Magnetic Boundary Conditions......Page 342 4.24 Force Between Two Parallel Current-Carrying Conductors......Page 345 4.25 Torque on a Current Loop in a Uniform Magnetic Field......Page 346 4.26 Magnetic Flux......Page 348 4.26 Magnetic Flux......Page 349 4.26 Magnetic Flux......Page 350 4.26 Magnetic Flux......Page 351 4.26 Magnetic Flux......Page 352 4.26 Magnetic Flux......Page 354 4.26 Magnetic Flux......Page 355 4.26 Magnetic Flux......Page 358 4.26 Magnetic Flux......Page 359 Exercises......Page 361 Exercises......Page 362 Exercises......Page 363 Exercises......Page 364 Exercises......Page 365 Exercises......Page 366 5.2 Diamagnetism, Paramagnetism and Ferromagnetism......Page 367 1......Page 371 5.3 Magnetization: Bound Currents......Page 372 5.4 Physical Interpretation of Bound Currents......Page 374 5.4 Physical Interpretation of Bound Currents......Page 375 5.4 Physical Interpretation of Bound Currents......Page 378 5.4 Physical Interpretation of Bound Currents......Page 381 5.5 Magnetic Field and Ampere’s Law in Magnetized Materials......Page 382 5.6 Linear and Nonlinear Media......Page 384 5.6 Linear and Nonlinear Media......Page 385 5.7 Boundary Conditions......Page 386 5.7 Boundary Conditions......Page 387 5.7 Boundary Conditions......Page 388 Exercises......Page 390 Exercises......Page 391 6.1.1 Ohm’s law......Page 392 6.2 Electromotive Force......Page 394 6.3 Motional emf......Page 396 6.4 Faraday’s Law......Page 399 6.5 Lenz’s Law......Page 401 6.5 Lenz’s Law......Page 403 6.5 Lenz’s Law......Page 404 6.5 Lenz’s Law......Page 405 6.5 Lenz’s Law......Page 406 6.5 Lenz’s Law......Page 407 6.6 Magnetic Circuits......Page 408 6.6 Magnetic Circuits......Page 412 6.6 Magnetic Circuits......Page 413 6.6 Magnetic Circuits......Page 414 6.7 Induction—Self Induction......Page 415 6.8 Induction—Mutual Induction......Page 416 6.8 Induction—Mutual Induction......Page 418 6.8 Induction—Mutual Induction......Page 419 6.8 Induction—Mutual Induction......Page 420 6.8 Induction—Mutual Induction......Page 421 6.8 Induction—Mutual Induction......Page 422 6.8 Induction—Mutual Induction......Page 423 6.10 Energy Stored in Magnetic Fields......Page 424 6.10 Energy Stored in Magnetic Fields......Page 426 6.10 Energy Stored in Magnetic Fields......Page 427 6.11 Maxwell’s Equation......Page 430 6.12 The Displacement Current......Page 432 6.13 Maxwell’s Equation in Matter......Page 434 6.14 Maxwell’s Equation in Integral Form and Boundary Conditions......Page 437 6.15 Potential Functions......Page 439 6.16 Gauge Transformations......Page 441 6.17 Retarded Potentials......Page 443 6.18 Time-Harmonic Fields......Page 445 6.18 Time-Harmonic Fields......Page 446 6.18 Time-Harmonic Fields......Page 447 6.18 Time-Harmonic Fields......Page 448 6.18 Time-Harmonic Fields......Page 449 6.19 Maxwell’s Equation in Phasor Form......Page 450 Exercises......Page 451 Exercises......Page 452 Exercises......Page 453 Exercises......Page 454 7.1 Introduction......Page 455 7.2 The Wave Equation......Page 457 7.3 Plane Electromagnetic Wave in Free Space......Page 458 7.4 Poynting Theorem......Page 465 7.5 Average Poynting Vector......Page 469 7.6 Poynting Vector for Wave Propagation in Free Space......Page 471 7.7 Plane Electromagnetic Waves in Lossy Dielectrics......Page 473 7.8 Plane Electromagnetic Waves in Good Conductors......Page 480 7.8 Plane Electromagnetic Waves in Good Conductors......Page 482 7.9 Plane Electromagnetic Waves in Lossy Dielectrics Using Maxwell’s Equation in Phasor Form......Page 483 7.10 Wave Polarization......Page 484 7.10 Wave Polarization......Page 489 7.11 Reflection and Transmission of the Plane Wave at Normal Incidence......Page 490 7.12 Reflection and Transmission of the Plane Wave at Oblique Incidence......Page 498 7.13 Total Reflection and Total Transmission......Page 509 7.14 Dispersion—Group Velocity......Page 512 7.14 Dispersion—Group Velocity......Page 513 7.14 Dispersion—Group Velocity......Page 514 7.14 Dispersion—Group Velocity......Page 515 Exercises......Page 517 Exercises......Page 518 8.1 Introduction......Page 519 8.2 Description of a Transmission Line......Page 520 8.3 Wave Propagation in a Transmission Line......Page 522 8.4 Transmission Line Equations......Page 523 8.5 Wave Propagation in the Transmission Line—Circuit Model......Page 525 8.6 Lossless Line......Page 527 8.7 Low-loss Line......Page 529 8.8 Distortionless Line......Page 530 8.9 Relationship Between G and C......Page 532 8.10 Determination of Transmission Line Parameters......Page 533 8.11 Field Approach—TEM Waves in a Parallel-Plate Transmission Line......Page 534 8.13 Finite Transmission Line......Page 546 8.14 Standing Waves......Page 551 8.15 Lossless Transmission Lines with Resistive Transmission......Page 560 8.16 Smith Chart......Page 563 8.17 Impedance Matching......Page 583 Exercises......Page 589 9.2 Transverse Electromagnetic, Transverse Electric and Transverse Magnetic Waves......Page 591 9.3 Wave Equation in Cartesian Coordinates......Page 592 9.4 Parallel-Plate Waveguide......Page 595 9.4 Parallel-Plate Waveguide......Page 600 9.5 Propagation of TEM, TE and TM Waves in Parallel-Plate Waveguide......Page 602 9.5 Propagation of TEM, TE and TM Waves in Parallel-Plate Waveguide......Page 603 9.6 General Solution of Wave Equation......Page 604 9.7 Rectangular Waveguides......Page 606 9.7 Rectangular Waveguides......Page 615 9.7 Rectangular Waveguides......Page 616 9.7 Rectangular Waveguides......Page 617 9.7 Rectangular Waveguides......Page 618 9.7 Rectangular Waveguides......Page 619 9.8 Cavity Resonators......Page 620 9.9 Quality Factor......Page 625 Exercises......Page 626 10.2 Types of Antenna......Page 628 10.3 Hertzian Dipole......Page 630 10.4 Magnetic Dipole......Page 639 10.5 Half Wave Dipole Antenna......Page 644 10.6 Antenna Arrays......Page 650 10.7 Receiving Antenna and Friis Equation......Page 654 10.8 The Radar Equation......Page 658 Exercises......Page 659*

S. Balaji

Electromagnetics Made Easy

Electromagnetics Made Easy

S. Balaji

Electromagnetics Made Easy

123

S. Balaji Indira Gandhi Centre for Atomic Research Kalpakkam, Tamil Nadu, India

ISBN 978-981-15-2657-2 ISBN 978-981-15-2658-9 https://doi.org/10.1007/978-981-15-2658-9

(eBook)

© Springer Nature Singapore Pte Ltd. 2020 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, speciﬁcally the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microﬁlms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a speciﬁc statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

Engineering electromagnetics is an interesting subject in the electrical/electronics engineering curriculum. However, students ﬁnd it difﬁcult to grasp the concepts and usually complain the paper is tedious. The common conception about electromagnetics is that the subject is hard to understand and is not as beautiful as other papers. As an example, mechanics appears to be friendlier to students. Concrete concepts and dealing with tangible objects make the learning of mechanics simple for students. Studies on projectile thrown in space, an object sliding on a floor, can be clearly visualized. On the contrary, electromagnetics is characterized by abstract concepts and intangible ﬁelds. Along with complex mathematics, imperceptible concepts and invisible ﬁelds confuse the students, and understanding electromagnetics becomes a dream for the reader. An attempt has been made in this book to change the above scenario. Real-life examples have been used throughout the book for easy grasping of the abstract concepts. This book is a result of a decade of teaching electromagnetics for electrical/electronics engineering students and physics postgraduates. Motivated by the positive feedback from students for the simple and lucid form of presentation of the subject, a need was felt to turn the presentation in the form of a textbook for a wider audience. The book contains ten chapters. The backbone of electromagnetics is vector analysis, and to make sure the reader becomes more familiar with vector calculus, Chap. 1 has been devoted to the study of vectors. Various theorems in vector calculus and coordinate systems, which are the fundamental concepts required to understand electromagnetics, are discussed in detail in the ﬁrst chapter. Chapters 2–10 can be grouped under four major sections—electrostatics, magnetostatics, time-varying ﬁelds and applications of electromagnetics. Chapters 2 and 3 focus on electrostatics. Chapter 2 commences with an introduction to Coulomb’s law and proceeds to calculate electric ﬁeld for various charge distributions. Chapter 3 continues with the calculation of electric ﬁeld and ends up with a detailed discussion on dielectrics and capacitors. Chapters 4 and 5 concentrate on magnetostatics. Chapter 4 introduces Biot–Savart law and continues with the calculation of magnetic flux density using various methods, and the methods are compared with v

vi

Preface

their electrostatic counterpart. Chapter 5 elaborates about magnetization and magnetic materials. Chapter 6 introduces time-varying ﬁelds by discussing Faraday’s law and induction, followed by a detailed account on Maxwell’s equation. The last sections of Chap. 6 describe the gauge transformations, retarded potentials and time-harmonic ﬁelds. Chapter 7 gives a detailed account on the flow of energy in electromagnetic wave—the Poynting theorem, wave polarization and reflections and transmission of electromagnetic waves. Chapters 8–10 describe the application part of electromagnetics. The theory so developed in Chaps. 1–7 is applied in the development of transmission lines, waveguides and antennas in Chaps. 8–10. Sequential development of the subject and utilizing concrete examples to explain abstract theory are the distinct features of the book. There are a number of topics in this book making it unique in the presentation of the concepts. Few such topics are mentioned below: 1. Real-life examples have been used to explain (a) that “Vectors are difﬁcult to deal with as compared to scalars” in Sect. 1.1 (b) Flux in Sect. 1.12 (c) Existence of source and sink as related to the divergence of a vector in Sect. 1.14 (d) Non-existence of magnetic monopoles in Sect. 4.12 (e) Deﬁciency as related to magnetic scalar potential in the calculation of B in Sect. 4.15. 2. In spherical polar coordinates, h is allowed to vary from 0 to p and not up to 2p. The reason for such a variation is explained with suitable ﬁgures in Sect. 1.19a. 3. For a given charge distribution all the three methods (Coulomb’s law, Gauss’s law and potential formulation) have been utilized to calculate the electric ﬁeld. 4. In Sect. 2.18, the three methods (Coulomb’s law, Gauss’s law and potential formulation) utilized for the calculation of the electric ﬁeld for a given charge distribution have been compared. 5. A detailed discussion on reference point R in the calculation of potential is elaborated in Sect. 2.22. Much deeper explanations are put forward for a number of concepts, which makes the subject easy and understandable. Here are few examples. When the instructor states “Vectors are difﬁcult to deal with as compared to scalars,” at once students question—Why is it so? Section 1.1 of this book explains why. When the students ask—why magnetic scalar potential Vm cannot be used in the calculation, the reply given will be Vm is a multi-valued function. At once the question to shoot up in the students’ mind is—let Vm be a multi-valued function, then how does it affect the calculation? How the multi-valuedness of Vm affects the calculation is explained in Sect. 4.15.

Preface

vii

Likewise, there are a number of questions in the students’ mind for which they get partial answers which do not satisfy the students’ quest to learn, leading to students not understanding the subject, and ﬁnally feel that electromagnetics is difﬁcult and can never be understood. So, as much as possible effort has been made in this book to make sure that all the questions are answered and students can grasp the subject easily. Understanding the concepts is a different skill, while applying the concepts to solve the problems is a different skill. To make sure that the students apply whatever they have learnt, in solving problems, a number of examples have been included throughout the text. A list of exercises is given at the end of each chapter, and it is emphasized that the students should attempt those exercises on their own to further strengthen their problem-solving skill. Finally, I should thank my wife E. Sumathi and my daughter B. S. Nakshathra for their extreme patience they had during the development of this book. Kalpakkam, India

S. Balaji

Contents

1

2

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

1 1 6 6 7 7 8 8

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

9 11 17 20 23 24 27 32 40 47 52 53 54 59 62 64 68

Electric Charges at Rest: Part I . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Coulomb’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Electric Field Intensity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

71 71 73

Vector Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Graphical Representation of Vectors . . . . . . . . . . . . . . 1.3 Symbolic Representation of Vectors . . . . . . . . . . . . . . . 1.4 Vector Addition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.5 Subtraction of Vectors . . . . . . . . . . . . . . . . . . . . . . . . . 1.6 Multiplication of a Vector by a Scalar . . . . . . . . . . . . . 1.7 Multiplication of Vectors: Dot Product of Two Vectors 1.8 Multiplication of Vectors—Cross—Product of Two Vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.9 Vector Components and Unit Vectors . . . . . . . . . . . . . 1.10 Triple Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.11 Line, Surface and Volume Integration . . . . . . . . . . . . . 1.12 Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.13 Vector Differentiation: Gradient of a Scalar Function . . 1.14 Vector Differentiation: Divergence of a Vector . . . . . . . 1.15 Vector Differentiation: Curl of a Vector . . . . . . . . . . . . 1.16 Divergence Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . 1.17 Stoke’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.18 The Gradient Theorem . . . . . . . . . . . . . . . . . . . . . . . . 1.19 Others Coordinate Systems . . . . . . . . . . . . . . . . . . . . . 1.19.1 Spherical Polar Coordinates . . . . . . . . . . . . . . 1.19.2 Cylindrical Coordinates . . . . . . . . . . . . . . . . . 1.20 Important Vector Identities . . . . . . . . . . . . . . . . . . . . . 1.21 Two and Three Dimensions . . . . . . . . . . . . . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ix

x

Contents

2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12 2.13 2.14 2.15 2.16 2.17 2.18 2.19 2.20 2.21 2.22

Electric Field Intensity Due to a Group of Discrete Point Charges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Continuous Charge Distributions . . . . . . . . . . . . . . . . . . . . . A Note about Coulomb’s Law . . . . . . . . . . . . . . . . . . . . . . . Calculating Electric Field E Using Coulomb’s Law . . . . . . . Solid Angle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gauss’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sketches of Field Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . Curl of E . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Potential of Discrete and Continuous Charge Distributions . . Calculating Electric Field Using Gauss’s Law and Potential . Electric Field Due to an Inﬁnite Line Charge . . . . . . . . . . . . Electric Field Due to the Finite Line Charge . . . . . . . . . . . . Electric Field Along the Axis of a Uniformly Charged Circular Disc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electric Field Due to an Inﬁnite Plane Sheet of Charge . . . . . Electric Field of a Uniformly Charged Spherical Shell . . . . . Comparison of Coulomb’s Law, Gauss’s Law and Potential Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electric Field of a Dipole . . . . . . . R. . . . . . . . . . . . . . . . . . . Calculation of Potential Using V ¼ E dl . . . . . . . . . . . . . The Conservative Nature of Electric Field . . . . . . . . . . . . . . RP The Reference Point R in the Equation V ¼ E dl . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . 119 . . 121 . . 128 . . . .

. . . .

..

136 141 146 151

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

153 161 163 167 170 175

Electric Charges at Rest—Part II . . . . . . . . . . . . . . . . . . . . 3.1 Work Done . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Energy in Electrostatic Fields . . . . . . . . . . . . . . . . . . . 3.3 Equipotential Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 A Note on Work Done . . . . . . . . . . . . . . . . . . . . . . . . 3.5 Method of Images . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6 Point Charge Near a Grounded Conducting Sphere . . . . 3.7 Laplace’s Equation—Separation of Variables . . . . . . . . 3.8 Separation of Variables Laplace’s Equation in Cartesian Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.9 Potential Between Two Grounded Semi Inﬁnite Parallel Electrodes Separated by a Plane Electrode Held by a Potential Vo . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

181 181 183 187 189 194 202 206

R

2.23 Poisson’s and Laplace’s Equation 2.24 Conductors . . . . . . . . . . . . . . . . . 2.25 Boundary Conditions . . . . . . . . . 2.26 Uniqueness Theorem . . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . . 3

74 75 77 78 89 93 96 98 101 105 107 113

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . . . 208

. . . . . . 209

Contents

xi

3.10 Potential Between Two Grounded Conducting Electrodes Separated by Two Conducting Side Plates Maintained at Vo Potentials Vo and Vo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.11 Separation of Variables—Laplace’s Equation in Spherical Polar Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.12 Separation of Variables—Laplace’s Equation in Cylindrical Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.13 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.14 Dielectrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.15 Dielectric in an Electric Field . . . . . . . . . . . . . . . . . . . . . . 3.16 Polar and Non-Polar Molecules . . . . . . . . . . . . . . . . . . . . . 3.17 Potential Produced by the Polarized Dielectric . . . . . . . . . . 3.18 Bound Charges rp and qp . . . . . . . . . . . . . . . . . . . . . . . . . 3.19 Electric Displacement Vector and Gauss Law in Dielectrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.20 Linear Dielectrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.21 Dielectric Breakdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.22 Boundary Conditions in the Presence of Dielectrics . . . . . . 3.23 Capacitance and Capacitors . . . . . . . . . . . . . . . . . . . . . . . . 3.24 Principle of a Capacitor . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.24.1 Capacity of a Parallel Plate Capacitor . . . . . . . . . . 3.24.2 Capacity of a Parallel Plate Capacitor with Two Dielectrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.25 Capacitance of a Spherical Capacitor . . . . . . . . . . . . . . . . . 3.26 Capacitance of a Cylindrical Capacitor . . . . . . . . . . . . . . . . 3.27 Capacitors in Parallel and Series . . . . . . . . . . . . . . . . . . . . 3.28 Energy Stored in a Capacitor . . . . . . . . . . . . . . . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

Magnetostatics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Lorentz Force Law . . . . . . . . . . . . . . . . . . . . . . 4.3 Applications of Lorentz Force—Hall Effect . . . . 4.4 Sources of Magnetic Field . . . . . . . . . . . . . . . . . 4.5 Magnetic Force Between Two Current Elements . 4.6 Biot–Savart Law . . . . . . . . . . . . . . . . . . . . . . . . 4.7 Current Distributions . . . . . . . . . . . . . . . . . . . . . 4.8 Magnetic Flux Density Due to a Steady Current in a Inﬁnitely Long Straight Wire . . . . . . . . . . . 4.9 Ampere’s Circuital Law . . . . . . . . . . . . . . . . . . 4.10 Equation of Continuity . . . . . . . . . . . . . . . . . . . 4.11 The Divergence of B . . . . . . . . . . . . . . . . . . . . 4.12 Magnetic Monopoles . . . . . . . . . . . . . . . . . . . . .

. . . 214 . . . 217 . . . . . . .

. . . . . . .

. . . . . . .

228 232 233 234 234 236 238

. . . . . . .

. . . . . . .

. . . . . . .

241 242 243 243 253 254 254

. . . . . .

. . . . . .

. . . . . .

256 257 258 260 263 267

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

275 275 276 277 280 280 281 283

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

287 289 296 297 299

xii

Contents

4.13 Magnetic Vector Potential . . . . . . . . . . . . . . . . . . . . . . . . 4.14 Magnetic Scalar Potential . . . . . . . . . . . . . . . . . . . . . . . . 4.15 Comments on Magnetic Vector Potential A and Magnetic Scalar Potential Vm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.16 B of a Current-Carrying Inﬁnitely Long Straight Conductor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.17 B of a Current-Carrying Finite Straight Conductor . . . . . . 4.18 B Along the Axis of the Current-Carrying Circular Loop . 4.19 B Inside a Long Solenoid . . . . . . . . . . . . . . . . . . . . . . . . 4.20 B of a Toroid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.21 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.22 Magnetic Dipole . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.23 Magnetic Boundary Conditions . . . . . . . . . . . . . . . . . . . . 4.24 Force Between Two Parallel Current-Carrying Conductors 4.25 Torque on a Current Loop in a Uniform Magnetic Field . . 4.26 Magnetic Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

6

. . . . 302 . . . . 306 . . . . 309 . . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

313 314 315 317 322 324 326 329 332 333 335 348

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

355 355 355 360 362

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

370 372 374 378

Time-Varying Fields and Maxwell’s Equation . . . . . . . . . . . 6.1 Introduction—Ohm’s law . . . . . . . . . . . . . . . . . . . . . . . 6.1.1 Ohm’s law . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Electromotive Force . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 Motional emf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4 Faraday’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5 Lenz’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6 Magnetic Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.7 Induction—Self Induction . . . . . . . . . . . . . . . . . . . . . . . 6.8 Induction—Mutual Induction . . . . . . . . . . . . . . . . . . . . . 6.9 Energy Stored in the Circuit in Terms of Self-Inductance 6.10 Energy Stored in Magnetic Fields . . . . . . . . . . . . . . . . . 6.11 Maxwell’s Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.12 The Displacement Current . . . . . . . . . . . . . . . . . . . . . . . 6.13 Maxwell’s Equation in Matter . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

381 381 381 383 385 388 390 397 404 405 413 413 419 421 423

Magnetic Fields in Materials . . . . . . . . . . . . . . . . . . . . . 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Diamagnetism, Paramagnetism and Ferromagnetism 5.3 Magnetization: Bound Currents . . . . . . . . . . . . . . . 5.4 Physical Interpretation of Bound Currents . . . . . . . 5.5 Magnetic Field and Ampere’s Law in Magnetized Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6 Linear and Nonlinear Media . . . . . . . . . . . . . . . . . 5.7 Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Contents

6.14 Maxwell’s Equation in Integral Form and Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.15 Potential Functions . . . . . . . . . . . . . . . . . . . . . . . . . 6.16 Gauge Transformations . . . . . . . . . . . . . . . . . . . . . . 6.17 Retarded Potentials . . . . . . . . . . . . . . . . . . . . . . . . . 6.18 Time-Harmonic Fields . . . . . . . . . . . . . . . . . . . . . . . 6.19 Maxwell’s Equation in Phasor Form . . . . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

8

xiii

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

426 428 430 432 434 439 440

Electromagnetic Waves . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Wave Equation . . . . . . . . . . . . . . . . . . . . . . . . . Plane Electromagnetic Wave in Free Space . . . . . . . . Poynting Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . Average Poynting Vector . . . . . . . . . . . . . . . . . . . . . Poynting Vector for Wave Propagation in Free Space . Plane Electromagnetic Waves in Lossy Dielectrics . . . Plane Electromagnetic Waves in Good Conductors . . . Plane Electromagnetic Waves in Lossy Dielectrics Using Maxwell’s Equation in Phasor Form . . . . . . . . 7.10 Wave Polarization . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.11 Reﬂection and Transmission of the Plane Wave at Normal Incidence . . . . . . . . . . . . . . . . . . . . . . . . . 7.12 Reﬂection and Transmission of the Plane Wave at Oblique Incidence . . . . . . . . . . . . . . . . . . . . . . . . . 7.13 Total Reﬂection and Total Transmission . . . . . . . . . . . 7.14 Dispersion—Group Velocity . . . . . . . . . . . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

445 445 447 448 455 459 461 463 470

Plane 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9

Transmission Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2 Description of a Transmission Line . . . . . . . . . . . . . 8.3 Wave Propagation in a Transmission Line . . . . . . . . 8.4 Transmission Line Equations . . . . . . . . . . . . . . . . . . 8.5 Wave Propagation in the Transmission Line—Circuit Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.6 Lossless Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.7 Low-loss Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.8 Distortionless Line . . . . . . . . . . . . . . . . . . . . . . . . . 8.9 Relationship Between G and C . . . . . . . . . . . . . . . . 8.10 Determination of Transmission Line Parameters . . . . 8.11 Field Approach—TEM Waves in a Parallel-Plate Transmission Line . . . . . . . . . . . . . . . . . . . . . . . . . . 8.12 The Inﬁnite Transmission Line . . . . . . . . . . . . . . . .

. . . . . . . 473 . . . . . . . 474 . . . . . . . 480 . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

488 499 502 507

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

509 509 510 512 513

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

515 517 519 520 522 523

. . . . . . . . 524 . . . . . . . . 536

xiv

Contents

8.13 Finite Transmission Line . . . . . . . . . . . . . . . . . . . . . . . . 8.14 Standing Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.15 Lossless Transmission Lines with Resistive Transmission 8.16 Smith Chart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.17 Impedance Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Waveguides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2 Transverse Electromagnetic, Transverse Electric and Transverse Magnetic Waves . . . . . . . . . . . 9.3 Wave Equation in Cartesian Coordinates . . . . . 9.4 Parallel-Plate Waveguide . . . . . . . . . . . . . . . . . 9.5 Propagation of TEM, TE and TM Waves in Parallel-Plate Waveguide . . . . . . . . . . . . . . . 9.6 General Solution of Wave Equation . . . . . . . . . 9.7 Rectangular Waveguides . . . . . . . . . . . . . . . . . 9.8 Cavity Resonators . . . . . . . . . . . . . . . . . . . . . . 9.9 Quality Factor . . . . . . . . . . . . . . . . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

10 Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . 10.1 Introduction . . . . . . . . . . . . . . . . . . . 10.2 Types of Antenna . . . . . . . . . . . . . . . 10.3 Hertzian Dipole . . . . . . . . . . . . . . . . 10.4 Magnetic Dipole . . . . . . . . . . . . . . . . 10.5 Half Wave Dipole Antenna . . . . . . . . 10.6 Antenna Arrays . . . . . . . . . . . . . . . . 10.7 Receiving Antenna and Friis Equation 10.8 The Radar Equation . . . . . . . . . . . . . Exercises . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

536 541 550 553 573 579

. . . . . . . . . . . . 581 . . . . . . . . . . . . 581 . . . . . . . . . . . . 581 . . . . . . . . . . . . 582 . . . . . . . . . . . . 585 . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

592 594 596 610 615 616

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

619 619 619 621 630 635 641 645 649 650

About the Author

Dr. S. Balaji is a Scientiﬁc Ofﬁcer at the Indira Gandhi Center for Atomic Research (IGCAR), Kalpakkam, India. He obtained his Ph.D. in physics from the University of Madras. His research primarily focuses on radiation damage, application of ion beams in material science, and ion beam analysis. He is also actively involved in the development of experimental facilities related to ion beam irradiation. He has published 20 articles in international peer-reviewed journals and conference proceedings. Having a passion for teaching, he has taught engineering electromagnetics/electromagnetic theory to engineering undergraduates and physics postgraduates for over a decade. Currently, he teaches the same topic to the junior research fellows at IGCAR.

xv

Chapter 1

Vector Analysis

1.1

Introduction

(a) Several complex problems in electromagnetism involve many physical quantities having both magnitude and direction. These quantities called vectors play an important role in understanding the multitude of phenomena in electrostatics, magnetostatics, etc. A scalar is a quantity which involves magnitude alone. A vector is a quantity which involves both magnitude and direction. Examples of scalar quantities are mass, length, temperature, etc., and examples of vector quantities are displacement, velocity, force, etc. The following example will make the difference between vector and scalar much clearer. Suppose, assume that there are two persons, person A and person B, trying to add rice in an empty box as shown in Fig. 1.1a, b. Person A is having 7 kg of rice and person B is having 2 kg of rice. In Fig. 1.1a person A and person B are standing in same direction and are adding the rice in the empty box. After adding the rice in the box, the box would contain 9 kg of rice. In Fig. 1.1b person A and person B are standing opposite to each other and are adding the rice in the empty box. Once again the box would then contain 9 kg of rice after the rice has been added in the box. So, the conclusion is, whatever direction the mass is added the net effect is same, we get 9 kg of rice in total. The total (9 kg in our example) simply depends on the initial magnitude (7 and 2 kg in our example) of the masses and not on the direction in which the masses are added. Now consider Fig. 1.2a, b. In Fig. 1.2a person A and person B are kicking a football with a force of 7 N and 2 N (Numbers are not to be taken seriously) in the same direction. That is person A and person B are standing adjacent to each other and kicking the football. The total effect or the total force is 7 N + 2 N = 9 N and the ball will move in the direction of the kick. In Fig. 1.2b person A and person B are standing opposite to each other and kicking the football with a force of 7 N and © Springer Nature Singapore Pte Ltd. 2020 S. Balaji, Electromagnetics Made Easy, https://doi.org/10.1007/978-981-15-2658-9_1

1

2

1 Vector Analysis

Fig. 1.1 a Person A and Person B standing in same direction and adding rice in the empty box. b Person A and Person B standing in the opposite direction and adding rice in the empty box

2 N, respectively. The net or total force is 7 N − 2 N = 5 N acting in the direction of kick of person A. Now, the conclusion is the total force (9 N and 5 N in our example) depends on, not only on the magnitude (7 N and 2 N in our example) but also on the direction of the initial forces. Person A and person B can stand in any possible direction and kick the football. Few such possible directions are shown in Fig. 1.2c. For such situations shown in Fig. 1.2c the net or total force and its direction must be obtained by using laws of vector addition. Thus, irrespective of whatever direction the initial masses are added we get the same total mass. However when forces are added, depending upon the direction of the initial forces, we get a different net or total force. Quantities which behave like mass are called scalars and quantities which behave like force are called vectors. Thus, the above example clearly distinguishes between a vector and scalar. We have another important information from the above example. During addition of the scalars one need not bother about their direction. One must take into account their magnitude alone. Because irrespective of whatever direction the scalars (masses in our example) are added we get the same total or net scalar (mass in our example) we need not bother about their direction. However, during addition of vectors we need to keep track of not only their magnitude, but also their direction. In our force example in Fig. 1.2a–c the initial direction of 7 N and 2 N must always be tracked if we want to get the total or net force. As we will see shortly the direction of the total or net force must also be tracked. Because the direction of the initial vector (7 N and 2 N in our force example in Fig. 1.2a–c) must always be tracked in order to obtain the value of the direction and magnitudes of the total or net vector vector addition becomes complicated than scalar addition. Scalar addition is simple. One needs to bother about their

1.1 Introduction

3

Fig. 1.2a, b, c Person A, B kicking the football from different directions. A detailed explanation is given below each ﬁgure

4

1 Vector Analysis

magnitude alone not their direction. In our mass example in Fig. 1.1, 7 and 2 kg can be added either in same direction or opposite direction or any possible direction we still get the same total or net mass 9 kg. So, no need to worry about direction of initial masses. The same is true for any scalar. However vector addition is comparatively complicated than scalar addition. In our force example in Fig. 1.2a–c the total or net force depends upon the direction and magnitude of the initial forces 7 N and 2 N. Hence, in addition to magnitude, the direction of initial forces needs to be always tracked. The same is true for any vector. This makes vector addition complicated. In order to get a feel how much difﬁcult it is to deal with vectors, assume that we are interested in adding ten masses m1, m2, … m10 and ten forces F1, F2 … F10. For all ten masses we need not keep track of their direction. Just add m1 + m2 + + m9 + m10. But what about forces. Each of the ten forces is going to have its own direction and we have to mess up with the direction. Now you can realize how much difﬁcult it is to deal with vectors. Similar to vector addition, vector multiplication, vector differentiation, vector integration, etc., become difﬁcult because vectors involve direction. For example if we want to multiply two scalars then we simply take their magnitude and multiply them. However, in the case of multiplication of two vectors, we have two types— dot product and cross product—as vectors involve direction. Vector multiplication thus is comparatively complicated than scalar multiplication. In Example 1.10 the difﬁculty involved in working with vectors is further explained. Hence, in electromagnetics wherever possible our aim will be to reduce a vector problem into a scalar problem because all operations can be carried out easily with scalars than with vectors. The ball example which we have shown in Fig. 1.2 must not be taken too seriously. Clearly the ball will hit person B and stop. In the case of electromagnetics we will be discussing with the invisible electric and magnetic forces where this problem doesn’t arise. As an example, in Fig. 1.2 replace the ball with a particle of charge q and the persons, with electric forces acting in the respective directions. We have taken the ball example—the real-life example for better grasping of the concept of vectors for the reader. In Figs. 1.1, 1.2 we have used scalar addition (normal usual addition) and vector addition, to differentiate between a scalar and vector which the reader will not ﬁnd it in any book. In standard books the vector addition will be deﬁned in the way, we have deﬁned in Sect. 1.4. The reader can use parallelogram law of vectors (which is not described in this book) to check whether we get similar results for Fig. 1.2. A ﬁnal point to be noted is some students raise an objection that if vectors are difﬁcult to deal with, then why we are considering force as a vector. Why force cannot be considered as a scalar. The reader might note that the direction depen-

1.1 Introduction

5

dency of the force shown in Fig. 1.2a, b is not our wish. We are not forcing the force to act in such a manner; adding and cancelling depending upon direction. Direction dependency of force (vectors) is naturally and intrinsically present in it and direction independency of mass (scalars) is naturally and intrinsically present in it. (b) Previously we noted that not only the direction of the initial forces must be tracked but the direction of the net or total force must also be tracked. In order to understand the above point consider Fig. 1.2d. In Fig. 1.2d person A and person B are kicking the football exactly opposite to the manner shown in Fig. 1.2b. Clearly, in Fig. 1.2d the resultant force is also 5 N but now it is opposite to the resultant force 5 N in Fig. 1.2b. Thus, the resultant force in Fig. 1.2b, d is 5 N but acting completely in opposite direction. The result is football in Fig. 1.2d moves completely in opposite direction as compared to Fig. 1.2b. Thus although the total force in Fig. 1.2b, d is 5 N the result they produce is different (football moving in opposite direction) because the direction in which the resultant force 5 N acting in Fig. 1.2b, d is exactly opposite.

Fig. 1.2d Person A and Person B kicking football in opposite direction as compared to Fig. 1.2b

6

1 Vector Analysis

Thus the conclusion is although the magnitude of two forces are same if their directions are different they are completely different forces. The same is true for any vector. However, we need not track the direction of total or net mass 9 kgs in Fig. 1.1a, b because mass doesn’t involve direction. The same is true for any scalar. In your lower classes you would have learnt how to add, subtract, multiply differentiate and integrate scalars. It’s very easy. Now in the coming chapters we will learn how add, subtract, multiply, differentiate and integrate vectors.

1.2

Graphical Representation of Vectors

A vector is graphically represented by an arrow. The length of the arrow represents the magnitude of the vector with respect to a pre-chosen scale. The direction of the vector is given by the arrowhead. For example Fig. 1.3 shows the graphical representation of a vector A. The arrow pointing from O to P represents the direction of the vector and the length represents the magnitude (to a suitable scale) of the vector. The end P containing the arrowhead is called the head of the vector whereas the other end O is called the tail of the vector. As for an example in Fig. 1.3 if A is the velocity say for example 20 km/h and in Fig. 1.3 if 1 cm corresponds to 5 km/h then the length of OP will be 4 cm which speciﬁes a velocity of 20 km/h. Fig. 1.3 Graphical representation of vector

1.3

Symbolic Representation of Vectors

The usual representation of a vector is in bold letter like A or a letter with an arrow ! on it A . The magnitude of a vector is usually represented by |A| or A without bold letter. In this book we will follow the notation bold letters A for the vectors and A without bold letter for its magnitude. Care must be taken by the reader to properly identify the vector notation. For example A in this book refers to a vector in both magnitude and direction. On the other hand writing the above A without bold face letter like A means that we are mentioning only the magnitude of the vector.

1.4 Vector Addition

1.4

7

Vector Addition

Figure 1.2 illustrates addition of vectors is not as simple as addition of scalars. In Fig. 1.4a the addition of two vectors A and B is shown. The tail of the B is placed on the head of A. The resultant R is given by the vector drawn from the tail of the vector A to the head of the vector B. In Fig. 1.4a we add vector B to A. In Fig. 1.4b we add vector A to B. In both the cases the resultant is same. Hence vector addition is commutative. AþB ¼ BþA

ð1:1Þ

Also vector addition is associative ðA þ BÞ þ C ¼ A þ ðB þ CÞ

1.5

ð1:2Þ

Subtraction of Vectors

Figure 1.5 shows two vectors A and B. For subtracting B from A reverse the direction of B (without changing the magnitude) as shown in Fig. 1.6 and then add B to A. Fig. 1.5 Subtraction of vectors

Fig. 1.4 Addition of vectors

8

1 Vector Analysis

Fig. 1.6 Subtraction of vectors

1.6

Multiplication of a Vector by a Scalar

If a vector is multiplied by a scalar u then the magnitude of the vector is increased by u times and the direction of the vector remains unchanged. If u = 3 then the magnitude of vector is increased by three times as shown in Fig. 1.7. On the other hand if u is negative then the direction of the vector is reversed. Multiplication of vector by a scalar is distributive. uðA þ BÞ ¼ uA þ uB

1.7

ð1:3Þ

Multiplication of Vectors: Dot Product of Two Vectors

The dot product of two vectors A, B is deﬁned by A B ¼ AB cos h

ð1:4Þ

h is the smallest angle between the two vectors when they are placed tail to tail. Dot product is also sometimes called scalar product because the result of a dot product is a scalar. Dot product obeys commutative and distributive laws. AB¼BA

ð1:5Þ

A ðB þ CÞ ¼ A B þ A C

ð1:6Þ

The geometrical meaning of the dot product can be understood by the aid of Fig. 1.8.

Fig. 1.7 Multiplication of a vector by a scalar

1.7 Multiplication of Vectors: Dot Product of Two Vectors

9

Fig. 1.8 Dot product of vectors

As can be seen from Fig. 1.8 the projection of A on B is A cosh: Geometrically the dot product A B is the product of B and the projection of A on B. If two vectors are parallel then A B ¼ AB cos 0 ¼ AB and if they are perpendicular then A B ¼ AB cos 90 ¼ 0.

1.8

Multiplication of Vectors—Cross—Product of Two Vectors

The cross product of two vectors is denoted by A B and is deﬁned by ^ A B ¼ AB sin h n

ð1:7Þ

^ is the unit vector in the direction perpendicular to the plane containing A, Here n B. However there are two perpendicular directions to the plane as shown in ^ is given by right-hand rule. Let all the ﬁngers of Fig. 1.9. The proper direction of n the right hand except the thumb point in the direction of A and let the ﬁngers curl towards B through the small angle between A and B. Then the thumb gives the ^ or the direction of A B. The entire sequence is shown in direction of n Fig. 1.10a. Similar sequence for obtaining the direction of B A is shown in Fig. 1.10b. The direction of A B and B A as per right-hand rule is indicated in Fig. 1.9. From the ﬁgure it is clear that A B ¼ ðB AÞ

Fig. 1.9 Cross product of vectors

ð1:8Þ

10

1 Vector Analysis

Fig. 1.10a Fixing the direction for the cross product A B

Hence the cross product is not commutative. However the cross product is distributive if the order of vectors is not changed. A ðB þ CÞ ¼ ðA BÞ þ ðA CÞ

ð1:9Þ

For two parallel vectors ^¼0 A B ¼ AB sin 0 n The geometrical meaning of the cross product is jA Bj is the area of the parallelogram generated by A and B in Fig. 1.9.

1.9 Vector Components and Unit Vectors

11

Fig. 1.10b Fixing the direction for the cross product B A

1.9

Vector Components and Unit Vectors

The above vector operations, which we discussed in previous sections, do not involve any particular coordinate system. We will now rewrite the above operations in terms of components in the most widely used coordinate system—Cartesian coordinate system. The unit vectors in the direction of x-, y- and z-axis in the ^ We will discuss vector components in two Cartesian coordinate system are ^i; ^j; k. dimensions and then extend the results to three dimensions.

12

1 Vector Analysis

(a) Two Dimensions Consider Fig. 1.11. The vector A shown in Fig. 1.11 can be resolved into components Ax and Ay. A can be written as A ¼ Ax^i þ Ay^j

ð1:10Þ

Also from Fig. 1.11 cos h ¼

Ax Ay ; sin h ¼ A A

ð1:11Þ

A the magnitude of A is A¼

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ A2x þ A2y

ð1:12Þ

Ax and Ay are themselves scalars. Ax^i; Ay^j are the component vectors. The A can point in any direction but the components vectors Ax^i; Ay^j have constant direction along x- and y-axis, respectively. In your lower classes you might have learnt that ^ where A ^ is the unit vector in the direction of A and is given by A ¼ AA

Fig. 1.11 Resolution of vectors

A ^ ¼ A ¼ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ﬃ A A 2 Ax þ A2y

ð1:13Þ

Ax Ay ^ ¼ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ﬃ ^i þ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ^j A A2x þ A2y A2x þ A2y

ð1:14Þ

1.9 Vector Components and Unit Vectors

13

Fig. 1.12 Example illustrating the resolution of forces

To make the above points clearer let us consider an example. In Fig. 1.12a person A is pulling a block of iron along x-axis with a force of 3 N using a thread tied to the block. The block will move along x-axis. In Fig. 1.12b person B is pulling the same block with a force of 4 N along y-axis using a thread tied to the block. Now the block will move along y-axis (In the above cases numbers are not to be taken seriously). Now consider Fig. 1.12c. Person A is pulling the block along xaxis with a force of 3 N while person B is pulling the block along y-axis with a force of 4 N using threads tied to the block. Assume that the threads are “invisible”—in the sense person A and person B are able to exert forces on the block (or) pull the block using the thread but the block is not constrained to move along xor y-axis because of the thread or in other words the block can move in the x–y plane because of the forces acting on it without getting constrained by the threads. Question: What is the total force Ft acting on the block. What is the direction of Ft or in which direction the block will move?

14

1 Vector Analysis

The total force Ft acting on the block is Ft ¼ 3 ^i þ 4^j The magnitude of the total force is Ft ¼

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 9 þ 16 ¼ 5 N

The angle made by the total force with x-axis is Fx 3 ¼ Ft 5 ) h ¼ 53:1

cos h ¼

The angle is made by Ft with y-axis is ¼ 90 53:1 ¼ 36:9 Thus in Fig. 1.12d a total force of 5 N acts on the block along the line OP, where the line OP makes an angle of 53.1° with the x-axis. Let us restate the above problem in a slightly different manner. In Fig. 1.13 person C is pulling the block along OP with an invisible thread, with a force of Ft = 5 N. The line OP makes an angle of 53.1° with the x-axis. Question: Resolve the force Ft into its x- and y-components. x-component Fx is given by Fx ¼ Ft cos h ¼ 5 cos 53:1 ¼ 3 N y-component Fy is given by Fy ¼ Ft sin h ¼ 5 sin 53:1 Fy ¼ 4 N: Thus person C pulling the block along line OP in Fig. 1.13 is equivalent to person A and B pulling the same block with a force of 3 N and 4 N along x- and yaxis, respectively, as in Fig. 1.12d. Or in other words, we have resolved Ft in

Fig. 1.13 Example illustrating the resolution of forcesExample illustrating the resolution of forces

1.9 Vector Components and Unit Vectors

15

Fig. 1.14 Figure illustrating that mass cannot be resolved into components

Fig. 1.13 into equivalent x- and y-components Fx and Fy, respectively. The same is true for any vector. That is, any vector can be resolved into its components. What about resolving scalars. To ﬁnd out, consider a particle of mass 5 kg located as shown in Fig. 1.14. Can you resolve the mass of 5 kg into its components along x- and y-axis. Mass being a scalar quantity is independent of direction and cannot be resolved into components. The same is true for any scalar. (b) Three Dimensions Now let us generalize the above results for three dimensions. In Fig. 1.15a we ^ The vector A shown in Fig. 1.15b show x, y, z along with their unit vectors ^i; ^j; k. ^ ^ ^ can be written in terms of i; j; k as ^ A ¼ Ax^i þ Ay^j þ Az k

ð1:15Þ

and cos a ¼

Ax Ay Az ; cos b ¼ ; cos C ¼ A A A

ð1:16Þ

A the magnitude of A is A¼

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ A2x þ A2y þ A2z

ð1:17Þ

Fig. 1.15 a Unit vectors in Cartesian Coordinate system. b A in Cartesian Coordinate system

16

1 Vector Analysis

^ are the component vectors. Ax, Ay, Az are themselves scalars. Ax ^i ; Ay ^j ; Az k ^ is given by The unit vector A ^ ^ ^ x i þ Ay j þ Az k ^ ¼A¼A qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ A A A2 þ A2 þ A2 x

Ax ^i + ^ ¼ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ A 2 Ax + A2y + A2z

y

ð1:18Þ

z

Ay qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ^j + 2 Ax + A2y + A2z

Az ^ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ k 2 Ax + A2y + A2z

ð1:19Þ Writing any vector in terms of its components has number of advantages. (c) Vector addition and subtraction in terms of components: To add any two vectors A and B just add the components ^ + (Bx^i + By^j + Bz kÞ ^ A + B ¼ (Ax^i + Ay^j + Az kÞ ^ A + B ¼ (Ax + Bx Þ^i + (Ay + By Þ^j + (Az þ Bz Þk

ð1:20Þ

To subtract two vectors A and B just subtract the components ^ ðBx^i + By^j + Bz kÞ ^ A B ¼ ðAx^i + Ay^j + Az kÞ ^ A B ¼ ðAx Bx Þ^i + (Ay By Þ^j + (Az Bz Þk

ð1:21Þ

Addition and subtraction of components are similar to that of scalars. (d) Multiplication of a vector by a scalar: To multiply a vector by a scalar say u, multiply each component by a respective scalar ^ uA ¼ uðAx^i + Ay^j + Az kÞ ^ uA ¼ uAx^i + uAy^j + uAz k

ð1:22Þ

(e) Dot product in terms of components: ^i; ^j; k ^ are mutually perpendicular to each other. Hence ^i ^i ¼ ^j ^j ¼ k ^ ^i ^j ¼ ^i k ^ ¼ ^j

^¼ 1 k ^¼ 0 k

ð1:23Þ

1.9 Vector Components and Unit Vectors

17

Thus the dot product can be written as ^ ðBx^i + By^j + Bz kÞ ^ A B ¼ ðAx^i + Ay^j + Az kÞ A B ¼ Ax Bx + Ay By + Az Bz

ð1:24Þ

(f) Cross product in terms of components: 9 ^i ^i ¼ ^j ^j ¼ k ^k ^ ¼ 0> > > > > = ^i ^j ¼ ^j ^i ¼ k ^ ^j k > ^ ¼ k ^ ^j ¼ ^i > > > > ; ^ ^i ¼ ^i k ^ ¼ ^j k

ð1:25Þ

^ ðBx ^i þ By ^j þ Bz kÞ ^ A B ¼ ðAx ^i þ Ay ^j þ Az kÞ

ð1:26Þ

Therefore

^ ¼ ðAy Bz Az By Þ ^i þ ðAz Bx Ax Bz Þ ^j þ ðAx By Ay Bx Þ k

ð1:27Þ

which can be written in much more familiar form as ^i A B ¼ Ax B x

1.10

^j Ay By

^ k Az Bz

ð1:28Þ

Triple Products

Scalar triple product is deﬁned as A ðB CÞ. The geometrical meaning of A ðB CÞ can be understood with the aid of Fig. 1.16. A ðB CÞ ¼ A cos h ðB CÞ

Fig. 1.16 Geometrical meaning of scalar triple product

ð1:29Þ

18

1 Vector Analysis

We have already seen that cross product B C is the area of the parallelogram (Area of the base in Fig. 1.16). Also from Fig. 1.16 A cos h is the vertical height of the parallelepiped. Hence, A ðB CÞ ¼ (Vertical height of parallelepiped) (Area of base) ¼ Volume of the parallelepiped: Thus A ðB CÞ is the volume of the parallelepiped, whose sides are A, B and C. As any face of the parallelepiped can be taken as the base A ðB CÞ ¼ B ðC AÞ ¼ C ðA BÞ Note that in the above expression cyclic alphabetical order A, B, C is maintained. In component form Ax Ay Az ð1:30Þ A ðB CÞ ¼ Bx By Bz Cx Cy Cz (b) Vector triple product: The vector triple product is deﬁned as A ðB CÞ ¼ BðA CÞ CðA BÞ

ð1:31Þ

The reader should note that we have written BðA CÞ. The statement B ðA CÞ has no meaning. Similarly the statement B ðA CÞ also has no meaning. Example 1.1 ^ and B ¼ 5^i þ t^j þ 6k ^ are perpenFind the value of t if the vectors A ¼ 2^i þ ^j þ k dicular to each other. Solution If A and B are perpendicular to each other then A B = 0 ^ ð5^i + t^j + 6kÞ ^ =0 ) ð2^i + ^j + kÞ ) 10 + t + 6 = 0 ) t ¼ 16 Example 1.2 ^ B ¼ 3 ^i 6^j þ 9k. ^ Calculate the angle between the vectors A ¼ 3^i þ ^j þ 2k,

1.10

Triple Products

19

Solution We have ^ 3^i 6^j þ 9k ^ A B ¼ 3^i þ ^j þ 2k ¼ 9 6 þ 18 ¼ 21: pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ pﬃﬃﬃﬃﬃ A ¼ 9 þ 1 þ 4 ¼ 14 pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ pﬃﬃﬃﬃﬃﬃﬃﬃ B ¼ 9 þ 36 þ 81 ¼ 126 We know that A B ¼ AB cos h AB 21 ¼ pﬃﬃﬃﬃﬃpﬃﬃﬃﬃﬃﬃﬃﬃ AB 14 126 21 1 pﬃﬃﬃﬃﬃpﬃﬃﬃﬃﬃﬃﬃﬃ h ¼ cos 14 126 h ¼ 60

cos h ¼

Example 1.3 ^ and B ¼ 2^i þ 3^j þ k ^ are parallel to Check whether the vectors A ¼ 6^i 9^j 3k each other. Solution We have ^ k 9 3 3 1 ^½18 18 ¼ 0 ¼ ^i½9 þ 9 ^j½6 6 þ k

^i AB ¼ 6 2

^j

¼0 AB ¼0 )A B sin h ¼ 0 As A, B the magnitude of A, B are not zero sin h ¼ 0 ) h ¼ 0: Hence A and B are parallel to each other.

20

1 Vector Analysis

Example 1.4 ^ B ¼ 9^i þ 3^j þ 6k ^ and C ¼ Check whether the vectors A ¼ 2^i þ 4^j þ 18k, ^ ^ ^ 4i þ 2j þ 6k are coplanar. Solution The condition that the three vectors are coplanar is that their scalar triple product is zero A ðB CÞ ¼ 0 we have

18 6 6

2 4 A ðBCÞ ¼ 9 3 4 2

¼ 2ð18 12Þ 4ð54 24Þ þ 18ð18 12Þ ¼ 2ð6Þ 4ð30Þ þ 18ð6Þ ¼ 0: Hence the vectors are coplanar.

1.11

Line, Surface and Volume Integration

In this section we will learn about the integration of vectors. There are three different types of vector integration, namely line, surface and volume integrations. (a) Line Integration Let us assume a scalar function u and a vector function A. There are three important line integrations ZQ

ZQ A dl;

u dl; P

ZQ

P

A dl; P

where in general the line integral will depend on the path. For example the above integrations will yield different results for path 1 and path 2 as shown in Fig. 1.17. RQ Out of the above three integrals P A d l will often appear in electromagnetics. If the path over which the line integral is carried out is closed we shall put a circle over the integration indicating that the path is closed

1.11

Line, Surface and Volume Integration

21

Fig. 1.17 Illustration to explain line integration

I A dl Although in general the line integration depends on path, for certain vectors the line integration doesn’t depend on path, but only on end points. For example for these certain vectors the line integration over the path 1 and path 2 in Fig. 1.17 will yield same result and will depend on only the end points P and Q. (b) Surface Integration Consider a surface S. Let the surface S be divided into small elements ds as shown in Fig. 1.18. There are three different surface integrals. ZZ ZZ ZZ u ds; A ds; A ds: S

Fig. 1.18 Figure showing Surface S

S

S

22

1 Vector Analysis

Fig. 1.19 a Closed surface showing direction of the surface. b Open surface showing direction of the surface

RR where u is a scalar. Out of the three integrals the surface integral S A ds will often appear in electromagnetics and it is customary to write the above integral for open and closed surface as Z I A ds and A ds S

S

Surface is a vector and has direction. For a closed surface (assume the surface of a tennis ball, which is a closed surface) the outward drawn normal is the direction of the surface at that point as shown in Fig. 1.19a. For an open surface the direction of the surface is the perpendicular direction to the surface at that point (assume a piece of paper which is an open surface). However there are two perpendicular direction to the open surface and hence the direction is arbitrary as shown in Fig. 1.19b. (c) Volume Integration Volume is a scalar and hence performing volume integration is easy. There are two different volume integrations ZZZ ZZZ u ds; A ds s

s

or Z

Z u ds;

s

A ds s

where s is the volume over which the integration is performed and ds is the volume element.

1.12

1.12

Flux

23

Flux

Flux is an important quantity which we will encounter often in electromagnetics. In this section we will learn how to deﬁne flux. Assume a rectangular cardboard abcd as shown in Fig. 1.20, whose surface area is S. Suppose the cardboard is exposed to sun rays and is held perpendicular to the sun rays as shown in Fig. 1.20a. A shadow forms below as shown in the ﬁgure. Because the cardboard is held perpendicular to the sun rays the area of the shadow formed will be exactly S. However now if the cardboard is held at an angle h to the sun rays that is, instead of being perpendicular to the sun rays, the cardboard is inclined with respect to the direction of the sun rays at an angle h as shown in Fig. 1.20b. Now the area of the shadow will be lesser. This is because the effective area shown by the cardboard to the incoming sun rays is lesser than S, when the cardboard is inclined at an angle h to the direction of sun rays. When the cardboard is perpendicular to the sun rays all the area S of the cardboard is shown to the sun rays. Hence the area of the shadow formed is S. But when the cardboard is inclined at angle h to the sun rays the effective area of the cardboard shown to the sun rays is S cos h and hence the area of the shadow formed is S cos h. In other words the projection of surface S is S cos h. Of course when the cardboard is held perpendicular to the sun rays h ¼ 0 and hence the effective area is S cos 0 ¼ S. Now suppose in the place of cardboard assume a surface abcd whose area is S and in the place of sun rays assume an electric ﬁeld E vector. The effective area shown by the surface abcd to the incoming vector E will be S cos h. Flux is deﬁned as the number of lines passing through the given area. Hence the total flux passing through the surface abcd in Fig. 1.20b is ¼ EðS cos hÞ ¼ E S

Fig. 1.20 a Cardboard held perpendicular to the sun rays. b Cardboard held at an inclined angle to the sun rays

24

1 Vector Analysis

However if we divide the surface into small elements ds then the total flux through abcd will be ZZ Flux ¼ E ds s

In Sect. 1.11b we have treated area as a vector. Throughout electromagnetics we will be treating area as a vector. Now a general question is whether area is vector or a scalar. In Sect. 1.1 and in Fig. 1.1 masses sum up and give the same result in whichever direction they are added. Quantities which behave like mass are called scalars. In Fig. 1.2 forces sum up or cancel out depending upon the direction in which they act. Quantities which behave like force are called vectors. If you add area of two plots (empty land) they sum up irrespective of whatever direction they are added and hence area must be a scalar. But in electromagnetics we treat area as vector. How is this possible? The answer lies in Fig. 1.20. The effective area shown by area abcd to the incoming vector depends on the “DIRECTION” of orientation of normal of the area abcd to the incoming vector E. If the “DIRECTION” of the normal of the area abcd is along the direction of the vector then the effective area shown to the incoming vector is S cos 0 ¼ S. If the “DIRECTION” of the normal of the area abcd is oriented at an angle h to the incoming vector the effective area shown to the incoming vector is S cos h. Because the amount of effective area shown by the area abcd to the incoming vector depends on the angle made by ‘DIRECTION’ of the normal of the area abcd to the incoming vector, area must be treated as vector.

1.13

Vector Differentiation: Gradient of a Scalar Function

There are three important types of vector differentiation—gradient, divergence and curl. In this section we will discuss gradient. In Sects. 1.14, 1.15 we will discuss divergence and curl. As we will see the gradient of a scalar function u in Cartesian coordinates is given by ru ¼

@u ^ @u ^ @u ^ k iþ jþ @x @y @z

The physical meaning of gradient will be clear to the reader if the reader considers ddyx where ddyx gives the rate of change of y with respect to x when y is a function of x alone y = y (x). However we have number of scalar functions which depend on three variables (x, y, z) say for example u = u (x, y, z). Similar to ddyx, r u gives the maximum rate of change of u with respect to x, y, z. As we have ddyx when y depends

1.13

Vector Differentiation: Gradient of a Scalar Function

25

Fig. 1.21 Figure showing Gradient of a Scalar

on single variable y = y (x), we have r u when the given scalar function u depends on three variables (x, y, z). Consider a scalar function u (x, y, z) which is constant over the surface 1 as shown in Fig. 1.21. Consider another surface, surface 2 on which the scalar function has a value u + du where du is small change in u. Point A lies on surface 1 and points B and C lie on surface 2. Consider an arbitrary origin and with respect to the origin let the position vectors of A and C be r and r + dr. We have AC = dr. In Cartesian coordinates ^ dr ¼ dx ^i + dy ^j + dz k

ð1:32Þ

From calculus du ¼

@u @u @u dx þ dy þ dz @x @y @z

ð1:33Þ

The above expression can be written as du ¼

@u ^ @u ^ @u ^ ^ ^ k : dx i þ dy ^j þ dz k iþ jþ @x @y @z

Let us denote the vector

@u @x

ð1:34Þ

^i þ @ u ^j þ @ u k ^ as G. @y @x

Hence du ¼ G dr

ð1:35Þ

Now let us assume point C lies on surface 1. In that case du = 0. Hence Eq. 1.35 is G dr ¼ 0: G dr cos a ¼ 0

ð1:36Þ

26

1 Vector Analysis

where a is the angle between G and dr. We will ﬁrst ﬁnd direction of G. Equation 1.36 can be zero only if a = 90°. As dr points from A to C, dr will be lying on surface 1 when point C lies on surface 1. As the angle a between dr and G is 90° and dr is lying on surface 1, the vector G is normal to the surface 1. We have identiﬁed the direction of G. It is normal to the surface 1. In Fig. 1.21 let AB = dn. From Fig. 1.21 cos h ¼

dn dr

) dn ¼ dr cos h

ð1:37Þ ð1:38Þ

^ be the unit vector in the direction of AB. Then Eq. 1.38 can be written as Let n ^ dr dn ¼ n

ð1:39Þ

As we move from point A to B u increases by du. Hence u ¼ uð nÞ @u du ¼ dn @n

ð1:40Þ

Substituting (1.39) in (1.40) du ¼

@u ^ dr n @n

ð1:41Þ

@u ^ n @n

ð1:42Þ

Comparing (1.41) and (1.35) G¼

Thus the magnitude of G is @@un and its direction is normal to the surface 1. @u @u @u @u Consider @u @n, @r in Fig. 1.21. We see that out of the two derivatives @n, @r , @n is greater @u than @r because dn in Fig. 1.21 is lesser than dr. If we consider D, E or any other point on surface 2 always AB = dn will be minimum as compared to AD, AE, etc. Hence @@un will be the maximum value of the derivative pointing normal to surface 1. Hence from Eqs. 1.34, 1.35, 1.42 G¼

@u ^ @u ^ @u ^ @u ^ k¼ n i + jþ @x @y @z @n

ð1:43Þ

“Thus the gradient of the scalar function u is a vector. Its magnitude is equal to the maximum rate of change of scalar function u with respect to scalar variables. The direction of this vector is along this maximum rate of change of scalar function”.

1.13

Vector Differentiation: Gradient of a Scalar Function

27

We denote the gradient of a scalar function as grad u or ru. G ¼ grad u ¼ ru ¼

@u ^ @u ^ @u ^ k iþ jþ @x @y @z

ð1:44Þ

The above equation can be written as @ ^@ @ ^ ^ þj þk ru ¼ i u @x @y @z

ð1:45Þ

@ @ ^ @ is called “del” operator. Here r ¼ ^i @x þ ^j @y þk @z

ru serves the same purpose in three dimensions as ddyx in one dimension. That is in Fig. 1.21 if we represent AD = dr1, AE = dr2, etc. We have number of derivations @@un; @@ur ; @[email protected] u1 ; @@ru2 . . . etc. Out of the above available derivatives @@un is the derivative selected to represent the rate of change of the scalar function with space variables (x, y, z) because @@ un gives the maximum rate of change of the scalar function with respect to space variables. y As d dx gives the rate of change of y with respect to x when y depends on x only that is y = y (x), ru gives the information how fast u varies with x, y, z when u depends on (x, y, z) that is u = u (x, y, z). Some of the examples in which the given scalar function depends on three variables (x, y, z) are pressure of a gas in a cubical box P = P (x, y, z), temperature inside a room T = T (x, y, z). The other derivatives like du dr are related to gradient as follows. From Eqs. 1.32, 1.34 and 1.44 we can write du ¼ ru:dr du ¼ ru ^r dr The magnitude of du dr depends on direction of dr and hence it is called as directional derivative.

1.14

Vector Differentiation: Divergence of a Vector

The word divergence reminds us the word “diverge” which means to “spread out”. Divergence is the measure of how much the given vector diverges or spreads out. For example for the vector shown in Fig. 1.22a the vector spreads out and is having large positive divergence. The vector shown in Fig. 1.22b is having large negative divergence. The vector shown in Fig. 1.22c is having zero divergence. Suppose we put a hole (marked p in Fig. 1.23a) in a floor and ﬁx a pipe in it and allow water to come out as shown in Fig. 1.23a, water spreads out; hence water is

28

1 Vector Analysis

Fig. 1.22 Divergence of vector ﬁelds

Fig. 1.23 a Water flowing outside the pipe. b Water flowing inside the pipe

diverging and is having positive divergence (in vectorial terms the velocity of water is diverging). On the other hand, we put a hole (marked q in Fig. 1.23b) in a cone-like structure and allowed water to flow inside, from the edges of the cone, then the water is having negative divergence (in vectorial terms it is the velocity of water having negative divergence). We can get another important physical information from Fig. 1.23. In Fig. 1.23a point p is acting like a source from which water emanates from and diverges. Hence presence of positive divergence implies implicitly that there is a source for the associated vector from which the vector emanates from. In Fig. 1.23b point q is acting like a sink into which water flows in. Hence presence of negative divergence implies that there is a sink for the associated vector to which the vector converges to. In Cartesian coordinates the divergence is deﬁnedas (for any vector A) i @ ^@ @ h ^ ^ ^ ^ r A¼ i þj þk : Ax i þ Ay^j þ Az k @x @y @z rA ¼

@Ax @Ay @Az þ þ @x @y @z

ð1:46Þ ð1:47Þ

1.14

Vector Differentiation: Divergence of a Vector

29

(1) In Fig. 1.22a if the diverging vector is A then r A [ 0 (This implies that there is a source for A. Point P in Fig. 1.22a is acting like a source for A). (2) In Fig. 1.22b if the vector depicted is A having negative divergence then r A \ 0 (This implies that there is a sink for A. Point Q in Fig. 1.22b is acting like a sink for A). (3) In Fig. 1.22c if the depicted vector is A then r A ¼ 0 (There is no source or sink for A). As we will see in Chap. 4 the above example will be used to explain magnetic monopoles do not exist. A vector alone can have divergence and a scalar cannot have divergence. For example if the mass of an object is 200 kg, it is 200 kg as itself and it cannot diverge or spread out. (If the mass of 200 kg explodes and spreads out, it will be the velocity of each piece which will be diverging, not the mass of the individual piece). We will now see the mathematical description of divergence. As shown in Fig. 1.22 the divergence of any vector A represents flux. In Fig. 1.22a there is outward flow of flux from the source at point P. In Fig. 1.22b there is inward flow of flux to the sink at point Q. In Fig. 1.22c there is an equal amount of inward and outward flux going through any volume resulting in zero net flux. “Thus the divergence of a vector function A is deﬁned as the flux per unit volume in the limit of volume enclosed by the closed surface tends to zero”. Consider a point P0 whose coordinates are x0 ; y0 ; z0 with respect to some arbitrary origin. Consider a small closed surface S surrounding the point P0 . Let s be the small volume enclosed by the surface S. Then the divergence of A at point P0 is deﬁned as H div A ¼ lim

s!0

A ds s

ð1:48Þ

H where in the above equation A ds represents flux (see Sect. 1.12). Consider an inﬁnitesimal rectangular box whose sides are D x; D y; D z as shown in Fig. 1.24. The box has six faces—ABQP, CRSD, ABCD, PQRS, BCRQ, ADSP. From Fig. 1.24 Area ðD sÞABQP ¼ Dy D z^i Area ðD sÞCRSD ¼ Dy D z^i Area ðD sÞABCD ¼ D xD z^j Area ðD sÞPQRS ¼ D xD z^j

30

1 Vector Analysis

Fig. 1.24 Differential volume in Cartesian coordinate system

^ Area ðD sÞBCRQ ¼ D xDyk ^ Area ðD sÞADSP ¼ D xDyk The vector A in Cartesian coordinates can be written as ^ A ¼ Ax^i + Ay^j + Az k For the box shown in Fig. 1.24 the integral I

Z

A:ds can be written as

Z

A ds ¼ S

H

Z

A ds þ ABQP

A ds þ CRSD

Z

þ

ABCD

Z

A ds þ PQRS

A ds

A ds þ BCRQ

R

Now consider the integral

Z

ð1:49Þ A ds

ADSP

A ds

ABQP

Z A ds ¼ ðAÞABQP ðDsÞABQP

ð1:50Þ

ABQP

h i ^ ¼ Ax^i + Ay^j + Az k

ABQP

DyDz^i

ð1:51Þ

Z A d s ¼ ½Ax ABQP DyDz ABQP

ð1:52Þ

1.14

Vector Differentiation: Divergence of a Vector

31

Because the box is inﬁnitesimal the value of Ax at the centre of the ABQP face can be taken to be average value of Ax over the entire face. Hence Eq. 1.52 is Z A ds ¼ ½Ax DyDzatðx0 þ Dx;y0 ;z0 Þ ð1:53Þ 2

ABQP

Applying Taylor’s series expansion and neglecting higher order terms as the box is inﬁnitesimal Dx @Ax ½Ax atðx0 þ Dx;y0 ;z0 Þ ¼ Ax ðx ; y ; z Þ þ 2 2 @x ðx0 ;y0 ;z0 Þ 0

0

0

Substituting the above equation in 1.53 Z ABQP

"

# D x @Ax A : ds ¼ Ax ðx ; y ; z Þ þ DyD z 2 @x ðx0 ;y0 ;z0 Þ 0

0

0

ð1:54Þ

Similarly for the face CRSD Z A d s ¼ ðAÞCRSD DsCRSD

ð1:55Þ

CRSD

h

^ ¼ Ax^i + Ay^j + Az k

i CRSD

DyD z ^i

¼ ½Ax DyD zatðx0 D x;y0 ;z0 Þ 2

ð1:56Þ ð1:57Þ

Using Taylor series expansion and neglecting higher order terms D x @ Ax ½Ax atðx0 þ D x;y0 ;z0 Þ ¼ Ax ðx ; y ; z Þ 2 2 @x ðx0 ;y0 ;z0 Þ 0

0

0

Substituting the above equation in 1.57 Z CRSD

"

# D x @ A x A d s ¼ Ax ðx0 ; y0 ; z0 Þ Dy Dz 2 @ x ðx0 ;y0 ;z0 Þ

ð1:58Þ

Adding Eqs. 1.54, 1.58 we get Z

Z A dsþ ABQP

A ds ¼ CRSD

@ Ax Dx Dy Dz @ x ðx0 ;y0 ;z0 Þ

ð1:59Þ

32

1 Vector Analysis

Applying the same procedure to ABCD and PQRS we get Z

Z A dsþ

ABCD

A ds ¼ PQRS

@ Ay @y

ðx0 ;y0 ;z0 Þ

Dx Dy Dz

ð1:60Þ

Dx Dy Dz

ð1:61Þ

Similarly for the faces BCRQ and ADSP Z

Z A dsþ

BCRQ

A ds ¼ ADSP

@ Az @z

ðx0 ;y0 ;z0 Þ

Substituting Eqs. 1.59, 1.60, 1.61 in Eq. 1.49

I A ds ¼ S

@ Ax @ Ay @ Az þ þ @x @y @z

ðx0 ;y0 ;z0 Þ

Dx Dy Dz

ð1:62Þ

With s ¼ Dx Dy Dz, the volume of the box, the above equation becomes I A ds ¼ S

@ Ax @ Ay @ Az þ þ s @x @y @ z ðx0 ;y0 ;z0 Þ

ð1:63Þ

Substituting Eq. 1.63 in Eq. 1.48 div A ¼

@Ax @Ay @Az þ þ @x @y @z

ð1:64Þ

The above equation applies for any point ðx0 ; y0 ; z0 Þ. Equation 1.64 can be written as

1.15

@ ^@ @ ^ ^ ^ þj þk div A ¼ i Ax^i þ Ay^j þ Az k @x @y @y

ð1:65Þ

div A ¼ r A

ð1:66Þ

Vector Differentiation: Curl of a Vector

In your daily life you could have come across the word “curly hair”. The meaning of word curl is circulation. Curl of the vector is the measure of how much the given vector circulates about the given point in question. For example, the vector shown in Fig. 1.25 is having high curl as it is circulating about a given point. A vector alone can curl, scalar cannot curl. For example if we consider a block of mass 200 kg, it is 200 kg as itself and the value 200 kg doesn’t

1.15

Vector Differentiation: Curl of a Vector

33

Fig. 1.25 Curl of a vector ﬁeld

Fig. 1.26 A point of high curl

circulate. On the other hand consider a swirl as shown in Fig. 1.26 which is having high curl or circulation. The velocity of water at each point is different and circulates about the central point. We will now see the mathematical description of curl. The curl of the vector ﬁeld at any point is a vector quantity, the magnitude of the vector quantity being given by maximum line integral per unit area, the line integral being carried out along the boundary of inﬁnitesimal test area at that point and the direction of the vector is perpendicular to the plane of the test area. As per the above deﬁnition the curl of any vector A is deﬁned as ^ n curl A ¼ lim

Ds!0

H

A dl

C

Ds

max

ð1:67Þ

Here Ds is the test area which is enclosed by contour C as shown in Fig. 1.27. ^ is the unit vector normal to the test area. We will see how to ﬁx the direction Here n ^ later on. of curl or n Fig. 1.27 Figure showing relation between different unit vectors

34

1 Vector Analysis

Now let us deﬁne the component of curl ^C ðcurl AÞ ðcurl AÞC ¼ n H A dl ¼ lim

CC

DSC !0

ð1:68Þ

DsC

^c is the unit vector normal to the area DsC , where DsC is the area bounded Here n by contour Cc. ^ in Fig. 1.27 is given by right-hand rule. When the ﬁngers curl The direction of n ^ which is perpendicular to the area Ds . around C the thumb gives the direction of n ^c which is perpendicular to area DsC , area DsC The same right-hand rule is true for n being bounded by contour CC. Contour CC is not shown in Fig. 1.27. Consider an inﬁnitesimal loop PQRS in the z – y plane surrounding the point P0 in the region of vector function A, as shown in Fig. 1.28. The positive direction around the loop is assigned by right-hand rule. When the ﬁngers of the right hand point in the direction of the arrows in the contour PQRS in Fig. 1.28 then the thumb gives the direction of the positive normal to the surface. This positive normal for the loop PQRS is along positive x-axis. Now let us calculate the line integral of A along PQRS.

Fig. 1.28 Loop PQRS in Cartesian coordinate system

1.15

Vector Differentiation: Curl of a Vector

35

The line integral of A along PQ is Z h i ^ Dy^j A dl ¼ Ax^i + Ay^j + Az k PQ

¼ Ay Dy at PQ

Because the loop is inﬁnitesimal the value of Ax at the centre of the line PQ can be taken to be average value of Ax over the entire line. Hence the above equation becomes Z

A dl ¼ Ay Dy atðx0 ;y0 ;z0 D zÞ ð1:69Þ z

PQ

The line integral of A along RS will be Z

h i ^ Dy^j A dl ¼ Ax^i þ Ay^j þ Az k

RS

ð1:70Þ

¼ Ay Dy atðx0 ;y0 ;z0 þ DzÞ z

The line integral of A along SP is Z

h i

^ Dzk ^ A dl ¼ Ax^i þ Ay^j þ Az k ð1:71Þ

SP

¼ ½Az Dzatðx0 ;y0 Dy;z0 Þ z

The line integral of A along QR is Z

h i

^ Dzk ^ A dl¼ Ax^i þ Ay^j þ Az k ð1:72Þ

QR

¼ ½Az Dzatðx0 ;y0 þ Dy;z0 Þ z

The line integral of I

PQRS

A dl can be written as

I A dl ¼

PQRS

H

I A dl þ

PQ

I A dl þ

RS

I A dl þ

SP

A dl QR

ð1:73Þ

36

1 Vector Analysis

Substituting Eqs. 1.69, 1.70, 1.71, 1.72 in Eq. 1.73 we get I h i A dl ¼ Ay ðx0 ;y0 ;z0 Dz;Þ Ay ðx0 ;y0 ;z0 þ Dz;Þ Dy z

PQRS

z

h

i

ð1:74Þ

þ ðAz Þðx0 ;y0 þ Dy;z0 Þ ðAz Þðx0 ;y0 Dz;z0 Þ Dz z

z

Using Taylor’s series expansion, neglecting higher order terms we get h

i

@ Ay Dz Ay ðx0 ;y0 ;z0 Dz;Þ Ay ðx0 ;y0 ;z0 þ Dz;Þ ¼ z z @z

ð1:75Þ

and h

i @ Az ðAz Þðx0 ;y0 þ Dy;z0 Þ ðAz Þðx0 ;y0 Dy;z0 Þ ¼ Dy z z @y

ð1:76Þ

Substituting Eqs. 1.75 and 1.76 in Eq. 1.74 I A dl ¼ PQRS

@ Az @ Ay D yD z D zD y @y @z

ð1:77Þ

With Ds ¼ DyDz we get

I PQRS

@ Az @ Ay A dl ¼ Ds @y @z

ð1:78Þ

Substituting Eq. 1.78 in Eq. 1.68 ðcurl AÞx ¼ ^i ðcurl AÞ H ¼ limDs!0

PQRS

A dl

Ds

¼

@ Az @ Ay @y @z

ð1:79Þ

Similarly ðcurl AÞy ¼

@ Ax @ Az @z @x

ð1:80Þ

1.15

Vector Differentiation: Curl of a Vector

37

and ðcurl AÞz ¼

@ Ay @ Ax @x @y

ð1:81Þ

We know that ^ ðcurl AÞ ¼ ðcurl AÞx^i þ ðcurl AÞy^j þ ðcurl AÞz k

ð1:82Þ

Substituting Eqs. 1.79, 1.80, 1.81 in Eq. 1.82 curl A ¼

@Az @Ay ^ @Ax @Az ^ @Ay @Ax ^ iþ jþ k @y @z @z @x @x @y

ð1:83Þ

The above equation can be written as h i @ ^@ @ ^ ^ ^ þj þk Ax^i þ Ay^j þ Az k curl A ¼ i @x @y @z

ð1:84Þ

curl A ¼ r A

ð1:85Þ

In determinant form r A is ^i @ r A ¼ @x A

x

^j

@ @y Ay

^ k @ @z Az

ð1:86Þ

We have seen three different types of vector differentiation. 1. Gradient: The operator r operates on an scalar function. In Cartesian coordinates ru ¼

@u ^ @u ^ @u ^ k iþ jþ @x @y @z

As seen above gradient of a scalar function is a vector. Divergence: The operator r operates on a vector by means of dot product. In Cartesian coordinates r A¼

@Ax @Ay @Az þ þ @x @y @z

As seen above divergence of a vector function is a scalar.

38

1 Vector Analysis

Curl: The operator r operates on a vector by means of cross product. In Cartesian coordinates rA¼

@Az @Ay ^ @Ax @Az ^ @Ay @Ax ^ iþ jþ k @y @z @z @x @x @y

As seen above the curl of a vector function is a vector. Because curl of any vector is a vector it must have both magnitude and direction. Let us see an example. ^ be curling around In Fig. 1.25 let the velocity vector v given by vx^i þ vy^j þ vz k with O as centre. Let us deﬁne curl of v as

@vz @vy ^ @vx @vz ^ @vy @vx ^ C¼rv¼ k iþ jþ @y @z @z @x @x @y Then the direction of C and the curl of v are given by right-hand rule. When the ﬁngers of right hand point in the direction of circulating v as shown in Fig. 1.25 then the thumb gives the direction of C which is along z-axis. Thus there are two vectors v, C :v the velocity which is circulating about and C, the curl of v pointing along z-direction. 2. We have seen that divergence of any vector means how much the vector is spreading out. Curl of any vector means how much the vector is circulating about. If r A ¼ 0 then A is called solenoidal vector If r A ¼ 0 then A is known as irrotational vector. For all the vectors shown in Fig. 1.22 curl of these vectors is zero because nothing is circulating. For the vector shown in Fig. 1.25 divergence is zero because nothing is spreading out. 3. Divergence and curl operate on a vector. As we have already seen divergence and curl of a scalar don’t have a meaning. Example 1.5 Find the gradient of u at the point (1, 2, 3) if u = x2 – yz. Solution

@ ^@ ^ @ x2 yz þj þk ru ¼ ^i @x @y @z ^ ^ ^ )ru ¼ 2xi þ zj þ yk Gradient of a scalar function u, r u is a vector as we discussed previously.

1.15

Vector Differentiation: Curl of a Vector

39

r u at the point (1, 2, 3) is then ^ ru ¼ 2^i þ 3^j þ 2k Example 1.6 ^ B ¼ x2 y^i þ z3 y^j. Find the divergence of the vector A ¼ x^i þ y^j þ zk; Solution @ ^@ ^ @ : x^i þ y^j þ zk ^ þj þk r A ¼ ^i @x @y @z r A ¼3 @ ^@ ^ @ x2 y^i þ z3 yk ^ þj þk r B ¼ ^i @x @y @z r B ¼ 2xy þ z3 Divergence of A is 2, divergence of B is 2xy þ z3 . Example 1.7 ^ Calculate curl of vector A ¼ yk. Solution ^i @ r A ¼ @x 0

^j @ @y

^ k @ @z y

0 @y ^ ½ 0 0 r A ¼ ^i 0 ^j½0 0 þ k @y r A ¼ ^i Example 1.8 Prove that for any vector A, div curl A = 0. That is prove that r ðr AÞ ¼ 0.

40

1 Vector Analysis

^i ^j ^ k @ @ @ r ðr AÞ ¼ r @x @y @z A A A x y z @ ^@ @ @Az @Ay ^ @Ax @Az ^ @Ay @Ax ^ ^ ^ ¼ i þj þk iþ jþ k @x @y @z @y @z @z @x @x @y @ @Az @Ay @ @Ax @Az @ @Ay @Ax þ þ ¼ @x @y @y @z @z @x @z @x @y ¼

@ 2 Az @ 2 Ay @ 2 Ax @ 2 Az @ 2 Ay @ 2 Ax þ þ ¼0 @x @y @x @z @y @z @y @x @z @x @z @y

Example 1.9 Prove that curl grad u = 0. That is prove that r ru ¼ 0 where u is a scalar function. Solution ^i ^j ^ k @ @ @ r ru ¼ @x @y @z @u @u @u @x @y @z 2 2 2 @ u @2u ^ @ u @2u ^ @ u @2u ^ ¼ iþ jþ k¼0 @y @z @z @y @z @x @x @z @x @y @y @x

1.16

Divergence Theorem

Divergence theorem is helpful in transforming a surface integral of a vector (flux of the vector) into a volume integral of the divergence of the vector and vice versa. The theorem states that the volume integral of the divergence of the vector A is equal to the flux of the vector over the closed surface S enclosing the volume s over which the volume integral is calculated, i.e. Z I r Ads ¼ A ds ð1:87Þ s

S

Divergence theorem is also known as Gauss’s divergence theorem. We have already seen that divergence of a vector is a scalar quantity (see Sect. 1.15). R The volume s is also a scalar quantity. Hence when dealing with integral s r H Ads we will be working with scalars, on the other hand in the integral S A ds,

1.16

Divergence Theorem

41

Fig. 1.29 a Arbitrary volume bounded by a surface. b Arbitrary volume divided into two small volumes

H surface ds is a vector quantity. Hence when working with integral S A ds we will be working with vectors. This point will be more clear in Example 1.10. Now we shall see the proof for divergence theorem. Let us consider some arbitrary volume s which is enclosed by surface S as shown in Fig. 1.29a. The arbitrary volume is present in a region of vector function A. The flux diverging from the surface S of volume s is I A ds

/¼

ð1:88Þ

S

Now let us divide the volume s into two parts with volume s1 and s2 which are enclosed by surfaces S1 and S2, respectively, as shown in Fig. 1.29b. We see that in Fig. 1.29b at the common internal surface for S1 and S2 the outward drawn normals of the two parts point in opposite direction. Hence the contribution of the internal common surface to the flux of the two parts will cancel each other. I The flux emerging out of surface S1 ¼

A ds1

ð1:89Þ

A ds2

ð1:90Þ

S1

I The flux emerging out of surface S2 ¼ S2

42

1 Vector Analysis

Because the flux due to the common internal surface for S1 and S2 cancel out each other, the flux from the rest of the surfaces of S1 and S2 will be equal to the flux from the original surface S. Thus I I I A ds1 þ A ds2 ð1:91Þ / ¼ A ds ¼ S

S1

S2

Suppose we divide the volume s into a large number of volumes s1 ; s2 . . .sn which are enclosed by the surfaces S1, S2, S3, … Sn, respectively, then I I I I A ds1 þ A ds2 þ þ A : dsn ð1:92Þ / ¼ A ds ¼ S

S1

S2

Sn

I A ds ¼

/¼

N I X j¼1

S

A dsj

ð1:93Þ

Sj

The above equation can be written as H

I A ds ¼

N X

sj

j¼1

S

A dsj

Sj

sj

ð1:94Þ

The volume sj is inﬁnitely small if N is very large, i.e. if N ! 1; sj ! 0. In this limit using Eqs. 1.48, equation 1.94 can be written as I A ds ¼

N X

Dsj r A

ð1:95Þ

j¼1

S

We write D sj instead of sj in the above equation because volume sj is inﬁnitely small. Converting the summation symbol into integration and expressing D sj as ds we can write the above equation as I

Z A ds ¼

S

r Ads

ð1:96Þ

s

Hence divergence theorem is proved. Divergence theorem is an important identity in vector analysis. We will use the theorem to establish other theorems in electromagnetics. Now we will see the physical interpretation of divergence theorem. In Fig. 1.23a we have shown a pipe ﬁtted in a floor and water spreads out from the floor. We used

1.16

Divergence Theorem

43

Fig. 1.30 Pipes ﬁxed inside a cubical structure

this example to explain that whenever there is divergence it implicitly means that there is a source present. Let us restate the above example in a slightly different manner. Instead of ﬁxing the pipe in the floor, let us ﬁx two pipes inside a cubical structure which will allow water to pass through (see Fig. 1.30). Assuming that water is incompressible whatever water comes out of the pipes flows out through the six faces of the cube. Now let us generalize the above example to any given vector. Consider Fig. 1.31 (Fig. 1.31 can be of any irregular shape). There are n numbers of sources giving rise to divergence. The n numbers of sources are contained in volume s which is H bounded by surface S. In Sect. 1.12 we saw that A ds is the flux of the vector. Just as in Fig. 1.30 whatever water comes out of the pipe flows out of the surfaces in Fig. 1.31 whatever diverges from the of the

R cube, similarly H n number of sources ¼ s r Ads flows out of surface in the form of flux ½¼ A ds. H R Hence s r Ads ¼ S A ds.

Fig. 1.31 Vectors diverging inside a irregular structure

44

1 Vector Analysis

Example 1.10 Verify divergence theorem for the vector ^ A ¼ x^i þ y^j þ zk for a unit cube which is located at the origin. Solution Z

I r Ads ¼

s

First let us evaluate the integral

A ds S

R s

r Ads

@ ^@ @ ^ ^ ^ þj þk rA¼ i x^i þ y^j þ zk @x @y @z ¼ 1þ1þ1þ ¼ 3 Now Z

Z1 Z1 Z1 r Ads ¼

s

3dx dy dz x¼0 y¼0 z¼0

¼ 3½ x10 ½ y10 ½z10 ¼ 3 Now let us evaluate the surface integral. The cube has six faces. So, the surface integral has to be evaluated at all the six faces as shown in Fig. 1.32.

Fig. 1.32 .

1.16

Divergence Theorem

45

(i) For face ABCD, x = 0 ds = –dy dz ^i as outward drawn normal to ABCD is along x-axis ZZ

¼ A dy dzð^iÞ ABCD

RR ABCD x¼0

^ dy dzð^iÞ x^i þ y^j þ zk

x¼0

(ii) For the face PQRS, x = 1 and ds = dy dz ^i . Hence RR

ZZ PQRS

A dy dz ^i ¼

^

^

^

PQRS x i þ y j þ z k x¼ 1 R R

x¼1

PQRS x¼1

dy dz ^i

x dy dz ¼ 1

(iii) For the face SADP, y = 0 and ds = –dx dz ^j as the outward drown normal to SADP is along y-axis RR ¼

SADP y¼0

A dx dzð^jÞ ¼

(iv) For the face RBCQ, y = 1 and ds = dx dz ^j ZZ

^ RBCQ A ðdx dzÞj y¼1 ZZ ^i þ y ^j þ z k ^ ðdx dzÞ^j ¼ x RBCQ ¼

ZZ ¼

y¼1 RBCQ y dx d z ¼ 1 y¼1

46

1 Vector Analysis

^ as the outward drawn normal (v) For the face PQCD, z = 0 and ds = dx dy k is along z-axis RR PQCD z¼0

¼

^ A ðdx dyÞðkÞ ¼

^ (vi) For the face SRBA, z = 1 and ds = dx dy k RR SRBA z¼1

¼

^ A dx dy k ¼

Thus I

Z A ds ¼

Z A ds þ

ABCD

Z

þ

A ds PQRS

Z

A ds þ SADP

Z A ds þ

RBCQ

Z A ds þ

PQCD

A ds ¼ 3 SRBA

R Hence divergence theorem is veriﬁed. We evaluated two integrals s r Ads H R and A ds. To evaluate s r Ads we did single integration. However to evalH uate A ds we performed six integrations. r A and volume are scalar quantiR ties. In s r Ads thus only scalar quantities are involved which doesn’t have direction.H Hence single integration is sufﬁcient to arrive at result. However in the integral A ds surface is a vector quantity which involves direction. For the cube shown in Fig. 1.31 there are six different faces or surfaces each having its own direction.H Hence six different integrations for six different directions. Thus evaluR ation of A ds becomes lengthy and laborious. s r Ads is easy as it involves H single integration because the integration involves scalars. A ds is lengthy as it involves vectors. The above example apart from verifying divergence theorem proves the advantage of working with scalars as compared to working with vectors. Scalars are easy to work with as compared to vectors, as scalars don’t involve direction.

1.17

1.17

Stoke’s Theorem

47

Stoke’s Theorem

Stoke’s theorem is helpful in transforming a surface integral of a vector into a line integral around the boundary C, where C is the boundary of the respective surface. The theorem states that Z

I ðr AÞ ds ¼

S

A dl

ð1:97Þ

C

Here C is the boundary of the surface S. ds; dl are the small elements of S and C, respectively. Both line and surface integrals involve vectors. But however as will see in Example 1.11 line integrals are easy to evaluate as compared to surface integrals. Now let us H see the proof for Stoke’s theorem. Consider C A dl where C is the boundary of surface S as shown in Fig. 1.33a in the region of vector A. Let us divide the surface S in two parts S1 and S2 having boundaries C1 and C2, respectively, as shown in Fig. 1.33b. of the H The line integral H vector A over the boundaries C1 and C2 can be written as C1 A dl1 and C2 A dl2 . From Fig. 1.33b as we traverse the line PQ along boundary C1 and line PQ along boundary C2, the line integral of A is equal and opposite and cancels out each other, the remaining boundaries being same as that of C. Hence I I I A dl ¼ A dl1 þ A dl2 ð1:98Þ C

C1

C2

In Fig. 1.33c we divide the surface S into number of surface s1, s2, s3, … sN whose boundaries are C1, C2, C3, … CN. A small part of Fig. 1.33c is shown in Fig. 1.34. Clearly at the common interface the line integral of A cancels out. Using the same argument as we did for Fig. 1.33b to arrive at Eq. 1.98, we can write for Fig. 1.33c I

I A dl ¼

C

I A dl1 +

C1

I A d l2 þ þ

C2

A dlN CN

ð1:99Þ

48

1 Vector Analysis

Fig. 1.33 a Surface S with boundary C. b Surface S divided into two equal parts. c Surface S divided into N number of parts Fig. 1.34 .

I )

A dl ¼

N I X j¼1

C

A dlj

ð1:100Þ

Cj

The above equation can be written as H

I A dl ¼

N X j¼1

C

sj

A dlj

Cj

sj

ð1:101Þ

Suppose if N is very large then the surface sj becomes very small, i.e. if N ! 1; sj ! 0 so that we can write Eq. 1.68 H lim

Dsj !0

A dlj

Cj

Dsj

^j ¼ curl A n

^j is the unit vector normal to surface Dsj . where n

ð1:102Þ

1.17

Stoke’s Theorem

49

In the limit N ! 1; sj ! 0 Eq. 1.101 can be written as I A dl ¼

H

N X

Dsj

j¼1

C

Cj

A dlj Dsj

ð1:103Þ

Here we have used Dsj instead of sj to note that sj ! 0 in the limit . Using Eq. 1.102 in Eq. 1.103 I A dl ¼

N X

^j Dsj curl A n

ð1:104Þ

j¼1

C

^j by n ^ Replacing the summation by integration and Dsj by ds, n I

Z ^ds ðr AÞ n

A dl ¼ C

ð1:105Þ

S

I )

Z A dl ¼

C

ðr AÞ ds

ð1:106Þ

S

Hence Stoke’s theorem is proved. The physical interpretation of Stoke’s theorem is as follows. We know that r A is the measure of how much the vector A circulates R about. As shown in Fig. 1.33c there is large circulation over surface S. Hence S ðr AÞ ds is the measure of total circulation over the surface S. However as shown in Fig. 1.34 at the common interface the line integral of A cancels out. So in Fig. 1.33c at all common interfaces the circulations are in opposite direction and cancel out. H Hence ﬁnally we arrive at Fig. 1.33a. In Fig. 1.33a we calculate the line integral C A dl R which is equivalent to S ðr AÞ ds as the above argument shows. Thus to calculate the curl over the surface S in Fig. 1.33c R you can go over the entire surface S and calculate total curl using the relation S ðr AÞ ds (or) go over the entire boundary C and H calculate the circulation along the boundary C in Fig. 1.33a using the relation C A dl. However while integrating the line integral which way to go around either clockwise or anticlockwise, along the boundary C of Fig. 1.33a. The direction over which the boundary C is to be traversed is ﬁxed by the direction of ds in Fig. 1.33. Suppose assume that ds in Fig. 1.33 is outward perpendicular to the plane of the paper. Let the thumb of your right hand point along ds, i.e. along the outward perpendicular to the plane of the paper. Then the ﬁngers of right hand will be along the arrow marks along the boundary shown in Fig. 1.33a which is the direction in H which the integration of line integral C A dl is to be carried out. Example 1.11

50

1 Vector Analysis

Verify Stoke’s theorem for the surface shown in Fig. 1.35 if A ¼ 4^i þ z2 y3^j. The square surface is of unit dimension. Solution First let us calculate the line integral (a) For the line a shown in Fig. 1.35 x ¼ 0; y ¼ 0 ! 1; z ¼ 0 Hence A ¼ 4^i Z A dl ¼ 4dx )

Z A dl ¼

4dx ¼ 0 as dx ¼ 0

(b) For the line b shown in Fig. 1.35 x ¼ 0; y ¼ 1; z ¼ 0 ! 1 Hence A ¼ 4^i þ z2^j Z

Z 4d x þ z2 dy ¼ 0 as dx; dy ¼ 0

A dl ¼

(c) For the line c shown in Fig. 1.35 x ¼ 0; y ¼ 1 ! 0; z ¼ 1 Hence A ¼ 4^i þ y3^j A dl ¼ 4dx þ y3 dy ¼ y3 dy as dx ¼ 0 0 Z0 y4 1 A dl ¼ y3 dy ¼ ¼ 4 1 4 1

(d) For the line d shown in Fig. 1.35 x ¼ 0; y ¼ 0; z ¼ 1 ! 0 A ¼ 4^i Fig. 1.35 A square surface ABCD in Cartesian coordinate system

1.17

Stoke’s Theorem

51

Z A dl ¼ 0 as dx ¼ 0 H Hence the total sum A dl ¼ 0 þ 0 14 þ 0 ¼ 14 We have r A ¼ 2zy3^i By right-hand rule ds = dy dz ^i R1 R1 R Thus S ðr AÞ ds ¼ 0 0 ð2zy3 Þdydz ¼ 14 Hence Stoke’s theorem is veriﬁed. In Fig. 1.35 there is only one surface ABCD to work with. Suppose assume that ABCD in Fig. 1.35 is an elastic membrane. We have redrawn ABCD elastic membrane in Fig. 1.36a. Let us stretch the elastic membrane in the form of a cube as shown in Fig. 1.36b. As ABCD the elastic membrane is stretched into a cube, face ABCD is now open. Thus there are ﬁve surfaces of cube to work with. Hence for the given loop ABCD in Fig. 1.36a there is only one surface, while in Fig. 1.36b there are ﬁve different surfaces. For the same loop ABCD if the elastic membrane in Fig. 1.36a is blown up in the form of balloon as shown in Fig. 1.36c surface integration becomes even more complicated. In all the three ﬁgures the loop ABCD is two dimensional, just one need z- and y-axis to describe loop ABCD. But the corresponding surface associated with ABCD can be three dimensional as in Fig. 1.36b, c and hence surface integration might become difﬁcult.

Fig. 1.36a, b a Elastic membrane ABCD. b Elastic membrane stretched in the form of cube

52

1 Vector Analysis

Fig. 1.36c Elastic membrane blown in the form of balloon

1.18

The Gradient Theorem

The Gradient theorem states that Z

b

ru d l ¼ uðbÞ uðaÞ

ð1:107Þ

a

where u is a scalar function, and a and b are end points as shown in Fig. 1.37. Because the integration depends on the end points a and b alone and not on the path, the integration will yield identical results for any path between a and b. For example the integration will yield same results for path 1 and path 2 in Fig. 1.37. For a closed path a = b and hence I ru d l ¼ u(a) u(a) ¼ 0 ð1:108Þ Fields which satisfy Eqs. 1.108 and 1.107 (which is independent of the path selected between points a and b in Fig. 1.37) are called conservative ﬁelds.

1.19

Others Coordinate Systems

53

Fig. 1.37 End points a, b along with Path 1 and Path 2 in Cartesian coordinate system

1.19

Others Coordinate Systems

In two dimensions the most familiar coordinate system is Cartesian coordinate system with x- and y-axis as shown in Fig. 1.38. Another coordinate system in two dimensions is polar coordinate system ðr; hÞ. The Cartesian coordinates and polar coordinates are related to each other by x ¼ r cos h; y ¼ r sin h The point P is Fig. 1.38 can be represented by either Cartesian coordinates x, y or polar coordinates ðr; hÞ. If we are interested in calculating electric (or) magnetic ﬁelds at point P due to some charge or current distributions then these physical quantities can be expressed in terms of Cartesian coordinates (x, y) or polar coordinates ðr; hÞ. Similarly in three dimensions in addition to Cartesian coordinates (x, y, z) there are number of other coordinate systems, out of which two are very important spherical coordinate system and cylindrical coordinate system.

Fig. 1.38 Cartesian and polar coordinate system

54

1 Vector Analysis

1.19.1 Spherical Polar Coordinates In Fig. 1.39a we show the Cartesian coordinate system (x, y, z) and the spherical coordinate system ðr; h; /Þ. Here r is the distance from origin. h is the angle between the z-axis and the line drawn from the origin to point P. / is the angle between the x-axis and line OQ. r varies from 0 to 1, h varies from 0 to p and / varies from 0 to 2p. h is allowed to vary up to p and not up to 2p The reason for such limited variation of h in explained in Fig. 1.40. In Fig. 1.40a–d h varies from 0 to p and forms a semicircle. After varying h from 0 to p and forming a semicircle, / is varied in Fig. 1.40e–g. In Fig. 1.40e we vary / from 0 to 90°. In Fig. 1.40f / is varied to 270° and in Fig. 1.40g / completes 2p degrees which ﬁnally forms sphere. Thus variation of from 0 to p forms a semicircle and when the semicircle is rotated through 2p degrees (i.e. we vary /) we ﬁnally form a sphere. To form a hemisphere we vary h from 0 to p=2 as shown in Fig. 1.40h–j. Then we rotate / from 0 to 2p in Fig. 1.40k–m which ﬁnally forms a hemisphere. An elemental length dl can be expressed in spherical polar coordinates as ^ d l ¼ dl1^r þ dl2 ^ h þ dl3 /

ð1:109Þ

Using the result from Fig. 1.39b–d ^ d l ¼ dr^r þ rdh ^ h þ r sin h d //

ð1:110Þ

The volume element ds shown in Fig. 1.39e, Fig. 1.41 can be expressed as ds ¼ (dl1 ) ðdl2 Þ ðdl3 Þ ds ¼ ðdr)(rdhÞðr sin h d/Þ

ð1:111Þ

ds ¼ r2 sin h dr dh d/

ð1:112Þ

Let us calculate the area elements in Fig. 1.41 9 ds1 ¼ dl1 d l3 ¼ r sin h dr d/ > = ds2 ¼ dl2 d l1 ¼ r dr dh > ; ds3 ¼ dl2 d l3 ¼ r2 sin h dh d/

ð1:113Þ

The transformation equation from Cartesian coordinates to spherical coordinates is 9 x ¼ r sin h cos / > = y ¼ r sin h sin / > ; z ¼ r cos h

ð1:114Þ

1.19

Others Coordinate Systems

55

Fig. 1.39a, b, c, d, e Spherical polar coordinates along with line and volume elements

and

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 9 x2 + y2 + z2 > > > pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ > > 2 2 x + y = 1 h ¼ tan > z > > > > 1 y ; / ¼ tan x r¼

ð1:115Þ

56

1 Vector Analysis

Fig. 1.40a, b, c, d Formation of sphere and hemisphere in spherical polar coordinate system

If u is any scalar function and if A is any vector function expressed in spherical coordinates as ^ A ¼ Ar ^r þ Ah ^ h þ A/ / then gradient, divergence and curl can be expressed in spherical coordinates as ru ¼ r A¼

@u [email protected]^ 1 @u ^ ^r þ / hþ @r r @h r sin h @ /

1 @ 2 1 @ 1 @A/ r Ar þ ½sin h Ah þ 2 r @r r sin h @ h r sin h @ /

ð1:116Þ ð1:117Þ

1.19

Others Coordinate Systems

57

Fig. 1.40e, f, g, h, i Formation of sphere and hemisphere in spherical polar coordinate system

58

1 Vector Analysis

Fig. 1.40j, k, l, m Formation of sphere and hemisphere in spherical polar coordinate system

Fig. 1.41 Area element in spherical polar coordinate system

1.19

Others Coordinate Systems

59

1 @ @Ah 1 1 @Ar @ ^r þ ðsin h A/ Þ ðrA/ Þ ^h rA ¼ r sin h @ h r sin h @ / @ r @/ ð1:118Þ 1 @ @Ar ^ ðrAh Þ / þ r @r @h The r2 operator called Laplacian is given by r2 u ¼

1 @ 2 @u 1 @ @u 1 @2u r sin h þ þ r2 @ r @r r2 sin h @ h @h r2 sin2 h @ /2

ð1:119Þ

1.19.2 Cylindrical Coordinates In Fig. 1.42 Cartesian coordinate system along with cylindrical coordinate system ðrC ; /; zÞ is shown. Here rC is the distance from point P to the z-axis, / is the angle between x-axis and line O Q and z is the usual Cartesian coordinate. Here rC varies from 0 to 1, / varies from 0 to 2p and z varies from 1 to 1. An element d l in cylindrical coordinates can be expressed as ^ þ d l3^z d l ¼ d l1^rC þ d l2 /

ð1:120Þ

Using the results from Fig. 1.43 we can write ^ þ d z^z d l ¼ drC^rC þ rC d //

Fig. 1.42 Cartesian and cylindrical coordinate system

ð1:121Þ

60

1 Vector Analysis

The volume element shown in Fig. 1.44 can be written as ds ¼ d l1 d l2 d l3

ð1:122Þ

ds ¼ rC drC d/ dz

ð1:123Þ

Now let us calculate the area element in Fig. 1.44 9 ds1 ¼ dl1 d l3 ¼ drC dz > = ds2 ¼ dl1 d l2 ¼ rC drC d/ > ; ds3 ¼ dl2 d l3 ¼ rC drC dz

ð1:124Þ

The transformation equation from Cartesian to cylindrical coordinate system is given by x ¼ rC cos /; y ¼ rC sin /; z ¼ z

ð1:125Þ

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 9 x2 + y2 > > > = y 1 / ¼ tan x > > > ; z¼z

ð1:126Þ

Also rC ¼

If u is any scalar function and if any vector A is expressed in cylindrical coordinates as ^ þ Az^z A ¼ ArC ^rC þ A/ /

Fig. 1.43 Line elements in cylindrical coordinate system

1.19

Others Coordinate Systems

61

Fig. 1.44 a Volume element in cylindrical coordinate system. b Area element in cylindrical coordinate system

Then the gradient, divergence and curl can be expressed in cylindrical coordinates as ru ¼ r A¼

@u 1 @u ^ @u ^rC þ ^z /þ @ rC rC @ / @z

1 @ 1 @A/ @Az þ ðrC ArC Þ þ rC @ rC rC @ / @z

@A r C @A z 1 @Az @A/ ^ ^rC þ rA ¼ / rC @ / @z @z @rC @ArC 1 @C ^z þ ðrC A/ Þ rC @ r @/

ð1:127Þ ð1:128Þ

ð1:129Þ

62

1 Vector Analysis

The Laplacian r2 is given by r2 u ¼

1 @ @u 1 @2u @2u rC þ þ 2 rC @ rC @ rC rC @ /2 @ z2

ð1:130Þ

Example 1.12 Convert the Cartesian coordinates (3, 4, 5) into spherical coordinates. Solution (a) In spherical coordinates pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ x2 + y2 + z 2 pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ x2 + y2 h ¼ tan1 z 1 y / ¼ tan x pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ r¼

r¼

32 + 42 + 52 ¼ 7:07 pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 2 2 1 3 + 4 h ¼ tan ¼ 45 5 4 / ¼ tan1 ¼ 53:13 3

1.20

Important Vector Identities

Let u be a scalar function A and B be vector functions. The following are few important vector identities. 1. 2. 3. 4. 5. 6. 7. 8. 9.

rðA BÞ ¼ A ðr BÞ þ ðA rÞB þ B ðr AÞ þ ðB rÞA r ðA þ BÞ ¼ r A þ r B r ðr AÞ ¼ 0 r ru ¼ 0 r ðA þ BÞ ¼ r:A þ r:B r ðuAÞ ¼ ur A þ ru A r ðr AÞ ¼ rðr AÞ r2 A r ðA BÞ ¼ B ðr AÞ A ðr BÞ r ðA BÞ ¼ ðB rÞA ðA rÞB þ Aðr BÞ Bðr AÞ

1.20

Important Vector Identities

63

10. r ðuAÞ ¼ ur A þ A ru ^ @ : ^i @ þ ^j @ þ k ^@ 11. r r ¼ ^i @@x þ ^j @@y þ k @z @x @y @z ) r2 ¼ @@x2 þ 2

@2 @ y2

þ

@2 @ z2

is called Laplacian operator.

The above form of Laplacian operator is in Cartesian coordinates. We have seen the form of Laplacian in spherical coordinates in Sect. 1.19a and in cylindrical coordinates in Sect. 1.19b. Example 1.13 ^ Calculate r r if r ¼ ðx2 þ y2 þ z2 Þ1=2 where ^r ¼ x^i þ y^j þ zk Solution @ r^ @ r^ @ r ^ k iþ jþ @x @y @z @ 2 @ 2 ðx þ y2 þ z2 Þ1=2^i þ ðx þ y2 þ z2 Þ1=2^j rr ¼ @x @y @ 2 ^ ðx þ y2 þ z2 Þ1=2 k þ @z 1 ¼ ðx2 þ y2 þ z2 Þ1=2 2x ^i 2 1 þ ðx2 þ y2 þ z2 Þ1=2 ð2yÞ^j 2 1 ^ þ ðx2 þ y2 þ z2 Þ1=2 ð2zÞk 2 ^ ¼ ðx2 þ y2 þ z2 Þ1=2 ðx^i þ y^j þ zkÞ

rr ¼

r r^r ¼ ¼ ¼ ^r r r Example 1.14 ^r Show that r 1r ¼ r r3 ¼ r2

^ where r is the position vector of a point r ¼ x^i þ y^j þ z k

Solution 1 @ 1 ^ @ 1 ^ @ 1 ^ r ¼ iþ jþ k r @x r @y r @z r

64

1 Vector Analysis

with r2 ¼ x2 þ y2 þ z2 ! @ 1 @ 1 ¼ @x r @ x ðx2 þ y2 þ z2 Þ1=2 i @ h 2 ðx þ y2 þ z2 Þ1=2 ¼ @x 1 2x ¼ 2 ðx2 þ y2 þ z2 Þ3=2 x x ¼ 3 ¼ 3 r ðx2 þ y2 þ z2 Þ2 Similarly @ 1 ¼ @y r @ 1 ¼ @y r

y r3 z r3

Thus 1 x y z ^ r ¼ 3 ^i þ 3 ^j þ 3 k r r r r ^ ^ ^ 1 ðxi þ yj þ zkÞ r ¼ 3 ¼ r r r3 r 1 r r^r ^r )r ¼ 3 ¼ 3 ¼ 2 r r r r

1.21

Two and Three Dimensions

Throughout electromagnetics we will be discussing two- and three-dimensional problems. If the reader has difﬁculty in visualizing two and three dimensions go to a corner of the room as shown in Fig. 1.45. In that ﬁgure, O is the corner of the room. Fix O, the corner of the room as the origin. Two walls along with the floor are shown in the same ﬁgure. Take the line joining the walls and floor as x-, y-, z-axis. Floor is the x–y plane. Wall 1 is the x–z plane; wall 2 is the z–y plane. We have shown an ant moving on x–z plane, that is a wall 1. If we want to describe the motion of the ant then we need x- and z-axis alone. No need for y-axis as long as the ant moves in x–z plane. Thus the problem is two dimensional. On the other hand if a bee is flying inside the room as shown in the same ﬁgure then we need all the three x-, y-, z-axis to describe the motion of bee then the problem is three dimensional.

1.21

Two and Three Dimensions

65

Fig. 1.45 Two and three dimensions

Example 1.15 Prove that I

Z ru ds ¼

S

udl C

where u is a scalar. Solution Consider any vector G, from Stoke’s theorem Z I ðr GÞ ds ¼ G dl S

C

66

1 Vector Analysis

Let us assume that G can be written as G = u N where u is a scalar function and N is a constant vector. Then Z I ðr uNÞ:ds ¼ uN dl ð1:131Þ C

From Sect. 1.20 using vector identity 6 r uN ¼ ur N þ ru N But r N ¼ 0 because N is a constant vector. Hence r uN ¼ ru N

ð1:132Þ

Substituting Eq. 1.132 in Eq. 1.131 we get Z I ður NÞ ds ¼ uN dl C

Z

)

I ðN ruÞ ds ¼ C

Z )

uN dl I

ðru dsÞ N ¼

uN dl C

As N is a constant vector, pulling N out of integral and cancelling both sides. Z )

I ru ds ¼

S

u dl C

Hence proved. Example 1.16 Prove that for any two vector Q, R where Q is a constant vector r ðQ RÞ ¼ Qðr RÞ ðQ rÞR Solution From Sect. 1.20 using vector identity 9 we get r ðQ RÞ ¼ ðR rÞQ ðQ rÞR þ Qðr RÞ Rðr QÞ

ð1:133Þ

1.21

Two and Three Dimensions

67

As Q is a constant vector r Q ¼ 0; ðR rÞQ ¼ 0 Hence r ðQ RÞ ¼ Qðr RÞ ðQ:rÞR

ð1:134Þ

Example 1.17 For a constant vector Q and scalar function u prove that I Q rdl

SQ¼ C

^ where S is the surface bounded by loop C and r ¼ x^i þ y^j þ zk. Solution Let u ¼ Q r where Q is a constant vector. From Sect. 1.20 using vector identity 1 ru ¼ rðQ rÞ ¼ Q ðr rÞ þ ðQ rÞr þ r ðr QÞ þ ðr rÞQ Because Q is a constant vector r Q ¼ 0 and ðr rÞQ ¼ 0 Hence ru ¼ rðQ rÞ ¼ Q ðr rÞ þ ðQ rÞr But r r ¼ 0. Hence ru ¼ rðQ rÞ ¼ ðQ rÞr @ @ @ ^ ¼ Qx þ Qy þ Qz x^i þ y^j þ zk @x @y @z ^ ^ ^ ¼ Qx i þ Qy j þ Qz k ¼ Q Subsituting ru ¼ Q and u = (Q rÞ in Eq. 1.133 I Z Q ds ¼ ðQ rÞdl S

C

68

1 Vector Analysis

Because Q is a constant vector I

Z ds ¼

Q S

I

)QS¼

ðQ rÞdl C

ð1:135Þ

ðQ rÞdl C

I )SQ¼

ðQ rÞdl

ð1:136Þ

C

Exercises Problem 1.1 ^ and Q ¼ ^i ^j z k ^ are perCheck whether the two vectors P ¼ 12 ^i 6 ^j þ 3 k pendicular to each other. Problem 1.2 ^ ﬁnd Q if Q is parallel to P. If P ¼ 10^i 4^j þ 2k Problem 1.3 ^ Check whether the vectors P ¼ 2^i þ 4^j þ 18k, ^ and R ¼ 10^i þ 5^j þ 15k ^ are coplanar. Q ¼ 9^i þ 3^j þ 6k Problem 1.4 ^ ﬁnd the unit vector in the direction of P. If P ¼ 3^i þ 2^j þ 5k Problem 1.5 ^ Check whether the three vectors A ¼ 2^i þ 3^j þ 5k, ^ C ¼ 3^i þ 4^j þ 4k ^ form a right angled triangle. and B ¼ ^i þ ^j k, Problem 1.6 A vector P is directed from (3, 4, 5) to (1, 3, 2). Determine jPj and unit vector in the direction of P. Problem 1.7 ^ and Q ¼ 3^i 2^j k ^ using Find the angle between the two vectors P ¼ ^i ^j þ k dot product and cross product. Problem 1.8 ^ Find the value of a such that the three vectors 10^i 5^j þ 5k, ^ and 3^i þ 2^j þ ak ^ are coplanar. 2^i þ 4^j 6k

Exercises

69

Problem 1.9 ^ on vector Q ¼ 2^i þ ^j 3k ^ Find the projection of vector P ¼ ^i þ 3^j þ 2k Problem 1.10 If P + Q + R = 0 then show that P Q = Q R = R P. Problem 1.11 Calculate (P Q) ∙ (R S) Problem 1.12 Show that r log r ¼ rr2 where r is the position vector. Problem 1.13 If r is the position vector show that (i) (ii) (iii) (iv)

rrn ¼ rnn2 r div r = 3 div rn r ¼ ðn + 3)rn curlðrn rÞ ¼ 0

Problem 1.14 Calculate r u at (1, 5; 3) if u ¼ x2 z þ y3 z2 . Problem 1.15 Show that r ðP þ QÞ ¼ r P þ r Q Problem 1.16 ^ is solenoidal. Calculate a if A ¼ ð3x þ yÞ^i þ ðz xÞ^j þ ðay þ zÞk Problem 1.17 ^ ﬁnd If P ¼ 3x2^i y2 z^j þ 2x3 zk (a) r P (b) r ðr PÞ Problem 1.18 2^ ^ ^ If P ¼ ð2x2 + R yÞi 10yzj þ 12xz k Calculate P dr along the straight line from (0, 0, 0) to (1, 1, 0) and then to (1, 1, 1). Problem 1.19 Calculate the gradients of (a) uðr; h; /Þ ¼ 3r sin h 4/ + 1 (b) uðr; h; zÞ ¼ 3cos / rz Problem 1.20 Calculate div P if ^ (a) P ¼ 3^r þ r sin h ^ h þ r/

70

1 Vector Analysis

Fig. 1.46 A rectangular area in Cartesian coordinates (Problem 1.24)

^ þ 3^z (b) P ¼ r^r þ z cos / / Problem 1.21 ^ emerging Calculate the net flux of the vector ﬁeld Pðx; y; zÞ ¼ 3xy2^i þ z2^j þ y3 k from a cube of dimensions 0 x; y; z 1 Problem 1.22 Solve Problem 1.21 using Gauss’s divergence theorem. Problem 1.23 ^ for the cube Verity divergence theorem for the vector P ¼ x3^i þ y3^j þ z3 k 0 x; y; z 1 shown in Fig. 1.25. Problem 1.24 ^ for a Verify Stoke’s theorem for the vector P ¼ ðx + yÞ^i þ ðy + zÞ^j þ ðx + zÞk plane rectangular area with vertices (0, 0), (2, 0), (2, 1), (0, 1) as shown in Fig. 1.46. Problem 1.25 Find the value of a, b and c if ^ is irrotational: P ¼ ð4x þ 5y þ a zÞ^i þ ð3x þ b y þ zÞ^j þ ðc x þ 2y 3zÞ k Problem 1.26 Calculate the surface area and volume of a sphere by integrating surface and volume elements in spherical polar coordinates. Problem 1.27 Calculate the surface area and volume of a cylinder by integrating surface and volume elements in cylindrical coordinates. Problem 1.28 R H Prove that for any vector G r G ds ¼ G ds. s

S

Chapter 2

Electric Charges at Rest: Part I

In this chapter we will learn how to calculate electric ﬁeld for static charge distributions. Greeks were the ﬁrst to observe electrical phenomena. They observed that a piece of amber when rubbed with fur or cat skin acquired the property of attracting small bits of paper and straw. Later on, Coulomb French Scientist conducted experiments using a torsion balance set-up to study forces between charged objects. In the following sections we will describe Coulomb’s law, discuss Gauss’s law and then see how potential is used to calculate electric ﬁeld. For complete grasp of the above topics reader is advised to read Sects. 2.1–2.18 continuously without breaking any section or example in between. After reading up to Sect. 2.18 the reader will be able to appreciate why there are three different methods, Coulomb’s law, Gauss’s law and potential to calculate electric ﬁeld for a given charge distribution.

2.1

Coulomb’s Law

Charles-Augustin de Coulomb, French Scientist, using his torsion balance set-up performed a series of experiments to study the nature of forces between two charged objects. The results obtained by Coulomb using his torsion balance set-up are called Coulomb’s law. As per Coulomb’s law the force between two charges (i) is directly proportional to the product of their charges. (ii) is inversely proportional to the square of the distance between the two charges and acts along the line joining the two charges. (iii) depends on the nature of the medium in which the charged objects are situated.

© Springer Nature Singapore Pte Ltd. 2020 S. Balaji, Electromagnetics Made Easy, https://doi.org/10.1007/978-981-15-2658-9_2

71

72

2

Electric Charges at Rest: Part I

Fig. 2.1 A source charge q separated from a test charge Q

A source charge q is separated from the test charge Q by a distance r as shown in Fig. 2.1. The force exerted by q on Q as per Coulomb’s law is then F¼k

qQ ^r r2

ð2:1Þ

Here ^r is the unit vector in the direction of F. k is the proportionality constant whose value depends on the medium in which the charges are situated, and units in which force, charge and distance are expressed. Because in Fig. 2.1 the charges are like charges, the forces are repulsive in nature. If the charges are unlike charges, the forces are attractive in nature. In SI units force is measured in Newton, distance in metre and charge in Coulomb. The proportionality constant k is written as k¼

1 4peo

So that Eq. 2.1 is F¼

1 qQ ^r 4peo r2

ð2:2Þ

Here eo is the absolute permittivity of free space whose value is given by eo ¼ 8:85 1012 C2 N1 m2 If the charges are not situated in vacuum but in a medium of permittivity e, then e ¼ eo er . The force between the two charges is 1 qQ ^r 4pe r2

ð2:3Þ

1 qQ ^r 4peo er r2

ð2:4Þ

F¼ F¼

Here er is the dielectric constant of the medium and is a dimensionless quantity. For air and vacuum er ¼ 1:

2.2 Electric Field Intensity

2.2

73

Electric Field Intensity

The electric ﬁeld intensity E is deﬁned as the force exerted on a unit positive charge placed at that point in an electrostatic ﬁeld. Here electrostatic ﬁeld refers to the space surrounding an electrostatic charge where the electrostatic forces can be detected. The electric ﬁeld produced by charge q in Fig. 2.1 is F Q

ð2:5Þ

1 q ^r 4peo r2

ð2:6Þ

E¼ Substituting Eq. 2.2 in 2.5 E¼

Consider Fig. 2.2. Point P is situated at a distance of r from the origin, and charge q is situated at a distance of r0 from origin. The distance between q and point P is r: The electric ﬁeld at point P is then EðrÞ ¼ where

1 q ^r 4peo r2

r ¼ r r0 .

Fig. 2.2 Charge q in Cartesian coordinate system

ð2:7Þ

74

2

2.3

Electric Charges at Rest: Part I

Electric Field Intensity Due to a Group of Discrete Point Charges

Consider a group of discrete point charges q1, q2, q3…qn as shown in Fig. 2.3 exerting force on test charge Q. We are interested in calculating the force exerted by the charges q1, q2, q3…qn on the test charge Q. The net force acting on charge Q is F ¼ F1 þ F2 þ F3 þ þ Fn F¼

n X

Fi

ð2:8Þ ð2:9Þ

i¼1

where the force Fi in the above equation is given by Eq. 2.3 replacing q by qi, r by

ri and ^r by ^ri .

Thus the net force acting on the test charge Q is found by calculating the force due to individual charges separately and then summing all the individual forces. However the addition is vector addition and not a scalar addition. That is forces in Eq. 2.8 must be vectorically added (see Sect. 1.1). Now consider Fig. 2.4. We are interested in calculating the net electric ﬁeld at point P. From Eq. 2.9 n F X Fi ¼ Q Q i¼1

)E¼

n X

Ei

ð2:10Þ

ð2:11Þ

i¼1

Once again the sum in Eq. 2.11 is vector addition and not scalar addition. Fig. 2.3 Figure showing number of source charges and a test charge

Fig. 2.4 Figure showing number of source charges

2.3 Electric Field Intensity Due to a Group of Discrete Point Charges

75

With the aid of Eqs. 2.7, 2.11 can be written as E¼

2.4

n 1 X qi ^r i 4peo i¼1 r2i

ð2:12Þ

Continuous Charge Distributions

In the previous section we deﬁned the electric ﬁeld intensity due to a group of discrete point charges. Now let us consider charges that are spread over space. There are three different continuous charge distributions as shown in Fig. 2.5. (1) Line Charge: Charge is distributed over a line as shown in Fig. 2.5a. Mathematically we describe the line charge with symbol k

k¼

dq dl0

ð2:13Þ

where k is charge per unit length.

Fig. 2.5 Continuous charge distributions, a. Line charge distribution, b. Surface charge distribution, c. Volume charge distribution

76

2

Electric Charges at Rest: Part I

(2) Surface Charge: Charge is distributed over a surface as shown in Fig. 2.5b. Mathematically we describe the surface charge with symbol r r¼

dq ds0

ð2:14Þ

where r is charge per unit area. (3) Volume Charge: Charge is distributed over a volume as shown in Fig. 2.5c. Mathematically we describe volume charge with symbol q q¼

dq ds0

ð2:15Þ

where q is charge per unit volume. For a small charge dq Eq. 2.7 can be written as Edq ¼

1 dq ^r 4peo r2

ð2:16Þ

(i) For a line charge dq ¼ kdl0 . Substituting for dq in Eq. 2.16 and integrating over the entire line l0 E¼

1 4peo

Z l0

kdl0 ^r r2

ð2:17Þ

(ii) For a surface charge dq ¼ rds0 . Substituting for dq in Eq. 2.16 and integrating over the entire surface S0 E¼

1 4peo

Z S

0

rds0 ^r r2

ð2:18Þ

(iii) For a volume charge dq ¼ q ds0 . Substituting for dq in Eq. 2.16 and integrating over the entire volume s0 1 E¼ 4peo

Z s0

qds0 ^r r2

ð2:19Þ

Equations 2.17, 2.18 and 2.19 give the electric ﬁeld intensity at point P corresponding to the charge distributions shown in Fig. 2.5a–c, respectively.

2.4 Continuous Charge Distributions

77

Finally the total charge enclosed for the line charge is Z qtot ¼ kdl0

ð2:20Þ

l0

surface charge is Z qtot ¼

rds0

ð2:21Þ

q ds0

ð2:22Þ

S0

and volume charge is Z qtot ¼ s0

In Fig. 2.5 we have used primed coordinates dl0 ; ds0 ; ds0 instead of dl; ds; ds. The above difference has been made in order to distinguish between source points where charge is located and ﬁeld points where the ﬁeld is calculated. For example in Fig. 2.5c within the volume all the points will be denoted with prime coordinates, i.e. ðx0 ; y0 ; z0 Þ because within the volume s0 all the charge is located. However for points like P in Fig. 2.5c we use unprimed coordinates (x, y, z) to denote that it is the point where ﬁeld is calculated. Thus we use primed symbols for source points and unprimed symbols for ﬁeld points. One must be careful enough to integrate with respect to either source points or ﬁeld points wherever necessary.

2.5

A Note about Coulomb’s Law

Equations (2.12), (2.17), (2.18) and (2.19) calculate electric ﬁeld vector E directly in terms of vector. The reader can observe that ^r is present in the equations mentioned above signifying that E is directly calculated in terms of a vector. We have already seen in Sect. 1.1 and Example 1.10 that vectors are very difﬁcult to deal with. Hence, calculation of E using the above-mentioned equations becomes complicated even for simple charge distributions. To circumvent this problem much easier methods to calculate E have been developed—Gauss’s law and potential formulation. As we will see, in Gauss’s law we use symmetry of the problem to calculate E easily, while in potential formulation we reduce the vector problem into a scalar problem so that E is calculated easily. This point will be more clear to the reader when the reader completes Sect. 2.18.

78

2

2.6

Electric Charges at Rest: Part I

Calculating Electric Field E Using Coulomb’s Law

Initially we will provide a note on calculating electric ﬁeld E in two dimensions in Cartesian coordinate system. The points discussed here are valid in three dimensions and other coordinate systems. The electric ﬁeld E in two dimensions is given by E ¼ Ex^i þ Ey^j While solving problems using Coulomb’s law it will be sufﬁcient to calculate Ex and Ey for the given charge distribution and substitute the values in the above equation to get E. However in calculating E for certain charge distributions like the ones solved in Examples 2.4, 2.5 and 2.7 either Ex or Ey might cancel out each other and the remaining component alone is to be calculated. But for certain charge distributions as shown in Example 2.9, Sect. 2.14, neither component will cancel out and hence Ex and Ey need to be calculated separately. Whether the components will cancel out (or) get added up depends on the charge distribution and the point where the electric ﬁeld is calculated. The above conclusions hold good for three dimensions and other coordinate systems. Example 2.1 Two charges of magnitude 5 10−9 C and 7 10−9 C are separated by 25 cm in vacuum. Find the force of interaction between them. If the same charges are separated by the same distance in mica (er = 6), what is the force of interaction? Solution We know that F¼ F¼

1 qQ 4peo r2

1 ð5 109 Þð7 109 Þ 4peo ð25 102 Þ2

1 1 ¼ 9 109 . ¼ 4peo 4pð8:85 1012 Þ ð5 109 Þð7 105 Þ Hence F ¼ ð9 109 Þ ð25Þ2

But

F ¼ 0:504 105 N For mica er ¼ 6

2.6 Calculating Electric Field E Using Coulomb’s Law

Hence Fmica ¼ F er ¼

79

0:504 105 6 Fmica ¼ 8:4 107 N:

Example 2.2 A line charge k ¼ 5ðy0 Þ4 nC b/m is situated along y-axis from 0 to 7 cm. Calculate the total charge. Solution From Eq. 2.20 Z qtot ¼ Z7 ) qtot ¼

kdl0 "

ðy0 Þ5 5ðy Þ dy ¼ 5 5 0 4

#7

0

0

0

) qtot ¼ 75 nC ) qtot ¼ 16:807 l C Example 2.3 In a circular disc of radius R charge density is given by r ¼ kðr0 Þ2 . Calculate the total charge. Solution From Eq. 2.21 Z qtot ¼

rds0

From Fig. 2.6 ds0 ¼ r0 dr0 d/0

Fig. 2.6 Charge distributed over a circular disc

80

2

Electric Charges at Rest: Part I

Fig. 2.7 Two point charges q separated by distance 2a

Hence qtot ¼

RR R2p

kðr0 Þ2 r0 dr0 d/0

0 0

) qtot ¼ 2pk ) qtot ¼

R4 4

pkR4 2

Example 2.4 Two charges each having magnitude q are separated by a distance 2a as shown in Fig. 2.7. Find the net electric ﬁeld at point P. Solution Let each charge q produce an electric ﬁeld of Eq at point P as shown in Fig. 2.8a. The magnitude of electric ﬁeld Eq produced by q at point P is Eq ¼ 1 q 4peo PB2 The line CPD shown in Fig. 2.8a is shown separately in Fig. 2.8b. Resolving Eq into Eqx and Eqy in Fig. 2.8b we see that the x components Eqx cancel out each other while Eqy components add up. Hence, the net electric ﬁeld at point P by the two charges each of magnitude q is given by E ¼ 2Eqy^j From Fig. 2.8c cosðp=2 hÞ ¼

Eqy Eq

) Eqy ¼ Eq sin h

2.6 Calculating Electric Field E Using Coulomb’s Law

81

Fig. 2.8 Electric ﬁeld produced by both the charges q and resolution of electric ﬁeld into components is shown in the ﬁgure

Substituting for Eqy in E E ¼ 2Eq sin h ^j From Fig. 2.8a sin h ¼

AP t ¼ pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 2 PB t þ a2

Substituting for sin h in E t E ¼ 2 Eq pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ^j t2 þ a2 Substituting for Eq E¼2

1 q t pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ^j 4peo PB2 t2 þ a2

82

2

E¼

1 qt ^j 4peo ½t2 þ a2 3=2

as

Electric Charges at Rest: Part I

PB2 ¼ t2 þ a2

Example 2.5 A line charge is present in the region a y0 a with a uniform line charge density k as shown in Fig. 2.9. Calculate the electric ﬁeld at point P at a distance t from point O, where O is the centre of the line charge. Solution In order to calculate the electric ﬁeld produced by the entire line charge let us divide the line charge into number of small strips each of length dl0 . Let us consider two Fig. 2.9 A line charge in Cartesian coordinate system

Fig. 2.10 The line charge divided into number of small strips

2.6 Calculating Electric Field E Using Coulomb’s Law

83

strips dy0 each holding a charge kdy0 as shown in Fig. 2.10. Let us calculate the electric ﬁeld by these two strips initially and ﬁnally integrate between 0 and a in order to calculate the net electric ﬁeld produced by the line charge lying between –a and a. The magnitude of electric ﬁeld produced by single strip kdy0 at point P as shown in Fig. 2.10 is jEs j ¼ jE s j ¼

1 kdy0 4peo r2

1 kdy0 h i 4peo ðy0 Þ2 þ t2

Resolving the electric ﬁeld Es due to the two strips into Esx and Esy components as shown in Fig. 2.10 we see that Esy components cancel out each other while Esx components add up. Let us denote the net electric ﬁeld at point P due to the two strips as dE and dE is given by dE ¼ 2 Esx^i From Fig. 2.10 we can write cos h ¼

Esx t ¼ r Es

t ) Esx ¼ Es r t Esx ¼ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Es ðy0 Þ2 þ t2 Substituting the value of Esx in dE t dE ¼ 2 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Es^i ðy0 Þ2 þ t2 Substituting the value of Es in dE t 1 kdy0 dE ¼ 2 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ h i2 ^i 4peo 2 2 2 2 0 0 ðy Þ þ t ðy Þ þ t dE ¼

1 k t dy0 h i3=2 ^i 2peo ðy0 Þ2 þ t2

84

2

Electric Charges at Rest: Part I

This is the ﬁeld produced by two line elements kdy0 at point P. The ﬁeld produced by the entire line charge at point P is Za E¼ 0

¼

kt dE ¼ 2peo

Za h 0

2

dy0 ðy0 Þ2 3a

þ t2

i3=2 ^i

0

kt 6 y 7 4 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ5 ^i 4peo 2 2 t ðy0 Þ þ t2 0

k a pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ^i E¼ 2peo t a2 þ t2 Example 2.6 Two point charges q1 = 10 nC and q2 = 20 nC are located at (1, 1, 3) m and at (2, 3, 2) m, respectively. Calculate the magnitude of force on q1 by q2. Solution The distance between the two charges is r¼

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ pﬃﬃﬃ ð2 1Þ2 þ ð3 1Þ2 þ ð2 3Þ3 ¼ 6 m

The force exerted by q2 onq1 is given by F¼

1 ð10 109 Þð20 109 Þ ^r pﬃﬃﬃ 4peo ð 6Þ 2

Here ^r is the unit vector pointing from q2 to q1 F ¼ 300 109 ^r N In order to ﬁnd ^r let us calculate the vector

r pointing from q2 to q1

r ¼ ð2 1Þ^i þ ð3 1Þ^j þ ð2 3Þk^ r ¼ ^i þ 2^j k^ and jr j ¼

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ pﬃﬃﬃ 1þ4þ1 ¼ 6 ^r ¼

^ r ^i þ p 2^j k ﬃﬃﬃ ¼ jr j 6

2.6 Calculating Electric Field E Using Coulomb’s Law

85

300 ^ N F ¼ pﬃﬃﬃ 109 ð^i þ 2^j kÞ 6 Now jFj ¼

300 109 pﬃﬃﬃ pﬃﬃﬃ 6 ¼ 300 109 N. 6

Example 2.7 Charge is uniformly distributed on a circular disc of radius R. Calculate the electric ﬁeld at point P at distance y from the centre of the disc. Charge density r on the surface of the disc is uniform. Solution In order to calculate the electric ﬁeld produced by the entire surface charge r divide the surface of the entire disc into number of small surface elements ds0 . Let us consider two surface elements ds0 each holding a charge rds0 as shown in Fig. 2.11a, b. Let us calculate the electric ﬁeld due to these two surface elements initially and ﬁnally integrate between 0 and p, 0 to R in order to calculate net electric ﬁeld produced by the entire disc. The electric ﬁeld produced by the single surface element ds0 at point P is jEe j ¼

1 rds0 h i 4peo y2 þ ðr0 Þ2

Fig. 2.11 a Chargeuniformly distributed over a circular disc. b Resolution of electric ﬁeld vectors due to two surface elements

86

2

Electric Charges at Rest: Part I

Resolve jEe j into Eex and Eey components as shown in Fig. 2.11b. Clearly Eex components cancel out, while Eey components add up. Let dE be the electric ﬁeld produced by the two elements at point P. Then as per the above discussion dE ¼ 2 Eey ^j From Fig. 2.11b cos h ¼

Eey y ¼ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Ee y2 þ ðr0 Þ2

y ) Eey ¼ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Ee y2 þ ðr0 Þ2 Substituting the value of Eey in dE y dE ¼ 2 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Ee y2 þ ðr0 Þ2 Substituting for Ee in dE y 1 rds0 h i ^j dE ¼ 2 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 2 4peo y2 þ ðr 0 Þ2 2 0 y þ ðr Þ dE ¼

ry ds0 ^j 2peo ½y2 þ ðr0 Þ3=2

Now the ﬁeld produced by the entire disc is then Z E ¼ dE Z ry ds0 ¼ h i3=2^j 2peo 2 2 0 y þ ðr Þ ry E¼ 2peo

Zp ZR 0

0

r0 dr0 d/0 ^ h i3=2 j y2 þ ðr0 Þ2

" # r y ^j E¼ 1 2 2eo y þ R2

2.6 Calculating Electric Field E Using Coulomb’s Law

87

Example 2.8 A 1 nC charge is situated at (2, 0, 0) m, and 2 nC charge is situated at (0, 1, 0) m. Find the electric ﬁeld at point P whose coordinates are (0, 0, 4) m. Solution The position vector of 1 nC charge with respect to point P is

r1 ¼ 2^i þ 4k^ The position vector of 2 nC charge with respect to point P is

r1 ¼ 1^j þ 4k^ The electric ﬁeld of 1 nC charge at point P is E1 ¼ Now jr 1 j ¼

1 1 109 ^r 1 4peo r21

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ pﬃﬃﬃﬃﬃ 4 þ 16 ¼ 20 ^r1 ¼

Hence, E1 ¼

^ r1 2p^i þ 4k ﬃﬃﬃﬃﬃ ¼ jr 1 j 20

^ 1 1 109 ð2^i þ 4kÞ pﬃﬃﬃﬃﬃ 4pe0 20 20

^ ) E1 ¼ 0:20125 ^i þ 0:4025 k Now jr 2 j ¼ ^r1 ¼

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ pﬃﬃﬃﬃﬃ 1 þ 16 ¼ 17 ^ r2 ^p j þ 4k ﬃﬃﬃﬃﬃ ¼ jr 2 j 17

) E2 ¼ ð9 109 Þ

^ ð2 109 Þ ð^j þ 4kÞ pﬃﬃﬃﬃﬃ ð17Þ 17

^ E2 ¼ 0:2568 ^j þ 1:027 k The net electric ﬁeld at point P is ^ þ ð0:2568 ^j þ 1:027 kÞ ^ E ¼ E1 þ E2 ¼ ð0:20125 ^i þ 0:4025 kÞ ^ E ¼ 0:20125 ^i 0:2568 ^j þ 1:4295 k

88

2

Electric Charges at Rest: Part I

Example 2.9 Three charges of magnitude q1, q2 and q3 are situated at the three corners of a square of side b. Calculate the net electric ﬁeld produced at the fourth corner of the square. Solution Consider Fig. 2.12. As shown the electric ﬁeld produced by q1 is acting along xaxis and is given by E1 ¼ jE1 j^i The electric ﬁeld produced by q3 is acting along y-axis and is given by E3 ¼ jE3 j^j The electric ﬁeld produced by q2 is along the diagonal and is denoted as E2. The net electric ﬁeld at point O is given by E ¼ E1 þ E2 þ E3 To ﬁnd E we need to resolve E2 into its x and y components as shown in Fig. 2.12b E2 ¼ E2x^i þ E2y^j

Fig. 2.12 a Threecharges located at the corner of a square. b Resolution of electric ﬁeld due to the charges

2.6 Calculating Electric Field E Using Coulomb’s Law

89

Thus E ¼ E1^i þ E2x^i þ E2y^j þ E3^j

E ¼ ðE1 þ E2x Þ^i þ E2y þ E3 ^j Example 2.10 Three charges shown in Fig. 2.13 are in equilibrium. Find the value of a. Solution For the charges to be in equilibrium ð27Þð12Þ ð27Þð102Þ ð12Þð102Þ ¼ ¼ a2 20 ð20 aÞ2 ) a2 ¼

ð27Þð12Þð20Þ ð27Þð102Þ

) a ¼ 1:5339 mm

2.7

Solid Angle

In the lower classes the reader would have learnt about ordinary angle which is angle in two dimensions. The ordinary angle h is shown in Fig. 2.14a. It is the angle between two lines. In Fig. 2.14a it is the angle between lines OA and OB. When h is in radians h¼

l r

where l is the arc length and OA = OB = r in Fig. 2.14a. As we increase OA, OB in Fig. 2.14b to OA1, OB1, h remains the same. l l1 That is h ¼ ¼ ½OA1 ¼ OB1 ¼ r1 . r r1 Fig. 2.13 Three charges separated by speciﬁed distance

90

2

Electric Charges at Rest: Part I

Fig. 2.14 Angle in two dimensions

As we increase l it ﬁnally forms a circle as shown in Fig. 2.14c. For a circle l = 2pr. l 2pr ¼ 2p h¼ ¼ r r Similarly for a semicircle l ¼ pr l pr h¼ ¼ ¼p r r Thus the ordinary angle is the angle in two dimensions. Solid angle is the angle in three dimensions. Take a plain sheet of paper, and rotate it in the form of cone as shown in Fig. 2.15. The angle subtended is the solid angle. The solid angle is denoted by X. Now consider Fig. 2.16a. We have shown solid circle X along with x-, y-, z-axes. Now in Fig. 2.16a AB is area. The solid angle is given by X¼

Area AB r2

From Fig. 2.16a as r is increased to r1 in Fig. 2.16b still X remains constant X¼

Area AB Area A1 B1 ¼ ½OA1 ¼ OB1 ¼ r1 r2 r21

2.7 Solid Angle

91

Fig. 2.15 Solid Angle—Angle in three dimensions

Fig. 2.16 Solid angle along with Cartesian coordinate system

As the area AB is increased and closed ﬁnally, we get a sphere as shown in Fig. 2.16c. For a sphere the surface area is 4pr2. Hence for a sphere solid angle is X¼

Area AB 4pr2 ¼ 2 ¼ 4p r2 r

For a hemisphere surface area is 2pr2. Thus solid angle for hemisphere is X¼

2pr2 ¼ 2p r2

92

2

Electric Charges at Rest: Part I

For a sphere we have seen that the solid angle is 4p. Although we have derived for the case of sphere the result holds good for any closed surface. That is, the solid angle subtended by any closed surface is 4p. Now consider Fig. 2.17. Let area BC be S. From Fig. 2.17 it is clear that AB is the projection of BC. Hence from Sect. 1.12 we can write Area AB ¼ Area BC cos a

ð2:23Þ

) Area AB ¼ S cos a

ð2:24Þ

Substituting 2.24 in solid angle X X¼

Fig. 2.17 Area S along with its projection AB is shown in the ﬁgure

Area AB S cos a ¼ r2 r2

ð2:25Þ

2.8 Gauss’s Law

2.8

93

Gauss’s Law

In this section we will prove Gauss’s law and later we will see how Gauss’s law simpliﬁes the calculation of electric ﬁeld as compared to Coulomb’s law. Consider an arbitrary surface S which is enclosinga charge q as shown in Fig. 2.18a. Let E be the electric ﬁeld produced by q as shown in Fig. 2.18a. ^r is the ^ is the unit unit vector in the direction of E. Consider a small strip of area ds, and n ^ and ^r is a. vector normal to ds. The angle between n The flux passing through the small surfaceds is ¼ E ds

ð2:26Þ

The total flux passing through the closed surface S is I ¼

E ds

ð2:27Þ

S

(Reader should note that we are using unprimed symbols instead of primed symbols because we are dealing with ﬁeld points). q ^r, ds ¼ ds n ^. But E ¼ 4peo r2 Substituting the above equations in Eq. 2.27 I

I E ds ¼

q 1 ^r n ^ ds 4peo r2

^ are unit vectors, the dot product ^r n ^ ¼ cos a. Thus Because ^r; n q Total flux ¼ 4peo

I

ds cos a r2

ð2:28Þ

Fig. 2.18 aArbitrary surface S enclosing charge q. b Charge q and small strip of area ds.

94

2

Electric Charges at Rest: Part I

dX

ð2:29Þ

But from Fig. 2.18b and using Eq. 2.25 Total flux ¼

q 4peo

I

H where dX is the solid angle as shown in Fig. 2.18b. But dX ¼ 4p because the solid angle subtended by any closed surface is 4p. Hence, I E ds ¼

q q 4p ¼ 4peo eo

ð2:30Þ

Suppose the charge q is situated outside the surface S as shown in Fig. 2.19. Consider two small strips ds1 and ds2. At surface ds1 the electric ﬁeld is pointing ^1 , the unit vector normal to ds1. Hence flux at ds1 is (using Eq. 2.30) opposite to n ¼

qdX 4pe0

(Here we need not integrate because we are considering only small strip ds1 and not the entire surface S). ^2 ; hence the flux at ds2 is At ds2 the E is pointing in the same direction of n ¼

qdX 4pe0

Hence the net flux through the elements ds1 and ds2 is ¼

qdX qdX þ ¼0 4peo 4peo

The above fact is true for any cone drawn through the surface S in Fig. 2.19. Thus if the charge is lying outside the surface the net flux enclosed by the surface due to the charge is zero.

Fig. 2.19 Charge q situated outside the surface S

2.8 Gauss’s Law

95

In Fig. 2.18 if the charge is situated in a medium of dielectric constant e1 then the flux is I q E ds ¼ ð2:31Þ eo er In Fig. 2.18 instead of one charge if the surface S encloses n number of charges q1, q2, q3, …qn, then told flux enclosed by the surface S is Total flux ¼ ðfluxÞ1 þ ðfluxÞ2 þ þ ðfluxÞn I q q q E ds ¼ 1 þ 2 þ þ n eo eo eo Here E is the “net” electric ﬁeld due to all charges I q þ q2 þ q3 þ þ qn E ds ¼ 1 eo

ð2:32Þ ð2:33Þ

ð2:34Þ

S

I E ds ¼ S

qenc eo

ð2:35Þ

where q1 + q2 + … + qn = qenc is the total charge enclosed by the surface S and (Flux)1, (Flux)2, … (Flux)n are the fluxes through the surface S due to charges q1, q2 … qn. From Eq. 2.22 I qenc ¼ q ds ð2:36Þ s

assuming that a volume charge density q is enclosed in surface S in Fig. 2.18. s is the volume enclosed by surface S in Fig. 2.18. In Eq. 2.36 we use unprimed symbol instead of primed symbol because we are dealing with ﬁeld points. With aid of Gauss’s divergence theorem I

Z E ds ¼ S

r Eds

ð2:37Þ

s

Substituting Eqs. 2.37 and 2.36 in Eq. 2.35 R

Z r E ds ¼ s

s

q ds eo

ð2:38Þ

96

2

Electric Charges at Rest: Part I

The above integration considers a small volume ds and then integrates over the entire volume s. Equation 2.38 is valid even for the small volume ds. In that case we need not integrate. Hence, r Eds ¼

q ds eo

ð2:39Þ

q eo

ð2:40Þ

)rE¼

The above form of Gauss’s law is called differential or point form of Gauss’s law.

2.9

Sketches of Field Lines

In Chap. 1 we discussed vectors in detail and the concept of flux was introduced in Sect. 1.12. Flux is deﬁned as the number of lines passing through the given area. Or in other words we have used “flux lines” to deﬁne flux. In Fig. 1.20 we have shown lines of the vector passing through the area abcd which are the flux lines. We have read about Coulomb’s law from which we observe that the ﬁeld of the point charge falls off as 1=r2 . Also Eq. 2.40 tells us the fact that divergence of the electric ﬁeld produced by a static charge distribution is nonzero. Recall that divergence means spreading out of vectors. Equation 2.40 says that the electric ﬁeld of a static charge distribution spreads out. For example we sketch the electric ﬁeld of a point charge in Fig. 2.20a and see that electric ﬁeld spreads out. In Fig. 2.20a we have sketched the electric ﬁeld in terms of vectors. Because the ﬁeld falls off as 1=r2 as per Coulomb’s law, the length of the vectors becomes shorter as we move away from the charge. Alternatively we can sketch the electric ﬁeld using “flux lines” or “ﬁeld lines” as shown in Fig. 2.20b. The flux lines or ﬁeld lines are

Fig. 2.20 aElectric ﬁeld of a point charge. b Field lines of a point charge

2.9 Sketches of Field Lines

97

Fig. 2.21 a Fieldlines of a positive and negative charge. b Field lines of two negative charges

directed lines. At each point surrounding the charge the direction of the line indicates the direction of the vector. The magnitude of the electric ﬁeld is indicated by the “density” of the ﬁeld lines. The density of the lines is stronger near the charge indicating that the ﬁeld at those points is stronger. Far away from the charge the density of the lines is less indicating the ﬁeld is weaker at those points. The electric ﬁeld pattern due to a positive and negative charge and two negative charges has been represented in Fig. 2.21 with the help of ﬁeld line concept. The ﬁeld lines have no real existence, but however as shown in Fig. 2.21 the ﬁeld lines are helpful in visualizing the ﬁelds. The following points should be noted about the ﬁeld lines. (a) Field lines cannot cross each other—otherwise ﬁeld would have two different directions at that point. (b) Field lines start on positive charge and end on negative charge as shown in Fig. 2.21a—otherwise, they extend to inﬁnity as shown in Fig. 2.21b. (c) Assuming a sphere of radius r the centre of which contains a point charge q the ﬁeld lines emanate outwards perpendicularly and uniformly through the surface of the sphere. As the surface area of the sphere is 4pr2 the density of the lines— that is the flux lines per unit area is inversely proportional to r2.

98

2

2.10

Electric Charges at Rest: Part I

Curl of E

Consider a point charge q situated at origin as shown in Fig. 2.22. The electric ﬁeld at point P is given by E¼ E¼

q 1 ^r 4pe0 r2

q 1 r 4pe0 r2 jrj

ð2:41Þ

From Fig. 2.22 ^ r ¼ x ^i þ y ^j þ z k j rj ¼

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ x2 þ y2 þ z 2

ð2:42Þ ð2:43Þ

Substituting Eqs. 2.43 and 2.42 in 2.41 E¼ )E¼

^ q x ^i þ y ^j þ z k 4peo ½x2 þ y2 þ z2 3=2

q x ^i 2 2 4peo ½x þ y þ z2 3=2 q y ^j þ 4peo ½x2 þ y2 þ z2 3=2 q z ^ þ k 4peo ½x2 þ y2 þ z2 3=2

^ ) E ¼ Ex^i þ Ey^j þ Ez k

Fig. 2.22 Point charge q situated in origin

ð2:44Þ

ð2:45Þ

ð2:46Þ

2.10

Curl of E

99

where Ex ¼

q x 4peo ½x2 þ y2 þ z2 3=2

ð2:47Þ

Ey ¼

q y 4peo ½x2 þ y2 þ z2 3=2

ð2:48Þ

Ez ¼

q z 4peo ½x2 þ y2 þ z2 3=2

ð2:49Þ

We know that ^ ^j ^ i k @ @ @ rE¼ @x @y @z Ex Ey Ez @Ez @Ey ^ @Ex @Ez ^ @Ey @Ex ^ ¼ k iþ jþ @y @z @z @x @x @y

ð2:50Þ

Let us calculate

@Ez @Ey @y @z " # @Ez q @ z ¼ 4peo @y ðx2 þ y2 þ z2 Þ3=2 @y

ð2:51Þ

" # @Ez 3q zy ¼ 4peo ðx2 þ y2 þ z2 Þ5=2 @y

ð2:52Þ

" # @Ey q @ y ¼ 4peo @z ðx2 þ y2 þ z2 Þ3=2 @z " # 3q yz ¼ 4peo ðx2 þ y2 þ z2 Þ5=2

ð2:53Þ

100

2

Electric Charges at Rest: Part I

Now " # " # @Ez @Ey 3q zy 3q yz þ ¼ 4peo ðx2 þ y2 þ z2 Þ5=2 4peo ðx2 þ y2 þ z2 Þ5=2 @y @z @Ez @Ey ) ¼0 @y @z

ð2:54Þ

@Ex @Ez ¼0 @z @x

ð2:55Þ

@Ey @Ex ¼0 @x @y

ð2:56Þ

Similarly

Substituting Eqs. 2.54, 2.55 and 2.56 in Eq. 2.50 rE¼0

ð2:57Þ

Thus curl of E is zero. Suppose there are a number of charges each producing the ﬁeld E1, E3, …En, then from Eq. 2.11 E ¼ E1 þ E2 þ þ En

ð2:58Þ

With the help of Eq. 2.57 we see that 9 r E1 ¼ 0 > > = r E2 ¼ 0 > > ; r En ¼ 0

ð2:59Þ

for individual charges. Thus for the net electric ﬁeld E due to all charges given in Eq. 2.58 r E ¼ r E1 þ r E2 þ þ r En

ð2:60Þ

From Eq. 2.59 then rE¼0

ð2:61Þ

The above equation is true for any static chargedistribution. Recall that curl is a measure of circulation or rotation. r E ¼ 0 says that electric ﬁelds due to static charges do not circulate about. Electric ﬁelds due to static charges are irrotational ﬁelds.

2.11

2.11

Potential of Discrete and Continuous Charge Distributions

101

Potential of Discrete and Continuous Charge Distributions

In this section we will derive expression for potential. In the later sections we will see how potential simpliﬁes the calculation of electric ﬁeld as compared to Coulomb’s law. From Eq. 2.61 rE¼0

ð2:62Þ

From Example 1.9, curl gradient of any scalar is zero. Hence, r ðrVÞ ¼ 0

ð2:63Þ

Here V is a scalar quantity and is known as electric potential. The inclusion of the negative sign in the above equation will be explained later. Comparing Eqs. 2.62 and 2.63 E ¼ rV

ð2:64Þ

Equation 2.64 can be written as ZP

ZP E dl ¼

R

rV dl R

Applying gradient theorem (Eq. 1.107) ZP

E dl ¼ VðPÞ VðRÞ R

where R the reference point is the “zero of the potential”, i.e. where we ﬁx the potential to be zero, and P is the point where the potential V is calculated. We will learn more about point R in future sections. For the time being V(R) = 0. Hence ZP VðPÞ ¼

E dl R

ð2:65Þ

102

2

Electric Charges at Rest: Part I

Let us ﬁx the “zero of the potential” at ∞. ZP E dl

V¼

ð2:66Þ

1

Consider Fig. 2.23. The potential at ∞ is zero. Charge q is located at origin. The point P lies at a distance r from charge q. In Fig. 2.23 E¼

1 q ^r 4peo r2

ð2:67Þ

From Fig. 2.23 dl ¼ dr ^r

ð2:68Þ

From Eqs. 2.67 and 2.68 q 1 ^r ðdr ^rÞ E dl ¼ 4peo r2 q 1 E dl ¼ dr 4peo r2

ð2:69Þ ð2:70Þ

Substituting Eq. 2.70 in Eq. 2.66 Zr V¼ 1

q 1 dr 4peo r2

q 1 r V¼ 4peo r 1 V¼

q 1 4peo r

ð2:71Þ

ð2:72Þ ð2:73Þ

The negative sign in Eq. 2.63 was selected to make the potential of the positive charge come out positive. Equation 2.73 holds good for Fig. 2.23 where the point charge q lies at origin. In general the point charge q need not lie in origin as shown in Fig. 2.24. For such situations the potential is given by

2.11

Potential of Discrete and Continuous Charge Distributions

103

Fig. 2.23 Point charge located at the origin

Fig. 2.24 Point charge q in Cartesian coordinate system

V¼

q 1 4peo r

ð2:74Þ

If there is a collection of n number of point charges as shown in Fig. 2.4, then the total potential at point P can be calculated as follows. From Eq. 2.11 the net electric ﬁeld at point P in Fig. 2.4 is E ¼ E1 þ E2 þ þ En

ð2:75Þ

Integrating ZP

ZP E dl ¼

1

ZP E1 dl1 þ

1

ZP E2 dl2 þ þ

1

En dln

ð2:76Þ

1

) V ¼ V1 þ V2 þ þ Vn

ð2:77Þ

Using Eq. 2.74 we get V¼

1 q1 1 q2 1 qn þ þ þ 4peo r1 4peo r2 4peo rn

ð2:78Þ

104

2

V¼

Electric Charges at Rest: Part I

n 1 X qn 4peo i¼1 rn

ð2:79Þ

Thus when there are a collection of point charges in Fig. 2.4 the potential at point P in Fig. 2.4 is given by Eqs. 2.77–2.79. Here r1 ; r2 ; r3 . . .rn are the distances of charges q1, q2, q3, … qn up to point P, respectively, in Fig. 2.4. For continuous charge distributions shown in Fig. 2.5a–c the potential is: For line charge in Fig. 2.5a V¼

1 4peo

Z l0

kdl0 r

ð2:80Þ

rds0 r

ð2:81Þ

q ds0 r

ð2:82Þ

For surface charge in Fig. 2.5b 1 V¼ 4peo

Z S

0

For volume charge in Fig. 2.5c V¼

1 4peo

Z s0

For a point charge situated at origin in Fig. 2.22 E from Eq. 2.46 is ^ E ¼ Ex ^i þ Ey ^j þ Ez k

ð2:83Þ

where the components Ex, Ey, Ez are given by Eqs. 2.47–2.49. For a point charge situated in origin in Fig. 2.23 the potential is given by Eq. 2.73 V¼

1 q 1 q pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ¼ 4peo r 4peo x2 þ y2 þ z2

As we see there are no components for V because V is a scalar. Vectors alone have components. Scalars don’t have components. This point has been discussed in detail in Sect. 1.9. Electric ﬁeld Eis a vector and hence can be resolved into components as shown inFig. 2.25a. However thepotential at point P in Fig. 2.25b is V and does not have components. That is along x-, y-, z-axes the potential is V, not Vx, Vy, Vz. To make this point more clear consider an object of mass m in Fig. 2.25c. The mass of the object along x-, y-, z-axes is m, not mx, my, mz, because mass is a scalar quantity. If the mass of the object is 50 kg, it will be always 50 kg along x-, y-, z-axes.

2.11

Potential of Discrete and Continuous Charge Distributions

105

Fig. 2.25 aResolution of electric ﬁeld into components. b Charge q and potential V in Cartesian coordinate system. c Mass m in Cartesian coordinate system

Similarly the potential V is a scalar and hence along x-, y-, z-axes its value is V and not Vx, Vy, Vz. However the electric ﬁeld E is a vector, and hence it can be split up into three components Ex, Ey and Ez. Hence if you are using Coulomb’s law directly, to calculate electric ﬁeld then you have to work with three components Ex, Ey and Ez. Hence you have to fuss with direction. However if you are working with potential, you are working with only one component V alone and need not bother about direction because along x-, y-, zdirections the value is V alone. This makes working with potential V easy as compared to working with E directly. Once V has been found, then E can be calculated using the relation E ¼ rV The way in which potential simpliﬁes calculation of electric ﬁeld as compared to Coulomb’s law will be more clear to the reader when the reader goes through Sect. 2.18. A ﬁnal point. The potentials calculated in Eqs. 2.80, 2.81 and 2.82 will tend to inﬁnity if the charge distribution itself extends to inﬁnity. In that case we select a suitable different reference point.

2.12

Calculating Electric Field Using Gauss’s Law and Potential

We know that Gauss’s law is I E ds ¼

qenc eo

Our aim is to calculate E using the above equation. However E is inside the integral. If we want to calculate E using the above equation, then E must be taken out of the integral. For this, for a given charge distribution we construct an imaginary Gaussian surface such that the following conditions are satisﬁed.

106

2

Electric Charges at Rest: Part I

(1) On the Gaussian surface E ds ¼ E ds cos h can be easily evaluated. (2) HOn the Gaussian surface E cos h is constant so that in the integral, E ds; E cos h can be pulled out. That is I

I E ds ¼

E cos h ds

ð2:84Þ

Because E cos h is constant over Gaussian surface I

I E cos h ds ¼ E cos h

ds

ð2:85Þ

H Now, ds, the calculation of area of the Gaussian surface is simple mathematics. Thus from Gauss’s law I E ds ¼

qenc eo

I )

E ds cos h ¼

qenc eo

Because E cos h is constant over Gaussian surface I q E cos h ds ¼ enc eo )E¼

qenc 1 H eo cos h ds

ð2:86Þ

ð2:87Þ ð2:88Þ

Hence from the above equation E can be easily calculated. If for a given charge distribution, a Gaussian surface cannot be constructed such that Eds cannot be easily evaluated (or) if E cos h is not constant over Gaussian surface, then E cannot be pulled out of the integral. In such situations calculation of E using Gauss’s law becomes difﬁcult. Then the alternative method to calculate E is by calculating potential initially and then determining E using the relation E ¼ rV In the following sections we will calculate the electric ﬁeld by using Coulomb’s law, Gauss’s law and potential for different charge distributions and compare the results and methods.

2.13

2.13

Electric Field Due to an Inﬁnite Line Charge

107

Electric Field Due to an Inﬁnite Line Charge

Consider an inﬁnite uniform line charge present along y-axis as shown in Fig. 2.26. The charge per unit length of the line charge is k. Let us calculate the electric ﬁeld at point P which is situated at a distance of r from point O as shown in Fig. 2.26. Now let us calculate the electric ﬁeld at point P using Coulomb’s law, Gauss’s law and potential formulation. (a) Using Coulomb’s Law Let us divide the entire line charge into small strips dy0 and consider only two strips as shown in Fig. 2.26. Each strip holds a charge of k dy0 . Let us initially calculate the electric ﬁeld due to these two strips. Finally we integrate to calculate the net electric ﬁeld at point P due to the entire line charge. Resolving the electric ﬁeld Es due to each strip into x and y components Esx and Esy we see that Esy components cancel out each other while Esx components add up. Let the net electric ﬁeld at point P due to the two strips be dE. The dE is given by dE ¼ 2 Esx ^i

ð2:89Þ

Esx Es

ð2:90Þ

From Fig. 2.26 cos h ¼

Fig. 2.26 A inﬁnite line charge situated along y axis. We denote the coordinates as y' to note that the points are source points

108

2

Electric Charges at Rest: Part I

Esx ¼ Es cos h

ð2:91Þ

dE ¼ 2 Es cos h ^i

ð2:92Þ

Substituting 2.91 in 2.89

The electric ﬁeld Es produced by the single element is given by Es ¼

1 kdy0 4peo r2

ð2:93Þ

From Fig. 2.26 y0 ¼ r tan h ) dy0 ¼ r sec2 h dh

ð2:94Þ

and r¼

r cos h

ð2:95Þ

Substituting Eqs. 2.95 and 2.94 in 2.93 Es ¼

1 k r sec2 hd h 2 cos h 4peo r2

ð2:96Þ

1 k dh 4peo r

ð2:97Þ

Es ¼ Substituting Eq. 2.97 in 2.92 dE ¼ 2:

1 k dh cos h ^i 4peo r

ð2:98Þ

The ﬁeld produced by the entire line charge at point P is 1 k E¼ 2peo r

Zp=2

cos h dh ^i

ð2:99Þ

0

In the above integration we are integrating from point O to ∞, only half of the entire line charge, because we are considering strips on two sides. For the point O the angle h is zero. For the inﬁnite point, as y ! ∞ the angle h ! p=2. Hence

2.13

Electric Field Due to an Inﬁnite Line Charge

E¼

109

1 k p=2 ½sin h0 ^i 2peo r

ð2:100Þ

k ^ i 2peo r

ð2:101Þ

E¼ (b) Using Gauss’s Law

In Fig. 2.27 we show the inﬁnite line charge. We have to calculate ﬁeld due to inﬁnite Line Charge by applying Gauss’s law. The ﬁrst step is to get “the feel of the ﬁeld”. In Fig. 2.26 we see that only x components survive, while y components cancel out. Hence the ﬁeld of the inﬁnite line charge “looks” radial as shown in Fig. 2.27a. Next step is to construct a Gaussian surface, keeping in mind that evaluation of E ds must be easy and E cos h must be constant over the Gaussian surface. A little thought shows that the cylindrical Gaussian surface satisﬁes the above requirement. So we construct the cylindrical Gaussian surface as shown in Fig. 2.27b. The Gaussian surface is an imaginary surface. Let r and L be the radius and length of the cylindrical Gaussian surface. Applying Gauss’s law to the cylindrical Gaussian surface I E ds ¼

qenc eo

Fig. 2.27 a Electric ﬁeld lines of inﬁnite line charge. b Cylindrical Gaussian surface for inﬁnite line charge

110

2

Z

Z

)

Z

E ds þ Endface 1

Electric Charges at Rest: Part I

E ds þ Curvedface

E ds ¼ Endface 2

kL eo

ð2:102Þ

where qenc = kL is the total charge enclosed by the Gaussian surface. End face 1, 2 and curved face of the cylindrical Gaussian surface are shown in Fig. 2.27b. At the end faces the angle between E and ds is 90° and at the curved surface the angle between E and ds is zero degrees Hence Eq. 2.102 is Z

E ds cos 90 þ

Endface1

Z

Z

E ds cos 0 þ

Curvedface

Endface2

E ds cos 90 ¼

kL eo

ð2:103Þ

As cos 90° = 0, cos 0° = 1 Eq. 2.103 is Z E ds ¼ Curvedface

kL eo

ð2:104Þ

From Fig. 2.27b it is clear that E is constant over the curved surface and hence Z ds ¼

E Curvedface

Now

R

kL eo

ð2:105Þ

ds is the area of the curved face which is equal to 2prL. Hence

Curvedface

) Eð2prLÞ ¼ E¼

k 2preo

kL eo

ð2:106Þ ð2:107Þ

While calculating electric ﬁeld using Coulomb’s law the Ey components cancelled out as shown in Fig. 2.26 and hence the difﬁculty involved in calculating electric ﬁeld using Coulomb’s law is less obvious in the above example. However future examples will show that Gauss’s law eases out the calculation of electric ﬁeld as compared to Coulomb’s law. (c) Using Potential Let us redraw Fig. 2.26 as shown in Fig. 2.28. Let us calculate the potential at point P in Fig. 2.28 due to a small element dy0 which carries a charge of kdy0 . We have

2.13

Electric Field Due to an Inﬁnite Line Charge

111

Fig. 2.28 .

V¼

1 kdy0 4peo r

ð2:108Þ

The total potential due to the entire line charge is 1 V¼ 4peo

Z1

kdy0 r

ð2:109Þ

kdy0 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ðy0 Þ2 þ r2

ð2:110Þ

1

From Fig. 2.28 r2 ¼ ðy0 Þ2 þ r2 . Hence 1 V¼ 4peo

Z1 1

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ1 k 0 ln y þ ðy0 Þ2 þ r2 V¼ 4peo 1

ð2:111Þ

V¼1 As previously mentioned if the charge distribution extends to inﬁnity, then the potential calculated using Eq. 2.80 goes to inﬁnity. However we can calculate the potential difference between two points. We have V¼

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ1 k ln y0 þ ðy0 Þ2 þ r2 4peo 1

ð2:112Þ

112

2

Electric Charges at Rest: Part I

The above equation can be written as qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃt k 0 V ¼ lim ln y þ ðy0 Þ2 þ r2 t!1 4peo t

ð2:113Þ

" pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ !# k t þ r 2 þ t2 pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ln V ¼ lim t!1 4peo t þ r2 þ t2

ð2:114Þ

For t r we have pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ r2 r2 þ t2 t1 þ 2 2t

3 2 2 t 2þ r2 k 2t V ¼ lim ln4 2 2 5 t!1 4peo r =2t

ð2:115Þ

Consider a reference point ro. The potential at the reference point ro is

2 r2 3 t 2 þ o2 k 2t Vo ¼ lim ln4 2 2 5 t!1 4peo r0 =2t

2 3 2 2 2 t 2þ r2 k r =2t 2t V Vo ¼ lim ln4 2 2 0 r2 5 t!1 4peo r =2t t 2 þ 02 2t

2 3 2 2 2þ r2 r 2 t

5 V Vo ¼ lim ln4 o2 r2 t!1 r 2 þ o2 2t

ð2:116Þ

ð2:117Þ

ð2:118Þ

Applying the limit k r2 ln o2 4peo r k r V Vo ¼ ln 2peo ro V Vo ¼

ð2:119Þ ð2:120Þ

2.13

Electric Field Due to an Inﬁnite Line Charge

113

Now the electric ﬁeld is @ ðV V o Þ @r @ k r E¼ ln @r 4peo ro

E ¼ rðV Vo Þ ¼

E¼

k 1 1 2peo r=ro ro

)E¼

k 2peo r

ð2:121Þ ð2:122Þ ð2:123Þ ð2:124Þ

The reason for using potential difference instead of potential to calculate E and the physical signiﬁcance of the reference point will be clear to the reader in future sections.

2.14

Electric Field Due to the Finite Line Charge

(a) Using Coulomb’s Law Consider a uniformly charged wire of length L, the charge per unit length being k. Let us calculate the electric ﬁeld at point P as shown in Fig. 2.29. The coordinates of point P are (x, y). Let us divide the entire line charge into small elements dx0 each carrying a charge of kdx0 . We will calculate the ﬁeld produced by the small element kdx0 and then integrate over the entire length to get the ﬁeld of the entire line charge. The point of intersection of point P on the line charge is at a distance of x from one end of the line charge, and the element dx0 is situated at a distance of x0 from origin O as shown in Fig. 2.29. The electric ﬁeld produced by the element dx0 is dE. Let us resolve dE into x and y components dEx and dEy, respectively. From Fig. 2.29 we can write dEx ¼

1 kdx0 cos h 4peo r2

ð2:125Þ

But from Fig. 2.29 r2 ¼ y2 þ ðx x0 Þ2

ð2:126Þ

x x0 cos h ¼ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ y2 þ ðx x0 Þ2

ð2:127Þ

114

2

Electric Charges at Rest: Part I

Fig. 2.29 Finite line of charge

Substituting Eqs. 2.126 and 2.127 in 2.125 dEx ¼

1 kdx0 ðx x0 Þ h i qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 4peo y2 þ ðx x0 Þ2 y2 þ ðx x0 Þ2

ð2:128Þ

1 kðx x0 Þdx0 h i3=2 4peo 2 y þ ðx x0 Þ2

ð2:129Þ

dEx ¼

The x component of the electric ﬁeld due to the entire line charge of length L is k Ex ¼ 4peo

ZL h 0

2 Ex ¼

ðx x0 Þdx0 y2 þ ðx x0 Þ2

ð2:130Þ

i3=2 3

k 6 1 1 7 4qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ5 2 2 4peo 2 x þ y 2 y þ ðx LÞ

ð2:131Þ

2.14

Electric Field Due to the Finite Line Charge

115

Now let us calculate the y component dEy ¼

1 kdx0 sin h 4peo r2

ð2:132Þ

From Fig. 2.29 y sin h ¼ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ y2 þ ðx x0 Þ2

ð2:133Þ

Substituting Eqs. 2.126 and 2.133 in Eq. 2.132 dEy ¼

1 kdx0 y h i qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 4peo y2 þ ðx x0 Þ2 y2 þ ðx x0 Þ2

ð2:134Þ

1 kydx0 h i3=2 4peo 2 y þ ðx x0 Þ2

ð2:135Þ

dEy ¼

The y component of the electric ﬁeld due to the entire line charge of length L is Ey ¼

ky 4peo

ZL h 0

dx0 y2

þ ðx

2 Ey ¼

x0 Þ 2

ð2:136Þ

i3=2 3

k 6 Lx x 7 4qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ þ pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ2ﬃ5 2þL 4peo y 2 y 2 y þ ðx LÞ

ð2:137Þ

The net electric ﬁeld at point P due to the entire line charge is then given by E ¼ Ex^i þ Ey^j

ð2:138Þ

where Ex and Ey are given by Eqs. 2.131 and 2.137, respectively. (b) Using Gauss’s Law Step one is to plot the electric ﬁeld or “to get a feel of the ﬁeld”. For comparison purposes let us sketch the ﬁeld due to the ﬁnite and inﬁnite line charge. Initially we show a line charge with two elements kdx’ in Fig. 2.30a. At point P the Ey components cancel out each other leaving only Ex components. Now let us consider the inﬁnite line charge as shown in Fig. 2.30b. In this case because the line charge is inﬁnite, for each line element kdx0 there will be always

116

2

Electric Charges at Rest: Part I

Fig. 2.30a Line charge with two line elements.

Fig. 2.30b, c b Electric ﬁeld of a inﬁnite line charge. c Electric ﬁeld of a ﬁnite line charge

opposite line element kdx0 such that Ey components cancel out and Ex components alone add. Hence it is either point P or P1 or P2 or any other point in Fig. 2.30b and the ﬁeld is always radial as shown in Fig. 2.30b because at every point in Fig. 2.30b the Ey components will cancel out and Ex components will add

2.14

Electric Field Due to the Finite Line Charge

117

up. Hence radial components are alone present in Fig. 2.30b. Therefore we were able to construct a cylindrical Gaussian surface in Fig. 2.27b such that the electric ﬁeld Eds can be easily evaluated for the cylindrical Gaussian surface and also E is constant over the cylindrical Gaussian surface. Thus using Gauss’s law we calculated E for inﬁnite line charge in Sect. 2.13b. Now let us consider ﬁnite line charge. In Fig. 2.30c we see that at point P the Ey components cancel out and Ex components are alone present. Thus ﬁeld at point P in Fig. 2.30c is radial. However at points like P1 and P2 the Ey components will not cancel out. Hence the ﬁeld will not be radial at points like P1 and P2 as shown in Fig. 2.30c. Now let us construct a cylindrical Gaussian surface for the ﬁnite line charge as shown in Fig. 2.31. Clearly at point like P, E and ds are in the same direction. But at points like P1, P2, E makes an angle with ds as shown in Fig. 2.31. Thus at edges of the cylindrical Gaussian surface E and ds have different directions. Hence at the edges or at any other point where E and ds are at some angle to each other evaluation of Eds is difﬁcult over the entire Gaussian surface. Even if we somehow manage to calculate Eds, E cos h is not H constantH over the Gaussian surface and hence taking E cos h out of the integral E ds ¼ Eds cos h becomes difﬁcult. It is still true that over the cylindrical Gaussian surface I E ds ¼

qenc eo

H But however evaluation of E ds and taking E cos h out of the integral E ds are difﬁcult. Hence using Gauss’s law to calculate electric ﬁeld in the above case, ﬁnite line charge, becomes tedious. Instead of selecting a cylindrical Gaussian Fig. 2.31 Cylindrical Gaussian surface for ﬁnite line of charge

118

2

Electric Charges at Rest: Part I

surface we can select a Gaussian surface of different shapes. However H until E ds can be easily evaluated and E cos h can be taken out of the integral E ds, easy evaluation of E using Gauss’s law is not possible. Thus we have seen using Gauss’s law evaluation of E for a ﬁnite line charge is very difﬁcult. “The same is true for any charge distribution. For a given charge distribution if we are interested in using Gauss’s law to calculate E, then we must be able to construct a Gaussian surface such that E ds must be evaluated easily andHE cos h must be constant over Gaussian surface so that E cos h can be pulled out of E ds. If we are unable to construct a Gaussian surface which satisﬁes the above requirement, then Gauss’s law cannot be easily applied to calculate E. It is still true H q that E ds ¼ eenc over the Gaussian surface, but Gauss’s law is not useful in o calculating E. (c) Using Potential The potential at point P in Fig. 2.29 due to the element dx0 which carries a charge of kdx0 is dV ¼

1 kdx0 4peo r

ð2:139Þ

The potential at point P due to the entire line charge in Fig. 2.29 is ZL

kdx0 r

ð2:140Þ

dx0 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ y2 þ ðx x0 Þ2

ð2:141Þ

1 V¼ 4peo

0

Substituting Eq. 2.126 in Eq. 2.140 k V¼ 4peo

ZL 0

Integrating k 1 x L 1 x sinh V¼ sinh 4peo y y

ð2:142Þ

Now let us calculate electric ﬁeld components Ex ¼

@V k @ xL x ¼ sinh1 sinh1 @x 4pe0 @x y y

ð2:143Þ

2.14

Electric Field Due to the Finite Line Charge

119

2 Ex ¼

3

k 6 1 1 1 17 6rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 7 ﬃ r ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 5

2 4peo 4 xL 2 y y x þ1 þ1 y y 2

Ex ¼

ð2:144Þ

3

k 6 1 1 7 4qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ5 2 þ y2 4peo 2 x ðx LÞ þ y2

ð2:145Þ

Now the Ey component is Ey ¼

@V k @ xL x ¼ sin h1 sin h1 @y 4peo @y y y 2 3 k 6 1 xL 1 x 7 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Ey ¼ 4qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 5 4peo y2 y x2 ðxLÞ2 þ 1 þ 1 2 y y2 2 Ey ¼

ð2:146Þ

ð2:147Þ

3

k 6 Lx x 7 4qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ þ pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ5 2 2 4peo y 2 x þy ðx LÞ þ y2

ð2:148Þ

Now E is given by E ¼ Ex^i þ Ey^j

ð2:149Þ

where Ex, Ey the components are given by Eqs. 2.145, 2.148.

2.15

Electric Field Along the Axis of a Uniformly Charged Circular Disc

Consider a circular disc of radius R on which charge is uniformly distributed with surface charge density r. Let us calculate the electric ﬁeld at point P at a distance y from the centre of the disc as shown in Fig. 2.32. (a) Using Coulomb’s Law We have calculated the electric ﬁeld for the above problem in Example 2.7 using Coulomb’s law.

120

2

Electric Charges at Rest: Part I

Fig. 2.32 Circular disc carrying uniform surface charge density

(b) Using Gauss’s Law In Fig. 2.33 ﬁeld distribution for a uniformly charged disc at few points is shown. Clearly construction of a Gaussian surface (over which E ds can be easily evaluated) is not that easy. Hence for similar reasons mentioned in Sect. 2.14b Gauss’s law is not useful here in calculating E. (c) Using Potential Let us calculate the electric ﬁeld at point P in Fig. 2.32 using potential. Consider a surface ds0 element situated at a distance of r from point P. The potential due to the surface element at point P is given by V¼

1 4peo

Z S

0

rds0 r

ð2:150Þ

From Fig. 2.32 ds0 ¼ r0 dr0 d/0

ð2:151Þ

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ y2 þ ðr0 Þ2

ð2:152Þ

r¼

Fig. 2.33 Electric ﬁeld distribution at few points for a circular disc carrying a uniform surface charge density

2.15

Electric Field Along the Axis of a Uniformly Charged Circular Disc

121

Substituting Eqs. 2.151 and 2.152 in Eq. 2.150 1 V¼ 4peo

Z2p ZR 0

2pr V¼ 4peo

0

ZR 0

r V¼ 2eo

r r0 dr0 d/0 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ y2 þ ðr0 Þ2

r0 dr0 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ y2 þ ðr0 Þ2

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 2 2 y þR y

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ @ ^@ r 2 2 ^ þj E ¼ rV ¼ i y þR y @x @y 2eo As

ð2:153Þ

ð2:154Þ

ð2:155Þ ð2:156Þ

@V ¼0 @x qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ @ r y2 þ R2 y ^j @y 2eo " # r y 1 pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ^j E¼ 2eo y2 þ R 2

E¼

ð2:157Þ

ð2:158Þ

Note Integration here is carried from 0 to 2p, that is over the entire disc, because we consider only one surface element ds0 . In Example 2.7 we carry out integration from 0 to p for /0 , that is over only half of the disc, because we consider two surface elements.

2.16

Electric Field Due to an Inﬁnite Plane Sheet of Charge

Consider an inﬁnite plane sheet of charge carrying a uniform surface charge density r as shown in Fig. 2.34. We are interested in calculating the electric ﬁeld at point P. We will use cylindrical coordinates (r, /, z). Consider two strips situated opposite to each other. The electric ﬁeld Es due to each strip is shown in the ﬁgure. Resolving the Es into components Esr, Esz we see that Esr components cancel out while Esz components add up.

122

2

Electric Charges at Rest: Part I

Fig. 2.34 Inﬁnite plane sheet carrying a uniform surface charge density

Hence the net ﬁeld due to the two strips at point P is dE ¼ 2Esz ^z

ð2:159Þ

From Fig. 2.34 sin h ¼

Esz Es

) Esz ¼ Es sin h

ð2:160Þ

2.16

Electric Field Due to an Inﬁnite Plane Sheet of Charge

123

The ﬁeld due to single strip ds0 at point P is Es ¼

1 rds0 4peo r2

ð2:161Þ

From Fig. 2.34 sin h ¼

z r

But

r¼

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ z2 þ ðr0 Þ2

ð2:162Þ

Hence z sin h ¼ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ z2 þ ðr0 Þ2

ð2:163Þ

Substituting Eqs. 2.161–2.163 in Eq. 2.159 dE ¼ 2Es sin h^z 1 rds0 sin h ^z 4peo r2

ð2:164Þ

1 rds0 z h i qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ^z 2 4peo z2 þ ðr0 Þ z2 þ ðr0 Þ2

ð2:165Þ

dE ¼ 2 dE ¼ 2

dE ¼

rz h 2peo

ds0 z2 þ ðr0 Þ2

i3=2 ^z

ð2:166Þ

The ﬁeld due to the entire plane sheet of charge is then E¼

rz 2peo

Z S

rz E¼ 2peo

ds0

h i3=2^z z2 þ ðr0 Þ2

0

Zp Z1 h 0

0

r0 dr0 d/0 i3=2 ^z ðr0 Þ2 þ z2

ð2:167Þ

ð2:168Þ

124

2

rz E¼ 2eo

Z1 h 0

r0 dr0 ðr0 Þ2 þ z2

2 E¼

rz 6 4 2eo

Electric Charges at Rest: Part I

i3=2^z

ð2:169Þ

31 1 ðr0 Þ2 þ z2

E¼

7

1=2 5 ^z

ð2:170Þ

0

rz 1 ^z 2eo z

ð2:171Þ

r ^z 2eo

ð2:172Þ

E¼ (a) Using Gauss’s Law

Now let us calculate the electric ﬁeld of an inﬁnite plane sheet of charge using Gauss’s law. Step one is to get feel of the ﬁeld. As shown in Figs. 2.34 and 2.35a the ﬁeld due to a inﬁnite plane sheet of charge points in z direction. Hence as shown in Fig. 2.35b we construct a Gaussian surface in the shape of pillbox with surface area S. Half of the pillbox lies above the plane sheet of charge, and half of the pillbox lies below the plane sheet of charge. Bottom half of the pillbox is not shown in the ﬁgure.

Fig. 2.35 a Electric ﬁeld lines for inﬁnite plane sheet carrying uniform surface charge density. b Gaussian Surface for inﬁnite plane sheet carrying uniform surface charge density

2.16

Electric Field Due to an Inﬁnite Plane Sheet of Charge

125

As per Gauss’s law I E ds ¼

qenc eo

The above integral can be written for the pillbox as Z Z Z E ds þ E ds þ E ds top

bottom

¼

sides

qenc eo

ð2:173Þ

For the sides of two pillboxes as shown in Fig. 2.35b the angle between E and ds is 90°. Hence E ds = 0 for sides of the pillbox. Equation 2.173 is then Z E ds

top

Z þ E ds

¼ bottom

qenc eo

ð2:174Þ

For the top and bottom of pillbox the angle between E and ds is zero. Hence Z jEjjdsj cos 0 S

top

Z þ jEjjdsj cos 0 S

Z jEjjdsj þ S

qenc eo

ð2:175Þ

bottom

Z

)

¼

jEjjdsj ¼ S

qenc eo

ð2:176Þ

Because |E| is constant over the top and bottom faces of the pillbox Z

Z jdsj þ jEj

jEj S

jdsj ¼ S

ESþES ¼

qenc eo

qenc eo

ð2:177Þ

ð2:178Þ

From Fig. 2.35 qenc = rS hence 2ES ¼

rS eo

ð2:179Þ

)E¼

r 2eo

ð2:180Þ

r ^z 2eo

ð2:181Þ

)E¼

126

2

Electric Charges at Rest: Part I

Fig. 2.36 Inﬁnite plane sheet carrying a uniform surface charge density

(b) Using Potential The potential at point P due to entire plane in Fig. 2.36 is 1 V¼ 4peo r V¼ 4peo

Z S0

Z2p Z1 0

0

rds0 r

r0 dr0 d/0 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ðr0 Þ2 þ z2

ð2:182Þ

ð2:183Þ

by substituting Eq. 2.162. r V¼ 2eo

Z1 0

r0 dr0 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ðr0 Þ2 þ z2

ð2:184Þ

2.16

Electric Field Due to an Inﬁnite Plane Sheet of Charge

127

The above equation can be written as Zt

r V ¼ lim t!1 2eo

o

r t!1 2eo

V ¼ lim

r0 dr0 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ðr0 Þ2 þ z2

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃt ðr0 Þ2 þ z2

ð2:185Þ

ð2:186Þ

0

i ﬃ r hpﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ t2 þ z2 z t!1 2eo

V ¼ lim

ð2:187Þ

As t z, pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ z2 t2 þ z2 t 1 þ 2 2t Hence r z2 t 1þ 2 z t!1 2eo 2t

ð2:188Þ

r z2 tþ z t!1 2eo 2t

ð2:189Þ

V ¼ lim

V ¼ lim

Consider a reference point z0. The potential at reference point z0 is r z2 t þ o zo t!1 2eo 2t

Vo ¼ lim

ð2:190Þ

The potential difference between points z and z0 is 2 r z z2o þ zo z V Vo ¼ lim 2eo t!1 2t 2t

ð2:191Þ

Applying the limit V Vo ¼

r ½ zo z 2eo

ð2:192Þ

Now the electric ﬁeld is E ¼ rðV Vo Þ

ð2:193Þ

128

2

Electric Charges at Rest: Part I

In cylindrical coordinates E¼

@ðV Vo Þ 1 @ðV Vo Þ ^ @ðV Vo Þ ^z /þ r^ þ @r r @/ @z E¼

@ðV Vo Þ ^z @z

ð2:194Þ ð2:195Þ

as V–Vo in Eq. 2.192 doesn’t contain terms with r, /. Substituting Eq. 2.192 in 2.195 E¼

r ^z 2eo

ð2:196Þ

The reason for using potential difference instead of potential to calculate electric ﬁeld and the physical signiﬁcance of reference point zo will be clear to the reader in future sections.

2.17

Electric Field of a Uniformly Charged Spherical Shell

Consider a spherical shell of radius R as shown in Fig. 2.37. Charge q is distributed uniformly on the shell with surface charge density r.

Fig. 2.37 a A spherical shell carrying uniform surface charge density. b. Surface element of a spherical shell carrying uniform surface charge density

2.17

Electric Field of a Uniformly Charged Spherical Shell

129

Fig. 2.38 Two opposite surface elements of a spherical shell carrying uniform surface charge density

(a) Using Coulomb’s Law (i) At a point interior to the spherical shell Consider point P in the interior of the spherical shell as shown in Fig. 2.38. Consider two cones with point P as their apex extending up to the surface of the sphere where they cut the sphere. Cone 1 and cone 2 subtend a solid angle of dX1 and dX2, respectively. Both the solid angles are equal dX1 ¼ dX2

ð2:197Þ

)

ds1 ds2 ¼ 2 r21 r2

ð2:198Þ

)

ds2 r22 ¼ ds1 r21

ð2:199Þ

Because the spherical shell is uniformly charged, the charge on each element is proportional to the area. Hence, dq2 ds2 ¼ dq1 ds1

ð2:200Þ

The ratio between electric ﬁelds at point P due to the elements ds1 and ds2 is 1 dq2 E2 4pe0 r22 ¼ 1 dq1 E1 4pe0 r21

ð2:201Þ

130

2

Electric Charges at Rest: Part I

E2 dq2 r21 ¼ E1 dq1 r22

ð2:202Þ

dq2 r22 ¼ dq1 r21

ð2:203Þ

E2 ¼1 E1

ð2:204Þ

From Eqs. 2.199 and 2.200

Hence

Thus the ﬁeld due to the elements ds1 and ds2 at point P is equal in magnitude. From Fig. 2.38 the ﬁeld due to ds1 and ds2 at point P is oppositely directed. Hence the ﬁeld at point P is zero. This is true for any point in the interior of the sphere. (ii) At a point exterior to the spherical shell Consider Fig. 2.37a. Clearly the x, y components of electric ﬁeld at point P due to the symmetrically placed surface elements cancel out and only the z components add up. Thus we have to integrate over the z components, dEs cos a of the elements, and here dEs is ﬁeld produced by the surface element at P given by dEs ¼

1 dq 4peo r2

ð2:205Þ

We have dq ¼ r ds0 ¼ r R2 sin h0 dh0 d/0

ð2:206Þ

in spherical polar coordinates. From Fig. 2.37b r2 ¼ R 2 þ z2 2R z cos h0

ð2:207Þ

(h in Fig. 2.37b is denoted as h0 in Eq. 2.207 to note that it is a source point.) Also, cos a ¼

z R cos h0 r

ð2:208Þ

Now the electric ﬁeld at point P due to the entire shell is Z Ez ¼

dEs cos a

ð2:209Þ

2.17

Electric Field of a Uniformly Charged Spherical Shell

131

Substituting Eq. 2.205 in Eq. 2.209 Z 1 dq cos a Ez ¼ 4peo r2

ð2:210Þ

Substituting Eqs. 2.206, 2.207 and 2.208 in Eq. 2.210 Ez ¼

rR2 4peo

Z2p

d/0

0

Ez ¼

Zp 0

2p r R2 4peo

Zp 0

Let t ¼ cos h0

ðz R cos h0 Þ sin h0 dh0 ðR2 þ z2 2Rz cos h0 Þ

ðz R cos h0 Þ sin h0 dh0 ðR2 þ z2 2Rz cos h0 Þ

ð2:211Þ

ð2:212Þ

dt ¼ sin h0 dh0 h0 ¼ 0 ) t ¼ 1 h0 ¼ p ) t ¼ 1

Hence 2 1 3 Z 2prR2 4 ðz RtÞdt 5 Ez ¼ 4peo ðR2 þ z2 2R ztÞ1=2

ð2:213Þ

1

" #1 2p r R2 1 ðzt RÞ pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Ez ¼ 4peo z2 R2 þ z2 2R zt 1 20 Ez ¼

ð2:214Þ

13

2R 1 6B z R z þ R C7 [email protected]ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ þ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃA5 2 4o z ðz RÞ2 ðz þ RÞ2

ð2:215Þ

2p r R2 2 4p r R2 1 ¼ 2 4peo z 4peo z2

ð2:216Þ

2

Ez ¼

But 4p R2 r ¼ q the total charge enclosed by the spherical surface. Hence Ez ¼

1 q^ k 4peo z2

ð2:217Þ

132

2

Electric Charges at Rest: Part I

Fig. 2.39 Electric ﬁeld lines of a spherical shell carrying uniform surface charge density

Now the total electric ﬁeld at point P due to the entire spherical shell is ^¼ E ¼ Ex^i þ Ey^j þ Ez k

1 q^ k 4peo z2

as Ex and Ey are zero. (b) Using Gauss’s Law First step is to plot the electric ﬁeld of the uniformly charged spherical shell as shown in Fig. 2.39. Clearly if we want to evaluate E ds easily, then we have to construct the Gaussian surface to be a sphere. (i) At a point interior to the spherical shell We construct a Gaussian surface as shown in Fig. 2.40 for an interior point P. Gauss’s law states I q E ds ¼ enc eo From Fig. 2.40 the angle between E and ds is zero and the charge qenc enclosed by the Gaussian surface is zero. Hence from Gauss’s law Z jEjjdsj cos 0 ¼ 0 ð2:218Þ ) jEj ¼ 0 Hence electric ﬁeld at the interior point of the charged spherical shell is zero.

2.17

Electric Field of a Uniformly Charged Spherical Shell

133

Fig. 2.40 Gaussian surface for an interior point of a spherical shell carrying uniform surface charge density

(ii) At a point exterior to the spherical shell We construct a Gaussian surface for an exterior point P as shown in Fig. 2.41. The angle between E and ds is zero, and the total charge enclosed by the spherical shell is q. Applying Gauss’s law I q E ds ¼ enc ð2:219Þ eo

Fig. 2.41 Gaussian surface for an exterior point of a spherical shell carrying uniform surface charge density

134

2

I jEjjdsj cos 0 ¼ I ) jEj

jdsj ¼

) jEj4pz2 ¼ )E¼

Electric Charges at Rest: Part I

q eo

q eo

q eo

q 4peo z2

ð2:220Þ ð2:221Þ ð2:222Þ ð2:223Þ

(c) Using Potential The potential at point P in Fig. 2.37a is given by V¼

1 4peo

Z

r ds0 r

ð2:224Þ

Substituting Eq. 2.207 and equation for ds′ we get V¼

Z

1 r 4peo

1 V¼ 4peo

Z2p d/

R2 sin h0 dh0 d/0 pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ R2 þ z2 2Rz cos h0 0

0

Zp 0

2pR2 r V¼ 4peo

Zp 0

R2 sin h0 dh0 pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ R2 þ z2 2Rz cos h0

sin h0 dh0 pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ R2 þ z2 2Rz cos h0

p 2pR2 r 1 pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 0 2 2 V¼ R þ z 2Rz cos h 4peo Rz 0

ð2:225Þ

ð2:226Þ

ð2:227Þ

V¼

2pR2 r 4peo Rz

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ðR þ zÞ2 ðR zÞ2

ð2:228Þ ð2:229Þ

2.17

Electric Field of a Uniformly Charged Spherical Shell

135

(i) Inside the spherical shell Inside the spherical shell z < R. Hence Eq. 2.229 is V¼

2pR2 r 4peo Rz

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ðR þ zÞ2 ðR zÞ2

ð2:230Þ

V¼

4pR2 r 4peo R

ð2:231Þ

V¼

q 4peo R

ð2:232Þ

Now E ¼ rV @ ^@ @ q ^ ^ þj þk ¼ i @x @y @z 4peo R E¼0 (ii) Outside the spherical shell Outside the spherical shell z > R. Hence Eq. 2.229 is V¼

2pR2 r 4peo Rz

V¼

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ðR þ zÞ2 ðz RÞ2

ð2:233Þ

2pR2 r ½ðR þ zÞ ðz RÞ 4peo Rz

ð2:234Þ

4pR2 r q ¼ 4peo z 4peo z

ð2:235Þ

V¼ Now

E ¼ rV @ ^@ @ q ^ ^ þj þk ¼ i @x @y @z 4peo z E¼

q ^ k 4peo z2

ð2:236Þ

ð2:237Þ

136

2

Electric Charges at Rest: Part I

Of the three methods discussed here Gauss’s law is the easiest as the reader may observe. The next two methods are Coulomb’s law and potential formulation. Luckily in this problem Ex, Ey components cancelled out and Ez component alone exists, and the reader can see how much steps are involved in calculating Ez component using Coulomb’s law. Suppose for a given charge distribution in addition to Ez component if Ex and Ey components do not cancel out then Ex and Ey must be calculated separately. That is, as we calculated Ez through Eqs. 2.205–2.217 we have to repeat similar procedure for Ex, Ey each separately. The reader can see how much tedious procedure it would be. The problem is Coulomb’s law calculates electric ﬁeld directly in terms of vector. There are three different directions x, y and z and hence three components Ex, Ey and Ez. However the potential V is a scalar quantity and we have already seen that it doesn’t have components. So we perform only single integration as we have done in Eqs. 2.224–2.229 and ﬁnally calculate E as E ¼ rV ¼

@V ^ @V ^ @V ^ k iþ jþ @x @y @z

^ E ¼ Ex^i þ Ey^j þ Ez k which is easy as compared to Coulomb’s law.

2.18

Comparison of Coulomb’s Law, Gauss’s Law and Potential Formulation

Let us now compare the methods we used to calculate the electric ﬁeld for a given charge distribution—Coulomb’s law, Gauss’s law and potential formulation. Out of the three methods we see that Gauss’s law is the easiest method. From sections 2.13, 2.16 and 2.17 we observe Gauss’s law is able to calculate the electric ﬁeld intensities for the respective charge distributions very easily as compared to Coulomb’s law and potential formulation. Successful application of Gauss’s law requires construction of Gaussian surface for a given charge distribution over which E ds can be easily evaluated and over the Gaussian surface jEj cos h must be constant so that jEj cos h can be pulled out of the integral I

I E ds ¼

I jEj cos h ds ¼ jEj cos h

ds

2.18

Comparison of Coulomb’s Law, Gauss’s Law and Potential Formulation

Evaluation of the area Gauss’s law

H

137

ds is simple mathematics. E can thus be calculated using

I

I E ds ¼ jEj cos h jEj ¼

ds ¼

qenc eo

1 qenc H cos h ds eo

The above conditions are easily satisﬁed when symmetry exists in the given charge distributions. In Sect. 2.13 cylindrical symmetry exists. In Sect. 2.16 planar symmetry exists. In Sect. 2.17 spherical symmetry exists. Existence Hof symmetries in the respective charge distributions facilitated easy calculation of E ds. If symmetry doesn’t exist in the given problem, then easy calculation of E ds is not possible using Gauss’s law. In such situations Gauss’s law is not useful in H q calculation of E although it is still true that E ds ¼ eenc . o This point has been made clear in Sects. 2.14 and 2.15. When Gauss’ law is not useful, we are left out with two options to calculate the electric ﬁeld—Coulomb’s law and potential formulation. As we have previously mentioned Coulomb’s law calculates E directly in terms of vector. This makes the calculation of E using Coulomb’s law difﬁcult. Section 2.17 illustrates this point. If we are interested in calculating E using Coulomb’s law, then we have to work with three components—Ex, Ey and Ez. However working with potential is easy. Potential is a scalar quantity and doesn’t have to be split into components. Working with vectors is comparatively difﬁcult than working with scalars. This fact has been made clear in Sect. 1.1 and Example 1.10. This point has also been made clear in Sect. 2.17. We will see one more example below to show working with vectors is difﬁcult as compared to working with scalars. We redraw Fig. 2.11a, 2.32 in Fig. 2.42a, b. In Fig. 2.42a the electric ﬁelds due to two surface elements ds01 and ds02 are shown. The electric ﬁeld produced by the surface elements at A, B acts along AC and BD, respectively. These ﬁelds are denoted as E1 and E2, respectively. Resolving E1 and E2 as shown in Fig. 2.43 into x and y components we see that x components cancel out while y components add up. Then we need to integrate the Ey components to calculate the net electric at point P due to entire disc. Now consider Fig. 2.44a. We are interested in calculating the electric ﬁeld at point P1 due to the charge density r over the entire disc, and we have resolved the electric ﬁelds, E1, E2, due to the elements ds1, ds2 into x, y components in Fig. 2.44b. Clearly both x and y components are adding up. E1x and E2x and E1y and E2y are in same direction.

138

2

Electric Charges at Rest: Part I

Fig. 2.42 a Charge uniformly distributed over a circular disc along with two surface elements and its ﬁeld vectors at point P. b Charge uniformly distributed over a circular disc along with one surface element. Fig. 2.43 Resolution of electric ﬁeld at point P due to two strips of surface elements shown in Fig. 2.42a.

Summing up we get x and y components due to the two elements. Then we integrate over the entire disc to get Ex and Ey and hence the electric ﬁeld. If the problem happens in three dimensions, then we need to keep track of three components Ex, Ey and Ez as we have mentioned at the end of Sect. 2.17. The trouble with above method is electric ﬁeld being a vector has a direction. For example in Fig. 2.44a E1 is in the direction from P1 to C1 while E2 is in the direction from P1 to D1. Hence you have to resolve E1, E2 into components, then add the x, y components separately and then ﬁnally integrate it to ﬁnd the ﬁeld at P1

2.18

Comparison of Coulomb’s Law, Gauss’s Law and Potential Formulation

139

Fig. 2.44 a Charge uniformly distributed over a circular disc along with two surface elements and its ﬁeld vectors at point P1. b Resolution of electric ﬁeld at point P1 due to two strips of surface elements shown in Fig. 2.44a

due to the entire disc. In Fig. 2.44a, the problem is two-dimensional - What about three dimensions? The problem becomes even more complicated in three dimensions because there are now three components to work with Ex, Ey and Ez. Now consider calculating the electric ﬁeld E using potential formulation. As shown in Fig. 2.45 we need to calculate the potential at point P1 without worrying about “what is the direction of the potential” because potential being a scalar doesn’t have direction. One need not get confused with components whenever working with V because potential being a scalar doesn’t have “direction and hence components”. Thus if we are going to calculate E using Coulomb’s law, then we have to keep track of the direction and mess up with components. “If the charge distribution is

140

2

Electric Charges at Rest: Part I

Fig. 2.45 Charge uniformly distributed over a circular disc along with one surface element

complex, it is very hectic to calculate E using Coulomb’s law because in that case working with components will be highly confusing”. In calculating potential V we need not bother about direction because V is a scalar quantity. Once V has been calculated, then we can calculate E using E ¼ rV. In summary we attempt to ﬁnd E for a given charge distribution by using Gauss’s law. However when Gauss’s law cannot be used, then we calculate E, ﬁrst by calculating potential V. Coulomb’s law is rarely used. Up to this point we have discussed how to calculate electric ﬁeld for various charge distributions. However based on symmetry and charge distribution we can ﬁnd with which coordinate the ﬁeld does not vary. For example consider the inﬁnite line charge distribution shown in Fig. 2.26. Considering cylindrical coordinates (rc, ф, z) when we keep rc and ф constant and vary z the line charge appears to be same for every point of z. Therefore the ﬁeld doesn’t vary with z. Similarly holding rc and z constant if we vary ф once again the line charge appears to be same at every point of ф. Hence the ﬁeld doesn’t vary with ф also. On the other hand if we hold ф and z constant and vary rc, then there is a change in the charge distribution as seen by every point corresponding to given rc and hence ﬁeld changes with rc. Considering the inﬁnite plane sheet of charge, the ﬁeld doesn’t vary with rc also (Eq. 2.172) and is constant to a near point or a very far away point from the plane. As we move away from the plane sheet of charge, more and more charge comes into the ﬁeld of view and hence the ﬁeld doesn’t decrease.

2.18

Comparison of Coulomb’s Law, Gauss’s Law and Potential Formulation

141

Finally we have developed three different methods to calculate the electric ﬁeld of the given charge distribution. Given the initial charge distribution we plug in charge distribution into the respective equations and ﬁnally ﬁnd the electric ﬁeld. However in most practical cases the “initial charge distribution” is not known. Without even knowing the initial charge distribution we are supposed to ﬁnd the electric ﬁeld. Two methods have been developed to calculate electric ﬁeld in such situations: method of images and Laplace’s equation method. Before discussing these methods we will discuss some intermediate topics like uniqueness theorem, boundary conditions, etc., which we will be using in the above methods.

2.19

Electric Field of a Dipole

We will calculate the electric ﬁeld of the dipole in this section. We will be using the concept of dipole in future sections. Two charges of equal magnitude and opposite sign separated by a small distance form a dipole. Figure 2.46a shows a dipole in which there are two charges q and –q separated by a small distance d.

Fig. 2.46a a A dipole - two equal and opposite charges separated by a small distance d

142

2

Electric Charges at Rest: Part I

Fig. 2.46b a The ﬁgure illustrates that at a distant point the lines from q and -q become almost parallel

We are interested in calculating the electric ﬁeld at point P in Fig. 2.46a due to the dipole. It is assumed that the point P is situated at a large distance from O so that d r. As said in Sect. 2.18 ﬁrst attempt will be to use Gauss’s law to calculate the electric ﬁeld. First step is to plot the electric ﬁeld of the dipole as shown in Fig. 2.47. Clearly there is no symmetry in the problem.

Fig. 2.47 .

2.19

Electric Field of a Dipole

143

H Construction of a Gaussian surface over which E ds can be easily evaluated is not possible. Hence Gauss’s law is not useful here. Let us construct a Gaussian surface as shown in Fig. 2.47. Let us apply Gauss’s law to the Gaussian surface. I q E ds ¼ enc eo Here qenc = + q – q = 0. Hence I E ds ¼ 0 Based on the above equation can we conclude E is Hzero everywhere due to the dipole? Clearly E is not zero everywhere. As we know E: ds gives the flux due to E. In Fig. 2.47 at point A ﬁeld lines are entering the Gaussian surface. At point B ﬁeld lines are leaving the Gaussian surface. Thus overH the entire Gaussian surface H E ds ¼ 0 which means the net flux is zero. “Thus E ds ¼ 0 might mean that either E can be zero as it happens in Eq. 2.218 or the net flux can be zero as it happens H in this dipole case. General thinking by plotting the electric ﬁeld will show, why E ds ¼ 0 for the given charge distribution”. With Gauss’s law not helpful here we now turn to potential to calculate the electric ﬁeld. The potential at point P due to +q charge is ¼

1 q 4peo r1

ð2:238Þ

The potential at point P due to –q charge is ¼

1 q 4peo r2

ð2:239Þ

The potential at point P due to both the charges +q and –q is then 1 q q 4peo r1 r2 q r2 r1 V¼ 4peo r1 r2 V¼

ð2:240Þ ð2:241Þ

From Fig. 2.46b we see that as d r 1 r1 r d cos h 2

ð2:242Þ

144

2

Electric Charges at Rest: Part I

1 r2 r þ d cos h 2

ð2:243Þ

and

1 d cos h r1 r2 r 2

2

2

r2

ð2:244Þ

Substituting Eqs. 2.242–2.244 in Eq. 2.241. V¼

q d cos h 4peo r2

If h ¼ 90 then V ¼ 0

ð2:245Þ ð2:246Þ

That is the plane at y = 0 is at zero potential. We know that in spherical polar coordinates E ¼ rV @V 1 @V ^ 1 @V ^ ^r þ / E¼ hþ @r r @h r sin h @/

ð2:247Þ

Differentiating Eq. 2.245 @V qd cos h ¼ @r 2peo r3

ð2:248Þ

@V qd sin h ¼ @h 2peo r2

ð2:249Þ

@V ¼0 @/

ð2:250Þ

Substituting Eqs. 2.248–2.250 in Eq. 2.247 E¼

qd ½2 cos h^r þ sin h ^h 4peo r3

ð2:251Þ

The above equation can be written as E ¼ Er þ Eh ¼ Er ^r þ Eh ^h

ð2:252Þ

2.19

Electric Field of a Dipole

145

where Er ¼

qd cos h ; 2peo r3

Eh ¼

qd sin h 4peo r3

The magnitude of the resultant dipole ﬁeld is given by E¼

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ E2r þ E2h

sﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ qd cos h 2 qd sin h 2 E¼ þ 2peo r3 4peo r2 E¼

ﬃ qd pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 2 hþ1 3 cos 4peo r3

ð2:253Þ ð2:254Þ

Now let us deﬁne the electric dipole moment. The electric dipole moment p is deﬁned as p = qd [d points from –q to +q]. Then we can write Eq. 2.245 as V¼

1 qd cos h 4peo r2

V¼

1 qd:^r 4peo r2

)V¼

1 p ^r 4peo r2

ð2:256Þ

From Fig. 2.48 we can write pr ¼ p cos h ph ¼ p sin h We have p ¼ pr ^r þ ph ^ h

ð2:257Þ

) p ¼ p cos h ^r p sin h ^h

ð2:258Þ

Now using Eq. 2.258 we can write 3ðp:^rÞ ^r p ¼ 3p cos h ^r ½p cos h^r p sin h ^h ¼ 2p cos h ^r þ p sin h ^h

ð2:259Þ

146

2

Electric Charges at Rest: Part I

Fig. 2.48 Figure shows electric dipole moment of a dipole

Equation 2.251 is E¼

1 ½2p cos h ^r þ p sin h ^h 4peo r3

ð2:260Þ

Substituting Eq. 2.259 in 2.260 E¼

1 ½3ðp ^rÞ^r p 4peo r3

ð2:261Þ

Equation 2.261 gives an interesting result. The electric ﬁeld of a dipole falls off like 1/r3. In comparison the electric ﬁeld of a point charge falls off like 1/r2. Thus electric ﬁeld of a dipole falls off more rapidly than that of the point charge. This is physically expected because at far away distances the opposite charges +q and –q appear to be closely spaced and hence the ﬁelds cancel out more rapidly than that of the point charge. In Fig. 2.49 we show the ﬁeld of a point dipole. Point dipole is one in which we allow d to approach zero while q approaches inﬁnity keeping the product p = qd constant.

2.20

Calculation of Potential Using V ¼

R

E dl

In the previous sections we saw how to calculate the electric ﬁeld E once the potential V is calculated. However if we are interested in calculating potential or potential difference between two points, instead of calculating electric ﬁeld, then

2.20

Calculation of Potential …

147

Fig. 2.49 Electric ﬁeld pattern of a point dipole

there are two methods to calculate it. The ﬁrst method is to calculate potential directly using Eqs. 2.80–2.82. Another method is to calculate E using Gauss’s law; if symmetry exists R for the given charge distribution, then calculate potential using the relation V ¼ E dl. We will discuss the second method here. (a) Potential due to inﬁnitely long charged wire In Sect. 2.13 we calculated the electric ﬁeld due to inﬁnitely long straight wire by applying Gauss’s law E¼

k 2peo r

In Fig. 2.50 we show the inﬁnitely long straight wire carrying a uniform line charge density k. Consider two points P and Q situated at a distance rP and rQ from the wire, respectively. We are interested in calculating the potential difference between the two points, P and Q. The potential at point P is given by ZP VP ¼

E dl 1

ð2:262Þ

148

2

Electric Charges at Rest: Part I

Fig. 2.50 A inﬁnitely long straight wire carrying a uniform line charge density

The potential at point Q is given by ZQ VQ ¼

E dl

ð2:263Þ

1

Now, ZP VP VQ ¼

ZQ E dl þ

1

1

ZP

Z1

VP VQ ¼

E dl 1

E dl

ð2:264Þ

E dl

ð2:265Þ

Q

ZP VP VQ ¼

E dl

ð2:266Þ

E dl cos h

ð2:267Þ

k dl cos h 2peo r

ð2:268Þ

Q

ZP VP VQ ¼ 1

ZP VP VQ ¼ Q

2.20

Calculation of Potential …

149

From Fig. 2.50 dl cos h ¼ dr

ð2:269Þ

Substituting Eq. 2.269 in Eq. 2.268 k Vp VQ ¼ 2peo Vp VQ ¼

Zrp

dr r

ð2:270Þ

rQ

k rQ ln 2pxo rP

ð2:271Þ

(b) Potential of a Spherical Shell The electric ﬁeld of spherical shell has been calculated by Gauss’s law in Sect. 2.17. E¼0 E¼

if

r\R

q ^r if 4peo r2

r[R

where R is radius of the spherical shell. For a point interior to the spherical shell as shown in Fig. 2.51a the potential is Zr V¼

E dl 1

Fig. 2.51 a An interior point in a spherical shell carrying a uniform surface charge density. b An exterior point in a spherical shell carrying a uniform surface charge density

150

2

Electric Charges at Rest: Part I

In spherical polar coordinates ^ dl ¼ dr ^r þ r dh^ h þ r sin h d/ / (i) Inside the Spherical Shell For a point inside the spherical shell as shown in Fig. 2.51a Zr V¼

ZR E dl ¼

1

ZR V¼ 1

Zr E dl

1

E dl

ð2:272Þ

R

q ^ ^r:ðdr ^r þ rdh ^ h þ r sin h d/ /Þ 4peo r2

Zr 0 dl

ð2:273Þ

R

ZR V¼ 1

V¼

q dr 4peo r2

q 1 R 4peo r 1

ð2:274Þ

q 4peo R

ð2:275Þ

V¼

Thus inside the spherical shell the potential is constant while the ﬁeld is zero. (ii) Outside the Spherical Shell At a point outside the spherical shell as shown in Fig. 2.51b Zr V¼

E dl 1

Zr V¼ 1

q ^ ^r ½dr ^r þ r dh ^ h þ r sin h d/ / 4peo r2 Zr V¼ 1

V¼

q dr 4peo r2

q 4peo r

ð2:276Þ

ð2:277Þ

ð2:278Þ

2.21

2.21

The Conservative Nature of Electric Field

151

The Conservative Nature of Electric Field

Consider a point charge situated in origin as shown in Fig. 2.52. The electric ﬁeld produced by the point charge at any point situated at a distance of r from origin is E¼

1 q ^r 4peo r2

Let us calculate the line integral of E from point A to B in Fig. 2.52. In spherical polar coordinates ^ dl ¼ dr ^r þ rdh ^ h þ r sin h d/ / Hence E dl ¼

Fig. 2.52 A point charge situated in origin

1 q dr 4peo r2

152

2

ZB E dl ¼

q 4peo

A

Thus integral

RB R BA

ZB

Electric Charges at Rest: Part I

dr q 1 1 ¼ r2 4peo rA rB

ð2:279Þ

A

E dl depends only on the end points rA and rB and is a constant. The E dl is independent of path.

RB One can go from A to B in any path in evaluating the integral A E dl, but still h i RB q q 1 whatever path selected the result of integration A E dl is always 4pe rB . o rA A

If the path AB is a closed path, then rA = rB and hence I E dl ¼ 0

ð2:280Þ

If the line integral of a vector ﬁeld around a closed path is zero, such a ﬁeld is known as conservative ﬁeld. From Eq. 2.280 I QE dl ¼ 0

ð2:281Þ

F dl ¼ 0

ð2:282Þ

W¼0

ð2:283Þ

I

where W is the work done. Thus the work done for a conservative ﬁeld around a closed path is zero. Equation 2.279 can be written as ZB E dl ¼

1 q q ¼ VA VB 4peo rA rB

ð2:284Þ

A

Hence potential difference is independent of the path selected between points A and B and depends only on end points A and B.

2.22

The Reference Point R in the Equation …

2.22

153

RP The Reference Point R in the Equation V ¼ E dl R

In Eq. 2.65 we have deﬁned the potential as ZP V¼

E dl R

and noted down that R is the reference point where we set the potential to be zero. In this section we will discuss about the reference point. If we want to measure the length of an object using a metre scale, then we ﬁx zero marking on the metre scale as zero length and with respect to the zero marking we measure the length of any object. If we want to measure the temperature of a body using mercury thermometer, the zero marking on the mercury thermometer is taken to be zero and with respect to the zero reference point we measure the temperature of the body. This is shown in Fig. 2.53. Similarly for the case of potential we require a reference point where the potential is set to be zero and with respect to the zero reference point we measure the potential of any other point. The normal procedure is that the inﬁnite point is taken as the reference point where we set the potential to be zero. Hence ZP V¼

E dl 1

This is shown in Fig. 2.54. ∞ is the point where we ﬁx the potential to be zero and from inﬁnity we measure the potential V at point P. We ﬁx the inﬁnity as reference point for free charges situated in space. But in actual practical problems ground is the reference point where we ﬁx the zero reference point for potential. That is we change the reference point from inﬁnity to ground. In order to understand this point consider Fig. 2.55 where we change the reference point from R to R1. It can be shown that V ¼ j þ V1

ð2:285Þ

where RP V ¼ E dl is the potential at point P with respect to reference point R in R

Fig. 2.55.

154

2

Electric Charges at Rest: Part I

Fig. 2.53 Figure showing zero reference point for a) Scale b) Thermometer

Fig. 2.54 Figure showing zero reference point for potential

2.22

The Reference Point R in the Equation …

155

Fig. 2.55 Figure showing change of reference point from R to R1 to measure potential at point P

V1 ¼

RP

E dl is the potential at point P with respect to reference point R1 in

R1

Fig. 2.55. j¼

R R1

E dl is a constant as the line integral depends only on the end points

R

(see Sect. 2.21). Taking gradient of Eq. 2.285 rV ¼ rj rV1 ) rV ¼ rV1 ) E ¼ E1

ð2:286Þ

Here rj ¼ 0 because j is a constant. That is changing the reference point doesn’t alter the actual electric ﬁeld. If you are bit confused, see Fig. 2.56. In Fig. 2.56a we are interested in measuring the length of the thread A1 A2 whose length is 5 cm. We see that the length of the thread A1 A2 with reference point set at zero cm is 5 cm. In Fig. 2.56b we have changed the zero of the reference point to 2 cm. We see that the length of the thread A1 A2 is still 7 cm – 2 cm = 5 cm. Hence the conclusion is if we change the reference point the ﬁnal reading gets changed (in Fig. 2.56b we see ﬁnal value 7 cm). But the actual quantity the length of the thread A1 A2 remains unchanged. Similarly when we change the reference point in our potential, the ﬁnal reading of potential changes because some constant j is added to existing potential. But the actual quantity the electric ﬁeld E remains unchanged (Eq. 2.286).

156

2

Electric Charges at Rest: Part I

Fig. 2.56 a Reference point is set as zero cm in the scale. b Reference point is changed to 2 cm in the scale

As the electric ﬁeld doesn’t change when we change the reference point, in Sects. 2.13c and 2.16c we calculate potential at some reference point ro to determine the electric ﬁeld whenever the charge distribution tends to inﬁnity. Another important point to be noted is it is the potential difference between two points which has physical signiﬁcance and not the potential itself. As an example if you are asked to measure the distance between earth and sun, then you measure the distance of sun from earth and report the value. In the above statement it is implicitly assumed earth is the zero reference point for distance. Fixing earth as zero reference point for distance, from earth the distance of sun is measured. Suppose assume that in this universe only sun is present no earth, no planets, no stars, nothing else. Then in that case distance between what and what? When sun alone is present in empty space, then distance has no physical meaning in the above example. Or in other words it is the difference in distance between two points that has physical signiﬁcance and not the distance itself. Similarly, it is the potential difference between two points, which has physical signiﬁcance and not the potential itself. Fixing earth as zero reference point for distance, from earth the distance of sun is measured in the above example. Similarly in Eq. 2.66 we measure the potential at point P from inﬁnity where we have ﬁxed the zero reference point of the potential. Some students raise objection by looking at Eq. 2.140 - that in Eq. 2.140 we are integrating from 0 to L and not from ∞ to L. But there is a difference between integrations in Eq. 2.66 and Eq. 2.140. Can the reader say what is the difference?

2.22

The Reference Point R in the Equation …

157

Example 2.11 Consider two concentric spherical shells of radii R1 and R2 as shown in Fig. 2.57. The inside shell with radius R1 carries a charge q1, and the outside shell with radius R2 carries a charge q2 with the charges q1 and q2 uniformly distributed over the shells, respectively. Find the electric ﬁeld at points A, B, C in Fig. 2.57. Solution Clearly there is spherical symmetry in the problem. Hence Gauss’s law can be applied to obtain electric ﬁeld. (i) At point A r R1 Construct a spherical Gaussian surface with OA as radius as shown in Fig. 2.58. Applying Gauss’s law I E ds ¼

qenc eo

Here qenc = 0. Hence I E ds ¼ 0

Fig. 2.57 Two concentric spherical shells of radii R1 and R2 carrying charges q1 and q2

158

2

Electric Charges at Rest: Part I

Fig. 2.58 .

On the Gaussian surface the angle between E and ds is zero. Hence I

I jEjjdsj cos 0 ¼ 0 )

jEjjdsj ¼ 0

On the Gaussian surface jEj is constant I jEj jdsj ¼ 0 ) jEj4pr2 ¼ 0 ) jEj ¼ 0 with r ¼ OA: (ii) Field at point B R1 \r R2 Construct a Gaussian surface with OB as radius as shown in Fig. 2.59.

Fig. 2.59 Construction of a Gaussian surface for a point between the two spherical shell

2.22

The Reference Point R in the Equation …

159

Applying Gauss’s law I E ds ¼

qenc eo

Here qenc = q1 and r = OB. I I I E ds ¼ jEjjdsj cos 0 ¼ jEjjdsj I ) jEj jdsj ) jEj4pr2 Hence Gauss’s law is jEj4pr2 ¼

qenc q1 ¼ eo eo

)E¼

1 q1 4peo r2

(iii) Field at point C r > R2 Consider Fig. 2.60. We construct a Gaussian surface with OC as radius. OC = r. Applying Gauss’s law I E ds ¼

qenc eo

Here qenc = q1 + q2 and I E ds ¼ jEj4pr2 with r ¼ OC

Fig. 2.60 Construction of a Gaussian surface for a point exterior of the spherical shell

160

2

Electric Charges at Rest: Part I

Hence Gauss’s law is jEj4p r2 ¼

q1 þ q2 eo

) jEj ¼

1 q1 þ q2 4peo r2

Example 2.12 Charge is uniformly distributed through out cylindrical volume of inﬁnite length of radius R. The volume charge density is q. Determine the electric ﬁeld intensity at a point inside the cylinder. Solution The problem has cylindrical symmetry. Hence Gauss’s law can be applied to calculate electric ﬁeld. Construct a Gaussian surface of radius r and length L as shown in Fig. 2.61. I E ds ¼

qenc eo

Here qenc ¼ q½pr2 L where pr2 L is the volume of the Gaussian surface. And I I I E ds ¼ jEjjdsj cos 0 ¼ jEjjdsj I ¼ jEj ds ¼ jEj 2prL Hence Gauss’s law is jEj 2prL ¼ qpr2 L ) jEj ¼

Fig. 2.61 Construction of a Gaussian surface for a point interior of the cylinder carrying a uniform volume charge density

qr 2eo

ð2:287Þ ð2:288Þ

2.23

2.23

Poisson’s and Laplace’s Equation

161

Poisson’s and Laplace’s Equation

We know that q eo

ð2:289Þ

E ¼ rV

ð2:290Þ

rE¼

From the above equations r ðrVÞ ¼ ) r2 V ¼

q e0

q eo

ð2:291Þ ð2:292Þ

The above equation is known as Poisson’s equation. In places where there is no charge q ¼ 0 and hence r2 V ¼ 0

ð2:293Þ

The above equation is known as Laplace’s equation. Example 2.13 Consider a uniform spherically symmetric distribution of charge in a sphere of radius R as shown in Fig. 2.62. The charge density is q. Calculate the ﬁeld at a point inside and outside the sphere using Poisson’s and Laplace’s equation. Plot the electric ﬁeld with respect to r.

Fig. 2.62 A sphere carrying uniform volume charge density

162

2

Electric Charges at Rest: Part I

Solution (a) At a point inside the sphere r2 V ¼

q eo

The Laplacian in spherical polar coordinates is given by 1 d 2 dV q r ¼ 2 r dr dr eo As the charge distribution is spherically symmetric no h, / terms appear. d 2 dV q 2 r ) r ¼ dr dr eo Integrating dV qr3 ¼ þj dr 3eo dV qr j ¼ ) þ 2 dr 3e0 r

r2

dV ^r Now Ei ¼ rV ¼ @@rV ^r ¼ dr

qr rj2 ^r Hence Ei ¼ 3e o At r = 0 Ei = ∞. An inﬁnite electric ﬁeld is impossible, and hence the constant j = 0. Thus Ei ¼

qr ^r 3eo

(b) At a point outside the sphere there is no charge and hence q = 0. Here therefore q ¼0 eo 1 d 2 dV r ¼0 ) 2 r dr dr

r2 V ¼

)

dV j ¼ 2 dr r

) Eo ¼ rV ¼ where j is a constant.

dV j ^r ¼ 2 ^r dr r

2.23

Poisson’s and Laplace’s Equation

163

Fig. 2.63 Electric ﬁeld variation for a sphere of radius R carrying a uniform volume charge density

At r ¼ R Hence

Eo ¼ Ei j qR ¼ 3eo R2 qR3 )j¼ 3e0

Substitute the value of j in Eo. Hence qR3 1 ^r 3eo r2 qR3 ^r ) Eo ¼ 3eo r2

Eo ¼

The electric ﬁeld of the spherical distribution of charge is plotted in Fig. 2.63. The electric ﬁeld within the spherical distribution of charge increases linearly. Outside the spherical distribution of charge the electric ﬁeld falls off as inverse square law.

2.24

Conductors

Up to this point we have calculated the electric ﬁeld of a static charge distribution in free space. However in number of practical cases we will be dealing with electric ﬁeld in materials. Materials in general are of three types—conductors, semiconductors and insulators (dielectrics).

164

2

Electric Charges at Rest: Part I

In this book we will be mainly dealing with conductors and dielectrics. Conductors are those materials which consist of free electrons (valence electrons loosely bound to atom) so that the electrons can easily move around within the material. In the case of dielectrics the electrons are tightly bound to the nucleus so that there are no free electrons to move around. In this section we will discuss in detail conductors. In later chapters we will deal with dielectrics. An isolated conductor contains equal number of positive and negative charges; thereby the conductor is electrically neutral. Suppose we charge the conductor with the help of a battery. Will the charge reside inside the conductor? If charge resides inside the conductor, then there will be force of repulsion which repels the charge to the surface of the conductor and hence charges reside on the surface of conductor. The process talks about 10−14 s in a conductor like copper. As net charge resides on the surface of the conductor, inside the conductor there are an equal number of positive and negative charges. Hence the net charge density inside the conductor is zero. q ¼ 0 inside a conductor. Suppose we place a conductor in an external electric ﬁeld as shown in Fig. 2.64. The electric ﬁeld will attract the free electrons towards the right, while a net positive charge will appear on the other side. As the conductor has been placed in free space, the electrons are bounded within the conductor due to the insulating property of free space. In Fig. 2.64 we see that due to separation of charges an internal electric ﬁeld Einternal is established within the conductor which is opposite to Eexternal. Within the conductor both the ﬁelds cancel out each other and hence the ﬁeld inside the conductor is zero. Charge flows inside the conductor to the surface until Einternal = Eexternal. So that the net ﬁeld inside the conductor is zero, i.e. E = 0. The charges residing over the surface of the conductor form a surface charge. The

Fig. 2.64 A conductor placed in an external electric ﬁeld

2.24

Conductors

165

surface charge produces an additional ﬁeld outside the conductor which modiﬁes the applied ﬁeld. Suppose assume that there is an empty cavity within the conductor as shown in Fig. 2.65a. Any net charge resides on the surface of the conductor as shown in the ﬁgure. However the surface of the cavity could also contain charges. Let us apply Gauss’s law to surface S in Fig. 2.65a I E ds ¼

qenc eo

At surface S which is well within the conductor E = 0 and hence qenc = 0 indicating that charges of single type + or – cannot be present on the cavity walls. However if the cavity surface contains an equal number of positive and negative charges as shown in Fig. 2.65b, still qenc = 0 and such a situation cannot be ruled out using Gauss’s law. From Eq. 2.280 I E dl ¼ 0 Let us apply the above equation to loop C1 shown in Fig. 2.65b. In loop C1 we have marked three points a, b and d. The electric ﬁeld points from positive to

Fig. 2.65 a A conductor with an empty cavity.b Equal number of positive and negative charges present on the cavity walls of the conductor

166

2

Electric Charges at Rest: Part I

negative charge as shown in Fig. 2.65b. We have constructed the loop such that E and dl have the same direction, so that the angle between them is zero. Hence the above equation for loop C1 becomes I E dl ¼ 0 Zb )

Zd

Za

E dl þ

E dl þ

a

b

ð2:294Þ E dl ¼ 0

d

The paths bd and da are within the conductor where E = 0. Hence Zb

Zd E dl þ

a

Za 0 dl þ

b

0 dl ¼ 0

ð2:295Þ

d

Zb )

E dl ¼ 0

ð2:296Þ

E dl cos 0 ¼ 0

ð2:297Þ

a

Zb ) a

Zb )

E dl ¼ 0

ð2:298Þ

a

The above integral can be zero only if E = 0. Thus there must not be any electric ﬁeld within the cavity, and hence positive and negative charges shown on the walls of the cavity in Fig. 2.65b cannot exist. Hence the walls of the cavity cannot contain any charges. If the walls of the cavity contain charges, then the electric ﬁeld inside the cavity is nonzero. However we have seen that E = 0 inside the cavity. Hence charges cannot be present on the walls of the cavity. An external electric ﬁeld can be present outside the conductor, but still the cavity inside conductor has zero electric ﬁeld. In general we need not have a bulk conductor as shown in Fig. 2.65. A thin sheath of conducting material is sufﬁcient to have zero electric ﬁeld inside. The conductor with a cavity inside known as “Faraday cage” has important applications. For example in a coaxial cable we will see that the inside material will be always surrounded by a outside conducting sheath in order to protect the inside material from any stray electric ﬁelds present external to the conductor.

2.24

Conductors

167

The conductor is an equipotential surface. As we have seen above the electric ﬁeld inside the conductor under electrostatic condition is zero, i.e. E = 0 inside the conductor. Hence for any two points a and b in the conductor Za VðbÞ VðaÞ ¼

E dl ¼ 0 b

Thus conductors are equipotential surfaces. Outside the conductor the electric ﬁeld is perpendicular to the surface of the conductor. In case if there is any tangential component, then charge flows due to the action of the tangential component of the ﬁeld and force of repulsion, until the tangential component of the ﬁeld is zero. The surface charge is conﬁned within the conducting object, and hence electric ﬁeld is perpendicular to the surface of the conductor.

2.25

Boundary Conditions

Boundary conditions play an important role in calculating how much the electric ﬁeld changes when we cross from one media to the other media. In this section we will see that whenever we cross a surface charge density r, the electric ﬁeld changes by an amount r=eo . Consider a surface charge density r as shown in Fig. 2.66. Consider a pillbox of length a. The area of top surface of the pillbox is A. Part of the pillbox lies above the surface charge, while the other part lies below the surface charge. Applying Gauss’s law to the pillbox I E ds ¼

qenc rA ¼ eo eo

ð2:299Þ

In the limit a ! 0 the contribution to the flux in Gauss’s law comes only from the normal component of the electric ﬁeld. Hence Eq. 2.299 can be written as Eln A E2n A ¼

rA eo

ð2:300Þ

) Eln E2n ¼

r eo

ð2:301Þ

168

2

Electric Charges at Rest: Part I

Fig. 2.66 a The ﬁgure shows a surface charge density between medium 1 and medium 2 and a pill box Gaussian surface. b, c Resolution of electric ﬁeld E1 and E2

Fig. 2.67 The ﬁgure shows a surface charge density between medium 1 and medium 2 along with a rectangular loop

Now consider a surface charge density r as shown in Fig. 2.67. Consider a rectangle of length L and breath a as shown in Fig. 2.66. We know I E dl ¼ 0

2.25

Boundary Conditions

169

In the limit a ! 0 the contribution for the above integral is only from sides with dimension L. Hence E1t L E2t L ¼ 0

ð2:202Þ

) E1t ¼ E2t

ð2:303Þ

Hence the tangential components are equal at the boundary between the two media. Thus the boundary conditions are E1n E2n ¼

r eo

E1t ¼ E2t Suppose medium 1 is free space and medium 2 is conductor as shown in Fig. 2.68. With E = 0 inside the conductor boundary conditions read E1n ¼

r eo

E1t ¼ E2t ¼ 0 where r is the “local” surface charge density.

Fig. 2.68 At the conductor free space boundary the electric ﬁeld is normal to the surface of the conductor

ð2:304Þ ð2:305Þ

170

2

Electric Charges at Rest: Part I

The above boundary conditions are pictorically depicted in Fig. 2.68. Thus at the conductor–free space boundary there are no tangential components and only normal component is present in the free space.

2.26

Uniqueness Theorem

As we have already mentioned previously that, we have considered charges distributed in free space. However in all practical cases we will be dealing with electric ﬁeld in materials. We have already discussed conductors. In reality the conductors will be maintained at different potentials by connecting them to external power sources like batteries. Let us consider an example, shown in Fig. 2.69. In this ﬁgure there are three conducting plates one at x = 0 maintained at potential Vo. The second plate is at y = 0 and is maintained at a potential V = 0. The third plate is at y = b and is maintained at potential V = 0. Thus in Fig. 2.69 we have the following conditions. (i) At x = 0, i.e. V(0, y) = Vo. (ii) At y = 0, i.e. V(x, 0) = 0. (iii) At y = b, i.e. V(x, b) = 0.

Fig. 2.69 Three conducting plates maintained at different potentials.

2.26

Uniqueness Theorem

171

These conditions are known as boundary conditions for given problem. We are interested in ﬁnding potential V(x, y) in the region between three plates in Fig. 2.69. The method to ﬁnd V(x, y) is being found will be discussed in future chapters (see Sect. 3.9). In Fig. 2.69 we are interested in knowing in the region between three plates whether there is only one value of potential V(x, y) (or) two values of potential V1(x, y) and V2(x, y) (or) multiple values of potential. This point will be discussed in this section. The above problem is only an example. In similar problems we are interested in knowing whether the problem has single solution or two solutions or multiple solutions. We have another important information from Fig. 2.69. Suppose assume that there are two potentials or two solutions in the region between the three plates – V1(x, y), V2(x, y). Both the solutions V1 and V2 must be equal to Vo at V(0, y). The potential Vo at V(0, y) is speciﬁed by us, by connecting the conducting plate at x = 0 to external power supplies, say batteries. Hence at the boundary x = 0, V1 = Vo and V2 = Vo. If we deﬁne Vd = V1 – V2, then at the boundary x = 0 Vd ¼ V1 V2 ¼ Vo Vo ¼ 0: Similar argument shows that (i) y = 0 for the plate at y = 0 V1 ¼ 0;

V2 ¼ 0

Vd ¼ V1 V2 ¼ 0 (ii) y = b for the plate at y = b V1 ¼ 0; V2 ¼ 0 Vd ¼ V1 V2 ¼ 0 0 ¼ 0: “The same is true for any given problem. Because we are specifying the potential at the boundaries by connecting the respective boundary to an external power source, the potential, or solution obtained in the region of interest, must be equal to potential given to the boundary, at the boundary. Hence if there are multiple solutions or potentials in the region of interest, the difference between any two solutions Vd must be equal to zero at the boundary”. Vd ¼ 0 at the boundary

ð2:306Þ

“Figure 2.69 shows a speciﬁc problem in which three electrodes are maintained at speciﬁed potentials. In general i number of conductors can be maintained at different potentials as shown in Fig. 2.70a. We are interested in ﬁnding the potential in the surrounding region of the conductors. We are interested in knowing whether

172

2

Electric Charges at Rest: Part I

Fig. 2.70 a Number of conductors maintained at speciﬁed potentials. b A surface So encloses all the conductors maintained at speciﬁed potentials

the problem has no solution, one solution or more than one solution”. The answer to the above question is given by uniqueness theorem. Uniqueness theorem states that the Poisson equation or Laplace’s equation satisfying the given boundary conditions has one and only solution, i.e. unique solution. Let us now consider Fig. 2.70a. A number of conductors whose surfaces are S1, S2…Si are maintained at speciﬁed potentials. Construct a surface So which encloses all the conductors, and the total volume is s. Contrary to uniqueness let us assume there are two solutions V1 and V2 in the region of interest. Then r2 V 1 ¼

q eo

ð2:307Þ

r2 V 2 ¼

q eo

ð2:308Þ

Let us denote the difference in potentials V1 and V2 as Vd. We have Vd ¼ V1 V2

ð2:309Þ

) r2 Vd ¼ r2 V1 r2 V2 q q ¼ þ ¼0 eo eo Hence r2 V d ¼ 0

ð2:310Þ

2.26

Uniqueness Theorem

173

From Sect. 1.20, using vector identity 10 we can write r ðVd rVd Þ ¼ Vd r:ðrVd Þ þ rVd :rVd

ð2:311Þ

r ðVd rVd Þ ¼ Vd r2 Vd þ jrVd j2

ð2:312Þ

Integrating the above equation over the volume s Z

Z

Z

r:ðVd rVd Þds ¼ s

jrVd j2 ds

Vd r Vd ds þ 2

s

ð2:313Þ

s

Substituting Eq. 2.310 in 2.313 Z

Z r ðVd rVd Þds ¼

s

jrVd j2 ds

ð2:314Þ

jrVd j2 ds

ð2:315Þ

s

Applying Gauss’s divergence theorem I

Z ðVd rVd Þ ^ ai ds ¼ s

S

Here the surface S encloses S0, S1, S2…Si. Hence I

Z ðVd rVd Þ ^ ai ds ¼

jrVd j2 ds

ð2:316Þ

s

S0 þ ðS1 þ S2 þ þ Si Þ

On the conducting boundaries S1, S2, … Si Vd is zero from Eq. 2.306. Hence Eq. 2.316 reduces to Z Z ðVd rVd Þ ^ ai ds ¼ jrVd j2 ds ð2:317Þ s

So

We have constructed surface S0, and we can construct the surface S0 as a sphere of inﬁnite radius. That is the surface S0 lies at very large distance from the conducting boundaries. Because the surface S0 is of large radii, the conducting boundaries will look like point charge to the surface S0. Hence V1 and V2 fall off as 1 1 1 r. So Vd ¼ V1 V2 also falls off like r :rVd then falls off like r2 . Thus the term Vd rVd falls off like r13 . The surface S0 for the sphere increases as 4pr2 (surface area H of sphere is 4pr2). Thus the total integral ðVd rVd Þ ds falls off like 1r. At large H S distances as r goes of to inﬁnity then ðVd rVd Þ ds falls off to zero. S

174

2

Electric Charges at Rest: Part I

Thus I ðVd rVd Þ ds ¼ 0

ð2:318Þ

Hence from Eq. 2.317 Z jrVd j2 ds ¼ 0

ð2:319Þ

s

Because jrVd j2 is positive everywhere, the above integration can be zero only when jrVd j ¼ 0

ð2:320Þ

) Vd ¼ j ½for all points within s

ð2:321Þ

for all point within s.

where j is a constant. We see that Vd = j at all points within s which includes surfaces S1, S2, …Si, but already we have seen that on the surfaces S1, S2, …Si the value of Vd is zero. Hence at the boundaries Vd ¼ j ¼ 0

ð2:322Þ

But from Eq. 2.321 j is equal to Vd for all points within s. From Eq. 2.322 at the boundaries S1, S2… Si which are enclosed in volume s, Vd = j = 0. Hence j is zero everywhere within volume s. Hence Eq. 2.321 is Vd ¼ j ¼ 0 for all points within s:

ð2:323Þ

) Vd ¼ 0 ) V1 V2 ¼ 0 ) V1 ¼ V2

ð2:324Þ

Hence Uniqueness theorem is proved. Thus as an example, in the region between three plates in Fig. 2.69 there is only one possible potential or solution, V (x, y), for the boundary conditions stated at the beginning of this section.

2.26

Uniqueness Theorem

175

Exercises Problem 2.1 Consider a inﬁnite long cylinder of radius R. The surface of the cylinder contains uniform charge density r. A uniform line charge of charge density k lies in the axis of the cylinder. Calculate the electric ﬁeld inside and outside the cylinder. Which method you will use to calculate the electric ﬁeld and why? Problem 2.2 Consider a circular loop of radius R. Calculate the electric ﬁeld at a point P as shown in Fig. 2.71. Point P lies along the axis of the loop and is at a distance of a from the centre of the ring. Problem 2.3 Three charges of magnitude q1 ¼ 2:5 nC, q2 ¼ 5 nC, q3 ¼ 7 nC are situated as shown in Fig. 2.72 with OA = 12 cm, OB = 9 cm. Calculate the force q1. Problem 2.4 Solve Example 2.13 using Gauss’s law. Problem 2.5 If three charges of magnitude q are situated at the three corners of the square, what would be the ﬁeld at point P in Fig. 2.73.

Fig. 2.71 A circular loop. Illustration for problem 2.2

176

2

Electric Charges at Rest: Part I

Fig. 2.72 Illustration for problem 2.3.

Fig. 2.73 Three charges situated at three corners of the square. Illustration for problem 2.5

Fig. 2.74 A semicircle carrying a uniform line charge density. Illustration for problem 2.6.

Problem 2.6 In Fig. 2.74 an upper portion of the semicircle carries a uniform line charge k. Calculate the ﬁeld at point P. The radius of the semicircle is R. Problem 2.7 Consider an annular disc of inner radius R1 and outer radius R2 having a uniform surface change density r. Determine the electric ﬁeld at point P along the axis of the disc situated at a distance a from the centre of the disc as shown in Fig. 2.75.

Exercises

177

Fig. 2.75 An annular disc carrying a uniform surface charge density. Illustration for problem 2.7

Fig. 2.76 I A hollow sphere carrying a volume charge density. Illustration for problem 2.8

Problem 2.8 Consider a hollow sphere as shown in Fig. 2.76. Within the region R1 r R2 the sphere contains a volume charge density q ¼ kðr0 Þ2 . Find the electric ﬁeld at a point r R1 ; R1 r R2 ; r R2 (k is a constant). Problem 2.9 Calculate the ﬁeld inside and outside of an inﬁnitely long hollow cylinder whose radius is R. The cylinder carries a uniform surface charge density r. Problem 2.10 Calculate the ﬁeld inside and outside of an inﬁnitely long solid cylinder whose radius is R. The cylinder carries a uniform volume charge density q. Problem 2.11 Consider inﬁnitely long hollow cylinder as shown in Fig. 2.77. The inner and outer radii of the cylinder are R1 and R2, respectively. The cylinder carries a charge density of q ¼ kðr0 Þ2 within R1 r R2 . Calculate the electric ﬁeld at a point r R1 ; R1 r R2 ; r R2 .

178

2

Electric Charges at Rest: Part I

Fig. 2.77 A hollow cylinder carrying a volume charge density. Illustration for problem 2.11

Problem 2.12 Check whether the following electrostatic ﬁelds are possible ^ (i) E ¼ x^i þ y^j þ zk 2^ 2^ ^ (ii) E ¼ xyz i þ 3x j þ 2yk Problem 2.13 ^ Given Ex ¼ 3z2 y. Find An electrostatic ﬁeld E is given by E ¼ Ex ^i þ Ey ^j þ Ez k. Ey, Ez components (Hint: use the relation r E ¼ 0). Problem 2.14 Consider a uniformly charged solid sphere of radius R with charge density q. Using inﬁnity as reference point calculate the potential inside and outside the sphere. Problem 2.15 Consider a rectangle of sides 3 and 7 cm as shown in Fig. 2.78. Four charges are situated at the corners of the rectangle as shown in the ﬁgure. Find the magnitude and direction of the resultant force on qA. Medium is free space. Problem 2.16 A unit positive charge is situated at point P as shown in Fig. 2.79. Here q1 = 1 nC and q2 = 3 nC. Calculate the force in the unit positive charge. Medium is free space.

Fig. 2.78 Illustration for problem 2.15

Exercises

179

Fig. 2.79 Illustration for problem 2.16

Problem 2.17 Three charges qA, qB and qC each of magnitude 3, 2 and –5 nC are located at (1, 0, 0), (0, 3, 0) and (0, 0, 7). Calculate the electric ﬁeld at (0, 5, 0). Medium is free space. Problem 2.18 A square consists of two charges qA = 2 nC and qB = 1 nC as shown in Fig. 2.80. The other side of the square consists of a line charge k ¼ 1010 coul=m as shown in the ﬁgure. Calculate the potential at point O. Problem 2.19 Calculate E if the potential is given by (k is a constant) V¼

k x2 þ y2

Problem 2.20 Consider a sphere of surface area 64 p m2. A point charge of magnitude 1 nC is located somewhere within the sphere. Calculate the electric flux at a point located at 5 m from the centre of the sphere.

Fig. 2.80 Illustration for problem 2.18

180

2

Electric Charges at Rest: Part I

Problem 2.21 Calculate the potential between two coaxial cylinders with inner radius R1 and outer radius R2. Fix the zero reference point of the potential at the outer cylinder. Problem 2.22 A spherical volume charge density distribution is given by h i q ¼ k 1 ðr0 Þ2 for r R and ¼0 Find the total charge q.

for r [ R ½k is a constant

Chapter 3

Electric Charges at Rest—Part II

3.1

Work Done

Suppose assume that we are interested in moving a charge Q from point a1 to point a2 in the presence of a charge q as shown in Fig. 3.1a. The work done in the process is Za2 W¼

F dl

ð3:1Þ

a1

Charge q will exert a force of QE to oppose the movement and hence we have exert a minimum force of –QE to make the charge Q move from a1 to a2. Hence Eq. 3.1 is Za2 W ¼ Q

E dl

ð3:2Þ

a1

With the aid of Eq. 2.284, Eq. 3.2 can be written as W ¼ Q½Vða1 Þ Vða2 Þ

ð3:3Þ

) W ¼ Q½Vða2 Þ Vða1 Þ

ð3:4Þ

Suppose we move charge Q from inﬁnity to point a2 as shown in Fig. 3.1b then W ¼ Q½Vða2 Þ Vð1Þ

ð3:5Þ

as Vð1Þ ¼ 0 we have © Springer Nature Singapore Pte Ltd. 2020 S. Balaji, Electromagnetics Made Easy, https://doi.org/10.1007/978-981-15-2658-9_3

181

182

3 Electric Charges at Rest—Part II

Fig. 3.1 a Charge Q is moved from a1 to a2 in the presence of charge q. b Charge Q is moved from inﬁnity to a2 in the presence of charge q

W ¼ Q V ð a2 Þ

ð3:6Þ

In Sect. 2.22 we have noted that occasionally we will be changing the zero reference point for calculating potential. One interesting point to be noted is that work done doesn’t change when the zero reference point is changed. Figure 2.55 is reproduced in Fig. 3.2. From Eq. 2.288 we see that Vða1 Þ ¼ j þ V1 ða1 Þ

ð3:7Þ

Vða2 Þ ¼ j þ V1 ða2 Þ

ð3:8Þ

where V(a1), V(a2) are the potentials at points a1, a2 with respect to reference point R. V1(a1), V2(a2) are the potentials at points a1, a2 with respect to reference point R1 (see Fig. 3.3).

Fig. 3.2 Figure showing change of reference point from R to R1 to measure potential at point P

3.1 Work Done

183

Fig. 3.3 Charge Q is moved from a1 to a2 in the presence of charge q. Reference point R and R1 are shown

From Eqs. 3.7 and 3.8 Vða1 Þ Vða2 Þ ¼ V1 ða1 Þ V1 ða2 Þ

ð3:9Þ

) Q½Vða1 Þ Vða2 Þ ¼ Q½V1 ða1 Þ V2 ða2 Þ

ð3:10Þ

Thus work done doesn’t get altered by changing the reference point.

3.2

Energy in Electrostatic Fields

Suppose assume that we are interested in moving a charge q1 from inﬁnity and placing it at a point A as shown in Fig. 3.4. The amount of work done in the process is zero because there is no electric ﬁeld against which we are doing work. Hence,

Fig. 3.4 Three charges q1 , q2 and q3 are moved from inﬁnity and placed at A, B and C

184

3 Electric Charges at Rest—Part II

W1 ¼ 0

ð3:11Þ

Next let us move charge q2 from inﬁnity and place it at point B at a distance of

r12 from charge q1 as shown in Fig. 3.4. The amount of work done in the process is W ¼ q2 V21 ðusing Eq. 3:6Þ where V21 is the potential produced by charge q1 at the site of charge q2 W ¼ q2 Because

q1 4peo r21

ð3:12Þ

r12 ¼ r21 in Fig. 3.4, Eq. 3.12 can be written as W 2 ¼ q1

q2 4peo r12

ð3:13Þ

W2 ¼ q1 V12

ð3:14Þ

where V12 is the potential at the site of q1 produced by charge q2. As W1 = 0 total work done in assembling charges q1 and q2 is 1 W2 ¼ q2 V21 ¼ q1 V12 ¼ ðq1 V12 þ q2 V21 Þ 2 1 1 q1 q2 q2 q1 W2 ¼ þ 2 4peo r12 r21

ð3:15Þ

Now let us bring another charge q3 from inﬁnity and place it at point C as shown in Fig. 3.4. The work done in placing charge q3 at point C is Wo ¼ q3 ½V31 þ V32

ð3:16Þ

where V31 and V32 have usual meaning similar to V12. Equation 3.16 is W o ¼ q3

q1 q2 þ 4peo r31 4peo r32

ð3:17Þ

The total work done in assembling three charges q1, q2, q3 is then W3 ¼ W1 þ W2 þ Wo Substituting Eqs. 3.11, 3.13, 3.17 in Eq. 3.18 we obtain W3 ¼

1 q1 q2 q3 q1 q3 q2 þ þ 4peo r12 r31 r32

ð3:18Þ

3.2 Energy in Electrostatic Fields

185

The above equation can be written as W3 ¼

1 1 q2 q1 q1 q2 q3 q1 q1 q3 q3 q2 q2 q3 þ þ þ þ þ 2 4peo r21 r12 r31 r13 r32 r23

ð3:19Þ

In general if we are assembling n number of charges we can write " # n X n q i qj 1 1 X W¼ 2 4peo i¼1 j¼1 rij

ð3:20Þ

with i 6¼ j Equation 3.19 can be rewritten as n n X 1X 1 qj W¼ q 2 i¼1 i j¼1 4peo rij )W¼

n n X 1X qi Vij 2 i¼1 j¼1

ð3:21Þ

ð3:22Þ

P where nj¼1 Vij is the total potential produced at the site of qi by all the remaining j charges. For continuous charges distributions Eq. 3.22 can be written as W¼

1 2

Z q Vds

ð3:23Þ

s

Poisson’s equation (Eq. 2.292) is q ¼ eo r2 V

ð3:24Þ

Substituting Eq. 3.24 in Eq. 3.23 W¼

eo 2

Z

V r2 V ds

s

From Sect. 1.20 using vector identity 10 we have r ðu AÞ ¼ u r A þ A r u Setting u = V, A ¼ r V We get

ð3:25Þ

186

3 Electric Charges at Rest—Part II

r (V r V) = V r ðr VÞ þ r V r V

ð3:26Þ

r ðV r VÞ ¼ V r2 V þ rV rV

ð3:27Þ

V r2 V ¼ r ðV r VÞ rV rV

ð3:28Þ

Noting down that E ¼ r V Vr2 V ¼ r ðVEÞ E2 Substituting Eq. 3.29 in Eq. 3.25 Z Z eo eo r ðVEÞds þ E2 ds W¼ 2 2 s

ð3:29Þ

ð3:30Þ

s

Applying Gauss divergence theorem I Z eo eo W¼ VE dS þ E2 ds 2 2 S

ð3:31Þ

s

In Eq. 3.31 we are integrating over the volume s: The volume s is the volume which encloses all the charges. However if we increase the volume s beyond the volume enclosing all the charges the work done given by Eq. 3.31 is not going to increase as s increases, because q = 0 in outer volume (q = 0 in Eq. 3.23 for outer volume that doesn’t contain charges). Hence let us select the sphere of very large radius. At such large distances charge density q will look like a point charge. Thus in the ﬁrst term of Eq. 3.31 V falls of as 1r ; E falls of as r12 while the surface S H increases as r2 so the entire term V E dS falls of as 1r : H S Thus for r going to inﬁnity S V E dS falls of to zero. Hence Eq. 3.31 is Z 1 W ¼ eo E2 ds ð3:32Þ 2 s In the above equation E2 falls of as r14 while the volume increases as r3 (volume R of sphere is 43 p r3 ). So the entire integral s E2 ds must fall of as 1r and at large r must go to zero. But it is incorrect. Can the reader say why? Finally for line and surface charge densities we have

3.2 Energy in Electrostatic Fields

187

1 W¼ 2 and

W¼

Z kVdl 1 2

ð3:33Þ

Z rV dS

ð3:34Þ

The work done in assembling the charges is stored as potential energy in the system. While Eq. 3.23 suggests electrostatic energy is stored in the charge, Eq. 3.32 suggests energy is stored in the ﬁeld surrounding the charges. The exact location where the energy is stored is not answerable and only thing that can be suggested is, both equations calculate the electrostatic energy for a given charge distribution correctly.

3.3

Equipotential Surfaces

Equipotential surfaces are those surfaces for which all the points on the surface have same potential. From Eq. 3.4 W ¼ Q½Vða2 Þ Vða1 Þ For an equi-potential surface V(a2) = V(a1) and hence work done on a equation potential surface is zero. Let us move a charge Q to distance dl on a equi-potential surface as shown in Fig. 3.5a. The work done in the process is zero because the surface is an equi-potential surface.

Fig. 3.5 a Charge q moved on an equipotential surface through a small distance. b The electric ﬁeld is perpendicular to the equipotential surface

188

3 Electric Charges at Rest—Part II

Hence dW ¼ F dl ¼ 0 ) QE dl ¼ 0 ) E dl ¼ 0 ) E dl cos h ¼ 0 As E 6¼ 0; dl 6¼ 0 cos h ¼ 0: ) h ¼ 90 Hence the angle between E, dl is 90° for an equipotential surface. This is shown pictorially in Fig. 3.5b. Thus E is perpendicular to a equipotential surface. Now consider a point charge q as shown in Fig. 3.6a. As the potential of the q point charge is given by 4peo r ; all points equidistant from the point charge are at same potential. Thus the equi potential surfaces for a point charge are concentric spheres as shown in Fig. 3.5a. The potential however will be different for different spheres. Because the electric ﬁeld is perpendicular to the equipotential surface its direction is radial as shown in the same ﬁgure.

Fig. 3.6 a Equipotential surface for a point charge are concentric spheres. b For an uniform electric ﬁeld equipotential surfaces are planes

3.3 Equipotential Surfaces

189

For a uniform electric ﬁeld in Fig. 3.6b the equipotential are planes which lie perpendicular to the electric ﬁeld. In Sect. 2.24 we have seen that under electrostatic conditions the electric ﬁeld inside a conductor is zero. From Eq. 2.284 ZB E dl = VA VB A

as E = 0 inside a conductor 0 ¼ VA VB VA ¼ VB : Thus any two points in a conductors are at same potential. Hence conductors are equi potential surfaces under electrostatic conditions. Thus work done to move a charge in conductor under electrostatic condition is zero.

3.4

A Note on Work Done

Similar to Eqs. 2.11, 2.58 and 2.77 can we write W ¼ W 1 þ W2 þ þ W n W¼

n X

Wi

ð3:35Þ ð3:36Þ

i¼1

See Problem 3.21. Example 3.1 A total charge q is uniformly distributed through out a sphere of radius R and with a volume charge density of q. Calculate the energy stored in the sphere. Solution From Example 2.13 we see that q R3 outside the sphere 3eo r2 qr Ein ¼ inside the sphere 3eo Eout ¼

190

3 Electric Charges at Rest—Part II

(a) Using the expression W ¼ e2o

R1 0

E2 ds

We have W¼

eo 2

eo W¼ 2

Z E2 ds Zp Z2p Z1 E2 ds 0

0

0

2 3 Zp Z2p ZR Z1 eo 4 E2 ds þ W¼ E2out ds5 in 2 0

eo W¼ 2

0

R

0

Zp Z2p ZR E2in 0

0

ð3:37Þ

eo ds þ 2

Zp Z2p Z1 E2out ds

0

0

0

R

Inside the sphere Zp Z2p ZR

Zp Z2p ZR E2in ds

0

0

¼

0

0

¼ ¼

0

q2 3e20

0

q2 r 2 2 r sin h dr dh d/ 9e20

ZR

Zp

0 2

Z2p sin h dh

r4 dr 0

d/

ð3:38Þ

0

5

4pq R 45 e20

Outside the sphere Zp Z2p Z1

Zp Z2p Z1 E2out 0

0

R

q R3 3 eo r 2

ds ¼ 0

0

q2 R ¼ 9 eo

6

R

Z1

dr r2

R

4p q2 R5 ¼ 9 e20 Substituting Eqs. 3.38, 3.39 in Eq. 3.37

2 r2 sin h dr dh d/

Zp

Z2p sin h dh

0

d/ 0

ð3:39Þ

3.4 A Note on Work Done

191

eo 4p q2 R5 eo 4p q2 R5 W¼ þ 2 2 45 e20 9 e20 )W¼

(b) Using equation W ¼

1 2

4p q2 R5 15 eo

Z q V ds

Z 1 q V ds to calculate the energy stored. 2 Let us calculate the potential V at a point inside the sphere

Now let us use the equation W ¼

Zr V¼

E dl 1

where r is a distance of point P inside the sphere as shown in Fig. 3.7 ZR )V¼

Zr E dl

1

ZR )V¼ 1

R

qR3 dl 3eo r2

q 2 )V¼ 3R r2 6eo

Fig. 3.7 A sphere with uniform volume charge density

E dl Zr R

qr dl 3eo

192

3 Electric Charges at Rest—Part II

Hence the work done is given by Z 1 q V ds W¼ 2 s Z 1 q 2 W¼ q 3R r2 r2 sin h dr dh d/ 2 3 eo s

1 q2 )W¼ 12 eo )W¼

ZR

Zp

3R r r dr 2

2

0

Z2p sin h dh

2

0

d/ 0

4p q2 R5 15 eo

Example 3.2 ^ Calculate the work done in moving An electric ﬁeld is given by E ¼ x ^i þ y ^j þ z k: a point charge –10 lC from (a) (0, 0, 0) to (2, 0, 0) (b) (2, 0, 0) to (2, 3, 0) (c) (2, 3, 0) to (0, 0, 0) as shown in Fig. 3.8 and hence show that electrostatic ﬁelds are conservative ﬁelds. Solution (a) We know that Z W ¼ Q

E dl

In Fig. 3.8 Q = –10 lC and ^ d l ¼ dx ^i þ dy ^j þ dz k Fig. 3.8 A point charge moved from (A) (0 , 0, 0) to (2, 0 , 0 ) (B) from (2, 0, 0) to (2, 3, 0) (C) (2, 3, 0 ) to (0, 0, 0)

3.4 A Note on Work Done

193

For path A dy = 0, dz = 0. Hence d l ¼ dx ^i Z2 ^ dx ^i Hence W ¼ 10 ðx ^i þ y ^j þ z kÞ 0

W ¼ 10

2 2 x ¼ 20 lJ 2 0

(b) For path B we move along y-axis and hence dx = 0, dz = 0. Therefore d l ¼ dy ^j Z3 W ¼ 10

^ dy ^j ðx ^i þ y ^j þ z kÞ

0

2 3 y W ¼ 10 ¼ 45 lJ 2 0

Therefore the work done in moving the charge from origin (0, 0, 0) to point (2, 3, 0) via the path A B is 20 + 45 = 65 lJ. (c) For path C, d l ¼ dx ^i þ dy ^j

ð0; Z0; 0Þ

Hence W ¼ Q

^ ðdx ^i þ dy ^jÞ ðx ^i þ y ^j þ z kÞ

ð2; 3; 0Þ

2 ð0; 0; 0Þ 2 ð0; 0; 0Þ x y Q ¼ Q 2 ð2; 3; 0Þ 2 ð2; 3; 0Þ W ¼ 65 lJ: The work done in moving the charge from (0, 0, 0) to (2, 3, 0) via AB is 65 lJ and then from (2, 3, 0) to (0, 0, 0) is –65 lJ. The total work done around the closed path ABC is zero and hence the electrostatic ﬁeld is conservative.

194

3 Electric Charges at Rest—Part II

Fig. 3.9 Four charges each of magnitude q placed in the corner of a rectangle whose sides are a, b

Example 3.3 Four charges each of magnitude q are placed at the corners of a rectangle of sides a, b as shown in Fig. 3.9. Calculate the energy stored in the system. Solution The total energy in the system is given by W ¼ 12

4 P

qi Vi

i¼1

1 W ¼ ½qVA þ qVB þ qVC þ qVD 2 Setting the zero reference of the potential at point A 1 W ¼ ½qVB þ qVC þ qVD 2 From Fig. 3.9 q 1 q ¼ 4p eo AB 4p eo b q 1 q pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ VC ¼ ¼ 4p eo AC 4p eo a2 þ b2 q 1 q VD ¼ ¼ 4p eo AD 4p eo a VB ¼

" # q2 1 1 1 þ pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ þ : Hence W ¼ a 8peo b a 2 þ b2

3.5

Method of Images

In Sects. 2.4, 2.8, 2.11 we saw how to calculate the electric ﬁeld once the charge distribution is given. The charge distribution is plugged into respective Eqs. 2.17 −2.19, 2.35, 2.80−2.82 and ﬁnally the electric ﬁeld is calculated for the given charge distribution. However let us consider the following example and check whether the problem can be solved using previous methods namely, Coulomb’s law, Gauss’s law and potential formulation.

3.5 Method of Images

195

Fig. 3.10 a A point charge q situated above the inﬁnite grounded conducting plane. b Illustration to show that the point charge q is nearer to point O than point A

Consider a point charge q situated above the inﬁnite grounded conducting plane at a distance d above the plane along the z-axis as shown in the Fig. 3.10a. The plane is situated at z = 0. Because the plane is grounded it must be at zero potential i.e., V(x, y, 0) = 0. Now our aim is to calculate the electric ﬁeld above the grounded conducting plane i.e., for all points with z > 0. Below the grounded conducting plane (i.e., for points z < 0) the electric ﬁeld is zero because there are no charges there. Let us calculate the electric ﬁeld at point P above the conducting plane. From qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ Fig. 3.10a we see that BP ¼ x2 þ y2 þ ðz dÞ2 : There are two charges responsible for electric ﬁeld at point P. First one is the point charge q situated at point B in Fig. 3.10a. The point charge q at point B will induce negative charges in the conducting plane and this induced charge will produce a electric ﬁeld at point P. In order to calculate the electric ﬁeld at point P our ﬁrst attempt will be to check whether Gauss’s law can be used to calculate the electric ﬁeld at point P. However a plot of ﬁeld lines for Fig. 3.10a would show you that Gauss’s law is not useful here for calculating electric ﬁeld.

196

3 Electric Charges at Rest—Part II

So we now try to calculate the electric ﬁeld using potential. The potential at point P is given by V¼

q 1 þ 4peo ðBPÞ 4peo

Z S

0

r ds0

r

ð3:40Þ

Here r is the induced negative charge density on the conducting surface and r is the distance between the small surface element ds0 and point P as shown in Fig. 3.10a. Z q 1 r ds0 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ þ V¼ 4peo r 4peo x2 þ y2 þ ðz dÞ2 S0

ð3:41Þ

However now trouble develops. What is the actual form of r? Without knowing r Eq. 3.41 cannot be solved and hence potential, electric ﬁeld cannot be calculated. We can manage to calculate r using the condition V(x, y, 0) = 0 but still once we have calculated r the surface integral involved in Eq. 3.40 is difﬁcult. We have another clue. The point O is nearer to charge q as compared to point A in Fig. 3.10b. Hence the induced charge at point O must be more as compared to point A (i.e.,) as we move away from point O the induced charge must decrease. Is this information sufﬁcient to know the actual form of r? The answer is no. We can write number of functions which satisfy the above requirement. For example all the functions r¼ r¼ r¼

k1 ðr 0

þ k2 Þ 2 k1

ðr0 þ k2 Þ5=2 k1 ð r 0 þ k1 Þ 3

with the constant k1, k2 > 0 satisfy the condition at point O ½r0 ¼ 0 the induced charge r is maximum and as we move away from O ðr0 increasesÞ the induced charge decreases. So without knowing the actual form of r (i.e.,) how the charges are distributed over the conducting plane, Eq. 3.41 cannot be used to calculate the potential. “In similar problems in which k; r; q are not known Coulomb’s law, Gauss’s law, potential formulation cannot be used to ﬁnd E.” For such problems new technique’s are required to calculate the electric ﬁeld. Two techniques have been developed

3.5 Method of Images

197

(i) Method of images (ii) Laplace’s equation. In this section we will see in detail about method of images and in later sections about Laplace’s equation method. Now we will calculate the electric ﬁeld at point P in Fig. 3.10a by method of images. First let us write the boundary conditions for Fig. 3.10a (i) V (x, y, 0) = 0 (i.e.,) the conducting plane is grounded and hence at z = 0, V=0 (ii) At inﬁnite point V ! 0 ði:e:; Þfor x2 þ y2 þ z2 d2

V ! 0:

As per uniqueness theorem for a given boundary condition there can be only one solution. So for the above boundary conditions for a point charge situated above the inﬁnite grounded conducting plane there can be only one solution. Now let us consider a completely different situation shown in Fig. 3.11. In Fig. 3.11 we have removed the grounded conducting plane at z = 0. Instead we have placed a –q charge at (0, 0, –d). The potential at point P1 in Fig. 3.11 is given by 2 V¼

3

1 6 q q 7 4qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ5 4p eo 2 2 x2 þ y2 þ ðz dÞ x2 þ y2 þ ðz þ dÞ

In Eq. 3.42 substituting z = 0 we see that

Fig. 3.11 Two point charges q and - q situated at d and - d respectively

ð3:42Þ

198

3 Electric Charges at Rest—Part II

2 V¼

3

1 6 q q 7 4qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ5 4p eo 2 2 2 2 2 2 x þ y þ ðdÞ x þ y þ ðdÞ

V¼0 Also in Eq. 3.42 we see that V ! 0 for x2 þ y2 þ z2 o d2 . Thus Eq. 3.42 which is the potential for the charges shown in Fig. 3.11 satisﬁes the boundary conditions for the charges shown in Fig. 3.10a, for z > 0. Hence by uniqueness theorem the only possible solution for Fig. 3.10a is Eq. 3.42. Clearly note down the steps we have carried out. For Fig. 3.10a we constructed a image charge as shown in Fig. 3.11. We obtained the potential for the image charge in Fig. 3.11 for z > 0 in Eq. 3.42. Then we checked whether the potential is obtained in Eq. 3.42 for image (Fig. 3.11) satisﬁes the boundary conditions for the point charge situated above the inﬁnite grounded conducting plane (Fig. 3.10a) for the region z > 0. It does. Then as per uniqueness theorem the potential obtained for the image charge in Eq. 3.42 is the potential for the point charge situated above the inﬁnite grounded conducting plane. Thus we have used a indirect method for calculating potential for point P in Fig. 3.10a. We never used Coulomb’s law, Gauss’s law or potential formulation. Note down the important role played by uniqueness theorem. Without uniqueness theorem we cannot guarantee that potential obtained in Eq. 3.42 for image charge in Fig. 3.11 is a solution for Fig. 3.10a for region z > 0. What about z < 0. In Fig. 3.10a for z < 0 ﬁeld is zero. But for Fig. 3.11 for z < 0 ﬁeld is non zero. Our problem was to ﬁnd the ﬁeld for z > 0 in Fig. 3.10a so we need not bother for region z < 0. Once we have calculated the potential let us calculate other quantities like electric ﬁeld etc. (a) Electric ﬁeld at any point above the conductor z > 0 Once the potential has been found it is very easy to calculate the electric ﬁeld E ¼ rV @V ^ @V ^ @V ^ E¼ k i j @x @x @z

ð3:43Þ

^ E ¼ Ex^i þ Ey^j þ Ez k

ð3:44Þ

Hence we have Ex ¼

@V @x

3.5 Method of Images

199

From Eq. 3.42 then 2 Ex ¼

3

q @ 6 1 1 7 4qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ5 4peo @x 2 2 x2 þ y2 þ ðz dÞ x2 þ y2 þ ðz þ dÞ 2

Ex ¼

q 6 4h 4peo

ð3:45Þ

3 x x2

þ y2

þ ð z dÞ

2

i3=2 h

x x2

þ y2

þ ð z þ dÞ

2

7 i3=2 5

ð3:46Þ

Similarly 2 Ey ¼

@V q 6 ¼ 4h @y 4peo

3 y 7 i3=2 h i3=2 5 2 x 2 þ y 2 þ ð z dÞ x 2 þ y 2 þ ð z þ dÞ y

2

2 Ez ¼

ð3:47Þ

3

@V q 6 zd zþd 7 ¼ 4h i3=2 h i3=2 5 @z 4peo x 2 þ y 2 þ ð z dÞ 2 x 2 þ y 2 þ ð z þ dÞ 2

ð3:48Þ

The electric ﬁeld at point P in Fig. 3.10a is given ^ E ¼ Ex ^i þ Ey ^j þ Ez k where Ex, Ey, Ez are given by Eqs. 3.46, 3.47 and 3.48. (b) Electric ﬁeld on the conductor On the conductor z = 0. From Eqs. 3.46, 3.47, 3.48 we have " # q x x ðEx Þz¼0 ¼ 4peo x2 þ y2 þ d2 3=2 x2 þ y2 þ d2 3=2 ðEx Þz¼0 ¼ 0 Ey z¼0 ¼ 0 and ðEz Þz¼0 ¼

2qd

3=2 4peo x2 þ y2 þ d2

ð3:49Þ

ð3:50Þ

200

3 Electric Charges at Rest—Part II

Thus on the conductor ^ E ¼ ðEx Þz¼0 ^i þ Ey z¼0 ^j þ ðEz Þz¼0 k ^ as ðEx Þ ¼ Ey E ¼ ðEz Þz¼0 k ¼0 z¼0 z¼0 E¼

2q d

3=2 4peo x2 þ y2 þ d2

^ k

ð3:51Þ

(c) Surface charge density on the conductor The surface charge density on the conductor which we initially were not able to ﬁnd can now be calculated. At the conductor free space interface the electric ﬁeld is given by Eq. 2.304 Ein ¼

r eo

where Ein is the normal component of the electric ﬁeld. The normal component of the electric ﬁeld to the conductor is Ez component. Hence ðEz Þz¼0 ¼

r eo

ð3:52Þ

) r ¼ ðEz Þz¼0 eo

ð3:53Þ

From Eq. 3.50 " r ¼ eo

2q d

3=2 4peo x2 þ y2 þ d2

)r¼

2q d

3=2 2 4p x þ y2 þ d2

# ð3:54Þ

ð3:55Þ

For problems which we have done in Chap. 2 we substituted the charge density in Eqs. 2.80–2.82 and then calculated the potential. However in this section we calculated the potential initially by indirect image method and from that we are now calculating the charge density because initially we don’t know what is the actual form of charge density. (d) Force of attraction between the point charge q and the grounded inﬁnite plane conductor Because the potential is same for the image charges in Fig. 3.11 and for the point charge above the grounded plane in Fig. 3.10a, so is the ﬁeld and hence the force.

3.5 Method of Images

201

The force of attraction between the +q charge and –q charge in Fig. 3.11 is F¼

1 qðqÞ ^ q2 ^ k ¼ k 4peo ð2dÞ2 16peo d2

ð3:56Þ

The negative sign indicates the force is attractive. (e) Work done in bringing the charge from 1 and placing it at d The work done in moving the charge q from 1 and placing it at point B in Fig. 3.10a is Zd W¼

F dl

ð3:57Þ

1

Zd

q2 ¼ 4peo W¼

dz 4 z2

ð3:58Þ

1 q2 4peo 4d

ð3:59Þ

1

(f) The total induced charge on the conducting plane We have r¼

2p

q d x2

The total induced charge is

þ y 2 þ d2 Z

qtot ¼ S

3=2

r ds0

ð3:60Þ

0

In the polar coordinates ðr0 Þ2 ¼ x2 þ y2 ; ds0 ¼ ðr0 Þðdr0 Þd/0 Z2p Z1 qtot ¼ 0

0

ðqdÞðr0 dr0 d/0 Þ h i3=2 2p ðr0 Þ2 þ d2

qtot ¼ q:

ð3:61Þ

202

3.6

3 Electric Charges at Rest—Part II

Point Charge Near a Grounded Conducting Sphere

Let a point charge q be placed near a grounded conducting sphere of radius R as shown in Fig. 3.12a. The distance between the center of the sphere O and the point A where the charge q is placed is d. The point charge will induce a charge on the sphere. We are interested in calculating the electric ﬁeld at point P. The electric ﬁeld at point P depends on point charge q and the induced charge. The actual form of induced charge on the sphere is not known and hence we go for solving the problem using method of images. Let us solve the problem in polar coordinates. First let us identify the boundary condition for the problem. One of the boundary conditions to the problem is that on the surface of the sphere, the potential is zero because the sphere is grounded. We will use the boundary condition to ﬁnd the magnitude and location of the image charge. In Fig. 3.12 suppose assume that the sphere is removed and a image charge q1 is placed at I at a distance from the center of the sphere O. By the boundary condition the potential at point B, C must be zero. At point B we have

Fig. 3.12 a A point charge q situated near a grounded conducting sphere. b, c Illustration to show point charge q and image charge q1

3.6 Point Charge Near a Grounded Conducting Sphere

203

q q1 þ ¼0 4peo ðABÞ 4peo ðIBÞ

ð3:62Þ

q q þ 1 ¼0 AB IB

ð3:63Þ

q q1 þ ¼0 dR Ra

ð3:64Þ

q q1 þ ¼0 4peo ðACÞ 4peo ðCIÞ

ð3:65Þ

At point C we have

q q þ 1 ¼0 AC CI

ð3:66Þ

q q þ 1 ¼0 Rþd Rþa

ð3:67Þ

) ) From Eq. 3.64

q1 ¼ q

Ra dR

ð3:68Þ

Substituting Eq. 3.68 in Eq. 3.67 q ð R aÞ q ¼0 Rþd ðR þ aÞðd RÞ )a¼

R2 d

ð3:69Þ ð3:70Þ

Substituting Eq. 3.70 in Eq. 3.68

q1 ¼

2 q R Rd ðd R Þ

) q1 ¼

q R d

ð3:71Þ ð3:72Þ

Equations 3.70, 3.72 gives the location and magnitude of the image charge respectively. 2 q R Thus a image charge of magnitude d and located at a distance of Rd from the center of the sphere satisﬁes the boundary condition of the grounded conducting sphere near a point charge.

204

3 Electric Charges at Rest—Part II

Fig. 3.13 Illustration to calculate the potential at point P due to charges q and q1

The grounded sphere along with the point charge q is shown in Fig. 3.12b. The image charge q1 along with charge q with the sphere removed is shown in Fig. 3.12c. (a) Potential at point P The potential at point P in Figs. 3.12c and 3.13 is V¼

1 hq qi þ 1 4peo AP IP

2 V¼

ð3:73Þ 3

7 1 6 q qR 6pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ7 4 5 2 4peo r2 þ d 2rd cos h d r2 þ R4 2R2 r cos h 2 d d

ð3:74Þ

The boundary conditions for Fig. 3.12a is (i) At the surface of the sphere the potential must be zero because the sphere is grounded i.e., at r = R, V = 0 (ii) As r ! 1 V ! 0: Both these boundary conditions are satisﬁed by Eq. 3.74. The boundary conditions are for Fig. 3.12b—a grounded sphere near a point charge. But the potential due to the two charges q and q1 in Fig. 3.13 at point P in Eq. 3.74 satisﬁes both the boundary conditions due to grounded sphere near the point charge in Fig. 3.12b. Hence by uniqueness theorem the potential at point P in Fig. 3.12b is given by Eq. 3.74. (b) Electric ﬁeld intensity at point P The electric ﬁeld at point P in Fig. 3.12a has two components Er ; Eh as shown in Fig. 3.13.

3.6 Point Charge Near a Grounded Conducting Sphere

205

Now Er ¼ 2

@V @r

ð3:75Þ 3

7 1 @6 q qR 6pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ7 4 5 4peo @r r2 þ d2 2rd cos h d r2 þ R2 2R2 r cos h 2 d d 2 3

R2 cos h qR r 1 6 qðr d cos hÞ d 7 Er ¼ 4 3=2 5 4 2 4peo r2 þ d2 2rd cos h 3=2 d r2 þ R2 2R d r cos h d ¼

ð3:76Þ

ð3:77Þ

The h component Eh is given by Eh ¼

1 @V r @h

93 8 > > > > =7 < 1 6 @ q q R 6 7 r ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ﬃ p ¼ 4 2 2 2 >5 4peo r @h > 2 > ; : r þ d 2rd cos h d r2 þ R2 2R r cos h> d d 2 3 2

Eh ¼

q 6 d sin h d sin h 7 h 2 4

i3=2 5 3=2 2 4peo r2 þ d 2rd cos h r2 d þ R2 2rd cos h R2

ð3:78Þ

ð3:79Þ

ð3:80Þ

(c) Electric ﬁeld at the surface of the sphere At the surface of the sphere r = R. From Eqs. 3.77, 3.80 " # q R d2 =R ðEr Þr¼R ¼ 4peo d2 þ R2 2Rd cos h 3=2

ð3:81Þ

ðEh Þr¼R ¼ 0

ð3:82Þ

(d) Surface density of the induced charge At the conductor, free space interface the electric ﬁeld is given by Eq. 2.304 Ein ¼

r eo

206

3 Electric Charges at Rest—Part II

where Ein is the normal component of electric ﬁeld. The normal component of the electric ﬁeld the sphere is Er component. ðEr Þr¼R ¼

r eo

) r ¼ ðEr Þr¼R eo

ð3:83Þ ð3:84Þ

From Eq. 3.81 r¼

q R 2 d2 4pR d2 þ R2 2Rd cos h 3=2

ð3:85Þ

(e) Force between the sphere and the point charge The force between the sphere of radius R and point charge q is same as that between image charge q1 and point charge q F¼

1 q q1 4peo ðIAÞ2

ð3:86Þ

2

But IA ¼ d Rd in Fig. 3.13. Substituting the values of q1 and IA in Eq. 3.86

q R 1 q d F¼

2 2 4peo d Rd

ð3:87Þ

1 q2 Rd 2 4peo d R2 2

ð3:88Þ

F¼

3.7

Laplace’s Equation—Separation of Variables

In the preceding sections we used method of images to calculate the electric ﬁeld for the given problem. However method of images requires isolated free charges near conducting boundaries to ﬁx a image charge. If the problem is such that there are no isolated free charges (for example see Fig. 3.14) method of images cannot be used. The potential in such situations can be obtained using Laplace’s equation. Laplace’s equation is a partial differential equation and can be solved by the method of separation of variables. The potential in Laplace’s equation is a function of three

3.7 Laplace’s Equation—Separation of Variables

207

Fig. 3.14 Three conducting plates maintained at speciﬁed potentials as shown in the ﬁgure

variables. To solve Laplace’s equations by separation of variables we express the potential in terms of products of functions, each function depending on only one variable. We will see the method of separation of variables in three different coordinate systems cartesian, spherical and cylindrical coordinate system. Another important point which we will be utilising in solving problems is that Laplace’s equation is linear. Suppose assume that V1, V2, V3, V4 … are solutions of Laplace’s equation then V ¼ a1 V1 þ a2 V2 þ a3 V3 þ is also a solution of Laplace’s equation (a1, a2, a3 … are constants). r2 V ¼ a1 r2 V1 þ a2 r2 V2 þ a3 r2 V3 . . .

ð3:89Þ

As V1, V2, V3 … are solutions to Laplace’s equation r2 V1 ¼ 0; r2 V2 ¼ 0; r2 V3 ¼ 0 . . . Hence r2 V ¼ a1 ð0Þ þ a2 ð0Þ þ a3 ð0Þ þ ) r2 V ¼ 0

ð3:90Þ

208

3.8

3 Electric Charges at Rest—Part II

Separation of Variables Laplace’s Equation in Cartesian Coordinates

We know that Laplace’s equation is given by r2 V ¼ 0 In cartesian coordinates @2V @2V @2V þ 2 þ 2 ¼0 @x2 @y @z

ð3:91Þ

We seek solutions for the above equation such that the potential V can be expressed as the product of three functions each function depending on only one variable i.e., V ¼ XðxÞ YðyÞ ZðzÞ

ð3:92Þ

Thus the functions X, Y, Z each depends on only one variable x, y, z respectively. Substituting Eq. 3.92 in Eq. 3.91 YZ

d2 X d2 Y d2 Z þ XZ þ XY ¼0 dx2 dy2 dz2

ð3:93Þ

Dividing throughout by XYZ 1 d2 X 1 d2 Y 1 d2 Z þ þ ¼0 X dx2 Y dy2 Z dz2

ð3:94Þ

2 The ﬁrst term in Eq. 3.94 X1 ddxX2 depends on only one variable x. It is independent 2 2 of y and z. The second and third terms Y1 ddyY2 ; Z1 ddzZ2 are independent of x and

depend on y and z alone respectively. Thus as we allow x to vary the second and third term in Eq. 3.94 must remain 2 unaltered because they are independent of x. But the ﬁrst term X1 @dxX2 ; depends on x. 2 2 Will the ﬁrst term X1 ddxX2 vary as we vary x. Suppose assume that X1 ddxX2 varies as 2 we vary x. If X1 ddxX2 alone varies as we vary x then Eq. 3.94 cannot be equal to zero.

d2 X must not vary. The term 1 ; d2 X X dx 2 dx 2 d2 X is constant as we vary x. Hence 1 d2 X must X dx2 dx 2

Thus even if X depends on x, as we vary x, must adjust themselves such that X1 be equal to a constant, say kx

1 X

3.8 Separation of Variables Laplace’s Equation in Cartesian Coordinates

1 d2 X ¼ kx X dx2

209

ð3:95Þ

Similar arguments holds good for second and third terms in Eq. 3.94 1 d2 Y ¼ ky Y dy2

ð3:96Þ

1 d2 Z ¼ kz Z dz2

ð3:97Þ

Substituting Eqs. 3.95, 3.96, 3.97 in Eq. 3.94 kx þ ky þ kz ¼ 0

ð3:98Þ

Thus instead of solving a partial differential equation in Eq. 3.91 to obtain V we now solve three ordinary differential equations 3.95, 3.96, 3.97 to obtain X, Y, Z. Once we have obtained X, Y, Z these values can be substituted in Eq. 3.92 to obtain V and hence ﬁnally electric ﬁeld by E ¼ r V:

3.9

Potential Between Two Grounded Semi Inﬁnite Parallel Electrodes Separated by a Plane Electrode Held by a Potential Vo

Consider two semi inﬁnite grounded parallel electrodes as shown in Fig. 3.14. The distance between the two electrodes is b. Another plane electrode lying at x = 0 is insulated from the semi inﬁnite parallel electrodes and is maintained at potential Vo. Because the plate at x = 0 is given a potential of Vo (by using a external power supply) it will have a surface charge densing r. We are now interested in ﬁnding the ﬁelds in the region between the three plates. Coulomb’s law, Gauss’s law and potential formulation cannot be used to ﬁnd the ﬁeld in the region between three electrodes because we don’t know the actual for of r: Method of images cannot be used here because there are no isolated free charges and hence ﬁxing a image is not possible. The solutions is then to solve Laplace’s equation r2 V ¼ 0

210

3 Electric Charges at Rest—Part II

From Eq. 3.94 1 d2 X 1 d2 Y þ ¼0 X dx2 Y dy2

ð3:99Þ

and the potential is independent of z. The potential V between the electrodes is given by V ¼ XðxÞ YðyÞ

ð3:100Þ

The boundary conditions for the potential V(x, y) is 9 ðiÞ Vðx; 0Þ ¼ 0 > > = ðiiÞ Vðx; bÞ ¼ 0 ðiiiÞ Vð0; yÞ ¼ Vo with 0\y\b > > ; ðivÞ Vð1; yÞ ¼ 0 with 0 \y\b

ð3:101Þ

Equation 3.99 can be de coupled into ordinary differential equations as we have done in Sect. 3.8. Then similar to Eqs. 3.95, 3.96 1 d2 X ¼ k1 X dx2

ð3:102Þ

1 d2 Y ¼ k2 Y dy2

ð3:103Þ

Substituting Eqs. 3.102, 3.103 in Eq. 3.99 k1 þ k 2 ¼ 0

ð3:104Þ

k1 ¼ k2 ¼ k2 k2 is a constant

ð3:105Þ

Substituting Eq. 3.105 in Eqs. 3.102 and 3.103 we get 1 d2 X ¼ k2 X dx2 )

d2 X ¼ k2 X dx2

ð3:106Þ

1 d2 Y ¼ k2 Y dy2 )

d2 Y ¼ k2 Y dy2

ð3:107Þ

3.9 Potential Between Two Grounded Semi Inﬁnite Parallel …

211

The solution for Eq. 3.106 is X ¼ A1 ekx þ B1 ekx

ð3:108Þ

Using the boundary condition Eq. 3.101(iv) Vð1; yÞ ¼ 0 0 ¼ A1 e1 þ B1 e1 0 ¼ A1 e1 Thus A1 ¼ 0

ð3:109Þ

X ¼ B1 ekx

ð3:110Þ

YðyÞ ¼ A2 sin ky þ B2 cos ky

ð3:111Þ

Substituting A1 ¼ 0 in Eq. 3.108

The solution for Eq. 3.107 is

Using boundary condition Eq. 3.101(i) in Eq. 3.111 0 ¼ A2 sin 0 þ B2 cosð0Þ 0 ¼ A2 ð0Þ þ B2 ð1Þ ) B2 ¼ 0

ð3:112Þ

Substituting B2 = 0 in Eq. 3.111 we get YðyÞ ¼ A2 sin ky

ð3:113Þ

Using boundary condition Eq. 3.101(ii) Vðx; bÞ ¼ 0 0 ¼ A2 sin kb Now assume A2 ¼ 0: If A2 ¼ 0 then by Eq. 3.113 Y = 0 and hence V ¼ XðxÞ YðyÞ ¼ 0: Thus the potential between three electrodes is zero if A2 ¼ 0 which is physically impossible. Hence the only option is sin kb ¼ 0 kb ¼ np np )k¼ b

ð3:114Þ n ¼ 1; 2; 3. . .

212

3 Electric Charges at Rest—Part II

We have excluded n = 0 because when n = 0 Y = 0 and V(x, y) = X(x) Y(y) = 0 everywhere in between the three electrodes which is physically impossible. Hence Eq. 3.113 is YðyÞ ¼ A2 sin

np y b

ð3:115Þ

Similarly Eq. 3.110 is X ¼ B1 enp x=b

ð3:116Þ

Substituting Eqs. 3.116, 3.115 in Eq. 3.100 we get np y b np y

Vðx; yÞ ¼ B1 A2 enp x=b sin

Vðx; yÞ ¼ D e np x=b sin

b

ð3:117Þ ð3:118Þ

where n = 1, 2, 3, 4 … and D = B1 A2 is a constant. Now the ﬁnal boundary condition to be satisﬁed is V(0, y) = Vo (boundary condition Eq. 3.101(iii)). Using this boundary condition in Eq. 3.118 Vo ¼ D sin

np y b

ð3:119Þ

Thus boundary condition Eq. 3.101(iii) demands that Vo to be of the form np y : But Vo is a constant. sin b In order to satisfy boundary condition Eq. 3.101(iii) we note that from Eq. 3.118 there are inﬁnite solutions possible because n = 1, 2, 3… and we have already seen in Eqs. 3.89 and 3.90 Laplace’s equation is linear. Thus each solution for each n can be summed up together so that boundary condition Eq. 3.101(iii) is satisﬁed. Thus similar to equation V ¼ a1 V1 þ a2 V2 þ a3 V3 . . . ¼

1 X

an V n

n¼1

in Sect. 3.8 we can write Eq. 3.118 as Vðx; yÞ ¼

1 X n¼1

Dn enp x=b sin

np y b

ð3:120Þ

The above equation satisﬁes the boundary condition Eq. 3.101(i), (ii), (iv). Now if we carefully select the values of Dn then the boundary condition Eq. 3.101(iii) also can be satisﬁed.

3.9 Potential Between Two Grounded Semi Inﬁnite Parallel …

213

Substituting the boundary condition Eq. 3.101(iii) in Eq. 3.120 Vo ¼

1 X

Dn sin

n¼1

Multiplying Eq. 3.121 by sin

np y b

ð3:121Þ

mpy and integrating between 0 and b b

Zb Vo sin

mp y dy b

0

¼

1 X n¼1

Zb But

sin

Zb Dn

ð3:122Þ np y mp y sin dy sin b b

0

np y mp y sin dy ¼ b b

0 b=2

if n 6¼ m if n ¼ m

ð3:123Þ

0

Thus substituting Eq. 3.123 in Eq. 3.122 Zb Vo sin

np y b dy ¼ Dn b 2

ð3:124Þ

0

Zb But Vo

( np y dy ¼ sin b

0

2bVo np 0

if n is odd if n is even

ð3:125Þ

Substituting Eq. 3.125 in Eq. 3.124 2bVo b ¼ Dn where n ¼ 1; 3; 5; 7. . . 2 np ( 4Vo p if n is odd ) Dn ¼ n 0 if n is even

ð3:126Þ ð3:127Þ

Substituting Eq. 3.127 in Eq. 3.120 Vðx; yÞ ¼

1 4Vo X 1 np x=b np y e with n odd sin b p n¼1 n

Equation 3.128 is the desired result.

ð3:128Þ

214

3.10

3 Electric Charges at Rest—Part II

Potential Between Two Grounded Conducting Electrodes Separated by Two Conducting Side Plates Maintained at Vo Potentials Vo and Vo

Two parallel plane conducting electrodes of width a are maintained at zero potential and are separated by two electrodes of width b at their sides as shown in Fig. 3.15. Both the side electrodes are maintained at potentials respectively. The boundary conditions to the problem are 9 ðiÞ Vðx; 0Þ ¼ 0 > > = ðiiÞ Vðx; bÞ ¼ 0 ðiiiÞ Vð0; yÞ ¼ Vo with 0\y\b > > ; ðivÞ Vða; yÞ ¼ Vo with 0\y\b

ð3:129Þ

As in Sect. 3.9 1 d2 X 1 d2 Y þ ¼0 X dx2 Y dy2 1 d2 X ¼ k1 ¼ k2 X dx2

ð3:130Þ

1 d2 Y ¼ k1 ¼ k2 Y dy2

ð3:131Þ

Fig. 3.15 Four conducting plates maintained at speciﬁed potentials as shown in the ﬁgure

3.10

Potential Between Two Grounded Conducting Electrodes …

215

Vðx; yÞ ¼ XðxÞ YðyÞ

ð3:132Þ

XðxÞ ¼ A1 ek x þ B1 ek x

ð3:133Þ

Y(y) ¼ A2 sin ky þ B2 cos ky

ð3:134Þ

As in Sect. 3.9 boundary condition Eq. 3.129(i) requires B2 to be zero. Hence ð3:135Þ

YðyÞ ¼ A2 sin ky Similar to Sect. 3.9 boundary condition ii requires k¼

np b

n ¼ 1; 2; 3 . . .

ð3:136Þ

Hence substituting Eq. 3.136 in Eqs. 3.135, 3.133 X(x) = A1 enp x=b þ B1 enp x=b YðyÞ ¼ A2 sin

ð3:137Þ

np y b

ð3:138Þ

Substituting Eqs. 3.137, 3.138 in Eq. 3.132 we get

Vðx; yÞ ¼ A1 enp x=b þ B1 enp x=b

A2 sin

npy b

npy ) Vðx; yÞ ¼ D enp x=b þ F enp x=b sin b

ð3:139Þ ð3:140Þ

where A1 A2 ¼ D; B1 A2 ¼ F are constants. Because Laplace’s equation is linear as in Sect. 3.9 we sum up Eq. 3.140 for all n Vðx; yÞ ¼

1 X n¼1

npy Dn enpx=b þ Fn enpx=b sin b

ð3:141Þ

Boundary condition Eq. 3.129(iii) requires that Vð0; yÞ ¼ Vo substituting this boundary condition in Eq. 3.141 Vo ¼

1 X n¼1

ðDn þ Fn Þ sin

npy b

ð3:142Þ

216

3 Electric Charges at Rest—Part II

Multiplying both sides by sin mbpy and integrating between 0 and b we get Rb mpy dy Vo sin b 0 ¼

1 Rb P n¼1 0

Zb ) Vo

ð3:143Þ

ðDn þ Fn Þ sin nbpy sin mbpy dy

sin

npy b dy ¼ ðDn þ Fn Þ b 2

ð3:144Þ

0

Zb As

npy dy ¼ Vo sin b

0

(

2bVo np 0

if n is odd if n in even

ð3:145Þ

Substituting Eq. 3.145 in Eq. 3.144 we get 2bVo b ¼ ðD n þ F n Þ 2 np 4Vo with n ¼ 1; 3; 5. . . ) Dn þ Fn ¼ np )

ð3:146Þ

Boundary condition Eq. 3.129(iv) requires Vða; yÞ ¼ Vo . Substituting this boundary condition in Eq. 3.141 we get Vo ¼

1 X n¼1

npy Dn enpa=b þ Fn enpa=b sin b

ð3:147Þ

Multiplying the above equation by a sin mbpy and integrating between 0 to b we get. Dn en pa=b þ Fn enpa=b ¼

4Vo with n ¼ 1; 3; 5. . . np

ð3:148Þ

Solving Eqs. 3.146, 3.148 we get 4Vo npa=b 1 enpa=b e Dn ¼ np 1 e2npa=b

ð3:149Þ

4Vo npa=b 1 enpa=b e Fn ¼ np 1 e2npa=b

ð3:150Þ

3.10

Potential Between Two Grounded Conducting Electrodes …

217

Equation 3.141 is the desired solution with Dn, Fn given by Eqs. 3.149, 3.150. Selection of solution for the differential equations in Sects. 3.9, 3.10 is made by examining the given problem. In Sect. 3.9, in Fig. 3.14 we see that as we move along y-axis at y = 0, y = b the potential needs to be zero. This condition is satisﬁed by sine function. Similarly as we move along x-axis from x = 0 we see that potential decreases gradually which is satisﬁed by exponential function, hence we selected k1 to be positive and k2 to be negative so that X(x) assumes exponential function, Y(y) assumes sine function.

3.11

Separation of Variables—Laplace’s Equation in Spherical Polar Coordinates

If the given problem involves spherical conductors then spherical polar coordinates is more appropriate for the calculation of electric ﬁeld. The Laplace’s equation for spherical polar coordinates is given by 2 1 @ 2 @V 1 @ @V 1 @ V r sin h þ 2 þ 2 2 ¼0 2 r @r @r r sin h @h @h r sin h @/2 Let us assume that V is independent of / then 1 @ 2 @V 1 @ @V r sin h þ 2 ¼0 r2 @r @r r sin h @h @h

ð3:151Þ

As in the previous problems we go for separation of variables V ¼ RðrÞHðhÞ

ð3:152Þ

Substituting Eq. 3.152 in Eq. 3.151 1 d 2 dR R d dH r sin h H 2 þ 2 ¼0 r dr dr r sin h dh dh d 2 dR R d dH r sin h )H þ ¼0 dr dr sin h dh dh

ð3:153Þ ð3:154Þ

Dividing Eq. 3.154 by V ¼ RðrÞHðhÞ 1 d 2 dR 1 d dH r sin h ) þ ¼0 R dr dr H sin h dh dh

ð3:155Þ

The ﬁrst term depends on R alone and the second term on H alone. So each must be separately equal to a constant.

218

3 Electric Charges at Rest—Part II

1 d 2 dR r ¼ k1 R dr dr 1 d dH sin h ¼ k2 H sin h dh dh

ð3:156Þ ð3:157Þ

From Eqs. 3.155, 3.156, 3.157 we have k1 þ k 2 ¼ 0

ð3:158Þ

) k1 ¼ k2 ¼ nðn þ 1Þ ðsayÞ

ð3:159Þ

Substituting Eq. 3.159 in Eqs. 3.156, 3.157 1 d 2 dR r ¼ nðn þ 1Þ R dr dr 1 d dH sin h ¼ nðn þ 1Þ H sin h dh dh The solutions for the above equations are Bn rn þ 1

ð3:160Þ

) HðhÞ ¼ Pn ðcos hÞ

ð3:161Þ

n 1 d where Pn ðcos hÞ ¼ n ðcos2 h 1Þn 2 n! dðcos hÞ

ð3:162Þ

RðrÞ ¼ An rn þ

Here Pn ðcos hÞ are known as Legendre polynamials. The potential V is given by V ¼ RðrÞH ðhÞ h i V ¼ An rn þ Bn rðn þ 1Þ Pn ðcos hÞ

ð3:163Þ

The most general solution is the linear combination of individual solutions V¼

1 h i X An rn þ Bn rðn þ 1Þ Pn ðcos hÞ

ð3:164Þ

n¼0

Example 3.4 Consider a uniformly charge disc with charge density r and radius R as shown in Fig. 3.16. Calculate the potential at any point P due to the disc.

3.11

Separation of Variables—Laplace’s Equation in Spherical Polar Coordinates

219

Fig. 3.16 A circular disc of radius R carrying a uniform surface charge density

Solution Let us calculate the potential at point P at a distance r from the center O of the disc. The potential at point P is 1 X Bn n V¼ An r þ n þ 1 Pn ðcos hÞ r n¼0

ð3:165Þ

There are two possible cases (i) r > R The boundary condition for the problem is V ! 0 as r ! 1 As r ! 1 the potential given by Eq. 3.165 tends to inﬁnity which is physically impossible. Hence the coefﬁcient An ¼ 0: Thus the potential is given by Vðr; hÞ ¼

X Bn Pn ðcos hÞ rn þ 1 n

Now let us evaluate Bn. We note that at h ¼ 0 Hence Eq. 3.166 is Vðr; 0Þ ¼

X Bn rn þ 1 n

ð3:166Þ

Pn ðcos 0Þ ¼ 1:

ð3:167Þ

220

3 Electric Charges at Rest—Part II

We have already seen in Sect. 2.15 Eq. 2.155 along z-axis (i.e.,) h ¼ 0 the potential of a uniformly charged disc is given by V¼

i ﬃ r hpﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ R 2 þ z2 z 2eo

ð3:168Þ

Let q be the total charge on the disc then q pR2

ð3:169Þ

i ﬃ q hpﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ R2 þ z2 z 2 2peo R

ð3:170Þ

r¼ Substituting Eq. 3.169 in Eq. 3.168 V¼

As we are measuring the potential along z-axis in Eq. 3.170 the point P in Fig. 3.16 lies along z-axis and hence r = z. Therefore Eq. 3.170 can be written as V¼

i q hpﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ R2 þ r2 r 2 2peo R

ð3:171Þ

Comparing Eqs. 3.167, 3.171 i X Bn q hpﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 2 2r ¼ þ r R rn þ 1 2peo R2 n

ð3:172Þ

" 1 # X Bn q R2 2 ) ¼ r þ r 1 þ 2 rn þ 1 2peo R2 r n

ð3:173Þ

Applying binomial theorem B0 B1 B2 þ 2 þ 3 þ r r" r !# 1 1 2 2 q 1 R2 R 2 2 ¼ r þ r 1 þ þ þ 2 r2 2! r2 2peo R2

ð3:174Þ

2 B0 B1 B2 q 1R 1 R4 1 R6 þ 2 þ 3 þ ¼ 3 þ ... ) 6r 16 r5 r r r 2peo R2 2 r

ð3:175Þ

)

Comparing the coefﬁcients of like powers of r on both sides q q R2 q R4 ; B2 ¼ ; B4 ¼ . . . and 4peo 16 peo 32peo B1 ¼ B3 ¼ B5 ¼ ¼ 0:

B0 ¼

3.11

Separation of Variables—Laplace’s Equation in Spherical Polar Coordinates

221

Substituting the coefﬁcients B in Eq. 3.165 q 1 R2 R4 P0 ðcos hÞ 3 P2 ðcos hÞ þ 5 P4 ðcos hÞ. . . Vðr; hÞ ¼ 4peo r 4r 8r

ð3:176Þ

The above equation gives the value of potential at any point P in Fig. 3.16 for r > R. (ii) r < R In Eq. 3.165 as r ! 0; V ! 1 which is physically impossible. Hence the coefﬁcients Bn ¼ 0: Thus the potential at r < R is given by X Vðr; hÞ ¼ An rn Pn ðcos hÞ ð3:177Þ n

Consider that point P in Fig. 3.16 lies along z-axis. Then h ¼ 0 and hence Pn ðcos 0Þ ¼ 1: Hence Eq. 3.177 is X Vðr; 0Þ ¼ An r n ð3:178Þ n

From Eq. 3.171 we have

" 1 # q r2 2 Vðr; 0Þ ¼ r þ R 1 þ 2 2peo R2 R

ð3:179Þ

Comparing Eqs. 3.178, 3.179 we have q R r 2 R r 4 r þ R þ þ ) A0 þ A1 r þ A2 r þ ¼ 2 R 8 R 2peo R2 ð3:180Þ 2

Comparing the coefﬁcients of like powers of r on both sides q q ; A1 ¼ 2peo R 2peo R2 q q A2 ¼ ; A4 ¼ ... 3 4peo R 10peo R3 and A3 ¼ A5 ¼ A7 ¼ ¼ 0: A0 ¼

Substituting the coefﬁcients A in Eq. 3.177 Vðr; hÞ ¼ A0 P0 ðcos hÞ þ A1 r P1 ðcos hÞ þ A2 r2 P2 ðcos hÞ þ

ð3:181Þ

222

3 Electric Charges at Rest—Part II

q 1 r r2 P1 ðcos hÞ þ Vðr; hÞ ¼ P2 ðcos hÞ þ 2peo R R2 2R3

ð3:182Þ

The above equation gives the value of potential at any point P in Fig. 3.16 for r < R. Example 3.5 ^ The Consider a conducting uncharged sphere placed in a uniform electric ﬁeld Eo k: induced positive and negative charges are shown in Fig. 3.17. Calculate the potential, dipole moment and electric ﬁeld outside the sphere. The radius of the sphere is R. Solution ^ the positive Once the conducting sphere is placed in the uniform electric ﬁeld Eo k charges are pushed in the direction of the ﬁeld and the negative charges opposite to the direction of the ﬁeld. Because of induced charge the uniform ﬁeld lines gets distorted as shown in the Fig. 3.17. Under electrostatic condition the sphere is an equipotential and we can take the value of equipotential to zero. Then the boundary conditions for the problem is ðiÞ V ¼ 0 for r ¼ R ðiiÞ V ¼ Eo z ¼ Eo r cos h for r R

ð3:183Þ

The second boundary condition is needed on the grounds that the distances far ^ and h has the usual meaning as in spherical away from the sphere the ﬁeld is Eo k polar coordinates.

Fig. 3.17 A conducting sphere placed in a uniform electric ﬁeld. The induced positive and negative charges are shown along with the electric ﬁeld lines in the ﬁgure

3.11

Separation of Variables—Laplace’s Equation in Spherical Polar Coordinates

223

We know that the potential is given by Vðr; hÞ ¼

X Bn An rn þ n þ 1 Pn ðcos hÞ r n

ð3:184Þ

Applying the boundary conditions Eq. 3.183(i) 0 ¼ An Rn þ

Bn Rn þ 1

) Bn ¼ An R2n þ 1

ð3:185Þ ð3:186Þ

Substituting Eq. 3.186 in Eq. 3.184 Vðr; hÞ ¼

X n

Vðr; hÞ ¼

X n

An R2n þ 1 An r n þ 1 Pn ðcos hÞ r

ð3:187Þ

R2n þ 1 An rn n þ 1 Pn ðcos hÞ r

ð3:188Þ

n

Applying the boundary condition Eq. 3.183(ii) and noting down that when r R the second term in Eq. 3.184 is negligible 1 X

An rn Pn ðcos hÞ ¼ Eo r cos h

ð3:189Þ

n¼0

) A0 P0 ðcos hÞ þ A1 r1 P1 ðcos hÞ þ A2 r2 P2 ðcos hÞ þ ¼ Eo r cos h

ð3:190Þ

Comparing the coefﬁcients of r A1 P1 ðcos hÞ ¼ Eo cos h ) A1 cos h ¼ Eo cos h

ð3:191Þ

) A1 ¼ Eo and A0 ¼ A2 ¼ A3 ¼ ¼ 0 Substituting the coefﬁcients A in Eq. 3.188 Vðr; hÞ ¼ Eo r cos h þ

Eo R3 cos h r2

ð3:192Þ

224

3 Electric Charges at Rest—Part II

The ﬁrst term corresponds to the external ﬁeld and hence the potential due to the induced charge is Vðr; hÞ ¼

E0 R3 cos h r2

ð3:193Þ

In spherical polar coordinates the electric ﬁeld is given by E ¼ rV ¼

dV 1 @V ^ ^r h dr r @h

ð3:194Þ

as there is no / dependence. Now the r component is given by @V R3 ¼ Eo 1 þ 2 3 cos h Er ¼ @r r

ð3:195Þ

and the h component is given by 1 @V R3 ¼ Eo 1 3 sin h Eh ¼ r @h r

ð3:196Þ

The electric ﬁeld outside the sphere due to the induced charge on the sphere is given by E ¼ Er ^r þ Eh ^ h

ð3:197Þ

where Er and Eh is given by Eqs. 3.195, 3.196 The induced surface charge is given by Eq. 2.304 Ein ¼

r eo

Here Ein ¼ ðEr Þr¼R : Hence ðEr Þr¼R ¼ ero ) rðhÞ ¼ eo ðEr Þr¼R

ð3:198Þ

rðhÞ ¼ eo ðEr Þr¼R ¼ 3eo E0 cos h

ð3:199Þ

From Eq. 3.195 we have

We know that the potential due to a dipole is given by Eq. 2.256

3.11

Separation of Variables—Laplace’s Equation in Spherical Polar Coordinates

V¼

1 p cos h 4peo r2

225

ð3:200Þ

Comparing Eqs. 3.200, 3.193 1 p cos h Eo R3 ¼ 3 cos h 4peo r2 r

ð3:201Þ

) p ¼ 4peo Eo R3

ð3:202Þ

is the dipole moment of the sphere. We have found the ﬁeld outside the sphere due to the induced charges on the sphere which is kept in the external electric ﬁeld, in Eq. 3.197. What about the ﬁeld inside the sphere. As explained in Sect. 2.24 the ﬁeld inside the sphere is zero. Example 3.6 Consider two conducting hemisphere’s placed one on the other and separated by a small insulating ring. The radius of the sphere is R. The top hemisphere is kept at potential Vo and the bottom hemisphere is kept at potential −Vo. Calculate the electrostatic potential inside and outside the sphere (Fig. 3.18). Solution We know that X Bn n V¼ An r þ n þ 1 Pn ðcos hÞ r n

ð3:203Þ

There are two possible cases Case (i) r < R Inside the sphere from Eq. 3.203 as r ! 0 V ! 1 which is physically impossible. Therefore the coefﬁcients Bn ¼ 0: Hence Vin ¼

X

An rn Pn ðcos hÞ

n

Let VS be the potential at the surface of the sphere

Fig. 3.18 Two conducting hemispheres placed one on the other and separated by a small insulating ring. The top and bottom hemispheres are maintained at potentials Vo and - Vo

ð3:204Þ

226

3 Electric Charges at Rest—Part II

Fig. 3.19 Two concentric cylindrical shells maintained at potentials Va and Vb. The radius of the inner cylinder is a and outer cylinder is b

VS ¼

X

An Rn Pn ðcos hÞ

ð3:205Þ

n

Now utilising the fact for Legendre polynamials

Zp Pn ðcos hÞ Pm ðcos hÞ sin h dh ¼ 0

0 2 2n þ 1

if n 6¼ m if n ¼ m

ð3:206Þ

Multiplying Eq. 3.205 by Pm ðcos hÞ sin h and integrating between 0 and p we get Zp VS Pm ðcos hÞ sin h dh ¼

X

Zp An R

n

0

n

Pn ðcos hÞPm ðcos hÞ sin h dh

ð3:207Þ

2 2n þ 1

ð3:208Þ

0

using Eq. 3.206 Zp VS Pn ðcos hÞ sin h dh ¼ An Rn 0

2 p 3 Z 2n þ 1 4 ) An ¼ VS Pn ðcos hÞ sin h dh5 2Rn

ð3:209Þ

0

2 ) An ¼

2n þ 1 6 4 2Rn

Zp=2

Zp Vo Pn ðcos hÞ sin h dh þ

0

3 7 ðVo ÞPn ðcos hÞ sin h dh5

p=2

ð3:210Þ and hence

3.11

Separation of Variables—Laplace’s Equation in Spherical Polar Coordinates

i Vo h p=2 ð cos hÞ0 ð cos hÞpp=2 ¼ 0 2 " p=2 2 p # Vo cos2 h cos h 3Vo þ ¼ A1 ¼ 2 2 R 2R 0 p=2

227

A0 ¼

. . . and so on:

Substituting the coefﬁcients An in Eq. 3.204 Vin ¼ Vo

3r 7 r3 P1 ðcos hÞ P3 ðcos hÞ þ 2R 8 R3

ð3:211Þ

Case (ii) r > R Outside the sphere by Eq. 3.203 as r ! 1 V ! 1 which is physically impossible. Hence An ¼ 0: Thus X Bn Pn ðcos hÞ rn þ 1 n

ð3:212Þ

X Bn P ðcos hÞ nþ1 n n R

ð3:213Þ

Vout ¼ On the surface of the sphere VS ¼

Multiplying the above equation by Pm ðcos hÞ sin h and integrating between 0 to p Zp VS Pm ðcos hÞ sin h dh ¼

XZ n

0

p

Bn Pn ðcos hÞ Pm ðcos hÞ sin h dh Rn þ 1

ð3:214Þ

Bn 2 Rn þ 1 2n þ 1

ð3:215Þ

0

using Eq. 3.206 Zp VS Pn ðcos hÞ sin h dh ¼ 0

2 p 3 Z 2n þ 1 n þ 1 4 ) Bn ¼ R VS Pn ðcos hÞ sin h dh5 2

ð3:216Þ

0

2 ) Bn ¼

2n þ 1 n þ 1 6 R 4 2

Zp=2

Zp Vo Pn ðcos hÞ sin h dh þ

0

3 7 ðVo ÞPn ðcos hÞ sin h dh5

p=2

ð3:217Þ

228

3 Electric Charges at Rest—Part II

so that B0 ¼

B1 ¼

3 2 Vo

R

2

i Vo R h p=2 ð cos hÞ0 þ ðcos hÞpp=2 ¼ 0 2

p=2 cos2 h 2 0

þ

p cos2 h 2 p=2

¼ 32 Vo R2 . . . and so on

Substituting the values of coefﬁcients B in Eq. 3.212 Vout

3.12

" # 3 R 2 7 R 4 ¼ Vo P1 ðcos hÞ P3 ðcos hÞ þ 2 r 8 r

ð3:218Þ

Separation of Variables—Laplace’s Equation in Cylindrical Coordinates

Let us assume that the problem is independent of z. In cylindrical coordinates then Laplace’s equation is given by (we have replaced rc by r for convenience) 1 @ @V 1 @2V r ¼0 þ 2 r @r @r r @ /2

ð3:219Þ

Now let us apply method of separation of variables Vðr; hÞ ¼ RðrÞUð/Þ

ð3:220Þ

Substituting Eq. 3.220 in Eq. 3.219 and dividing throughout by RðrÞU ð/Þ we get r d d RðrÞ 1 d2 Uð/Þ r ¼0 þ RðrÞ dr dr U ð/Þ d/2

ð3:221Þ

The ﬁrst term depends only on R(r) and the second term only depends only on U ð/Þ. Hence each term must be separately equal to a constant

where k is a constant.

r d d RðrÞ r ¼ k2 RðrÞ dr dr

ð3:222Þ

1 d2 U ð/Þ ¼ k2 U ð/Þ d/2

ð3:223Þ

3.12

Separation of Variables—Laplace’s Equation in Cylindrical Coordinates

229

The solutions for Eq. 3.223 is U ð/Þ ¼ Ak cos k/ þ Bk sin k/

ð3:224Þ

The potential should be a single valued function at a given point. This condition is fulﬁlled when Uð/ þ 2pÞ ¼ Uð/Þ. In Eq. 3.224 Uð/ þ 2pÞ ¼ Uð/Þ is satisﬁed only when k = 0, 1, 2, 3 … Now k = 0 in Eq. 3.224 gives Uð/Þ ¼ A0

ð3:225Þ

which is a constant. Substituting k = 0 in Eq. 3.223 d2 Uð/Þ ¼0 d/2

ð3:226Þ

Solving the above equation the solution is Uð/Þ ¼ a þ b/

ð3:227Þ

However the condition Uð/ þ 2pÞ ¼ Uð/Þ is not satisﬁed for the above solution and hence the term b/ is not acceptable. The solution for Eq. 3.222 is RðrÞ ¼ Ck rk þ Dk rk

ð3:228Þ

where we have already identiﬁed the values of k as 0, 1, 2, 3 … For the case of k = 0, R(r) = 0 from Eq. 3.225; substituting k = 0 in Eq. 3.222 d dRðrÞ r r ¼0 dr dr

ð3:229Þ

R ¼ a0 þ b0 ln r

ð3:230Þ

Solving

From Eqs. 3.230, 3.228 and 3.224 the solution for potential can be written as ½for k ¼ 0; and k ¼ 1; 2; 3. . . 1

X Vðr; /Þ ¼ a0 þ b0 ln r þ Ck rk þ Dk rk ½Ak cos k/ þ Bk sin k/ k¼1

ð3:231Þ

230

3 Electric Charges at Rest—Part II

Example 3.7 Consider an inﬁnitely long conducting cylinder of radius R placed in a uniform electric ﬁeld Eo . The z-axis of the cylinder is perpendicular to the ﬁeld Eo . The center of the cylinder coincides with the origin. Find the potential outside the cylinder. Solution Consider a long conducting cylinder with its centre line coinciding with z-axis. As per the given problem the ﬁeld Eo is perpendicular to the z-axis. Therefore let us assume the ﬁeld is along x-axis. The potential outside the cylinder is given by Eq. 3.231 Vðr; /Þ ¼ a0 þ b0 ln r þ

1

X Ck rk þ Dk rk ½Ak cos k/ þ Bk sin k/

ð3:232Þ

k¼1

The cylinder is an equipotential surface and we can set its potential to zero. The boundary conditions to the problem are ðiÞ V ¼ 0 at r ¼ R ðiiÞ V ¼ Eo x ¼ Eo r cos / at r R

ð3:233Þ

The second boundary condition is needed on the grounds that at distances far away from the cylinder the ﬁeld is Eo^i. As r ! 1 the potential should be ﬁnite. Hence b0 = 0 in Eq. 3.232. Also at r R the potential should contain only terms of cos / and no sin / components as per boundary conditions (ii) in Eq. 3.233. Hence Bk = 0 in Eq. 3.232. Therefore the potential in Eq. 3.232 is ðiÞ Vðr; /Þ ¼ a0 þ

1

X

Ck rk þ Dk rk Ak cos k/

k¼1

Let Ak Ck ¼ Fk ; Ak Dk ¼ Gk then Vðr; /Þ ¼ a0 þ

1

X Fk rk þ Gk rk cos k/

ð3:234Þ

k¼1

Now applying the ﬁrst boundary condition (i) V = 0 at r = R from Eq. 3.233 0 ¼ Fk rk þ Gk rk ) Gk ¼ Fk R2k

ð3:235Þ

3.12

Separation of Variables—Laplace’s Equation in Cylindrical Coordinates

231

Substituting for Gk in Eq. 3.234

R2k Vðr; /Þ ¼ a0 þ Fk r k cos k/ r k¼1 1 X

k

ð3:236Þ

Substituting boundary condition (ii) from Eq. 3.233 in Eq. 3.236 and noting 2k down that the term Rrk can be neglected for r R E0 r cos / ¼ a0 þ

1 X

Fk rk cos k/

ð3:237Þ

k¼1

¼ a0 þ F1 r cos / þ F2 r cos 2/ þ 2

Comparing the like coefﬁcients of R on both sides E0 ¼ F1 ; a0 ¼ F2 ¼ F3 ¼ ¼ 0

ð3:238Þ

Substituting the above values in Eq. 3.236 E0 R2 Vðr; /Þ ¼ E0 r þ cos / r

ð3:239Þ

is the required result. Example 3.8 Consider two concentric inﬁnite cylindrical shells maintained at potentials Va and Vb. The radii of inner shell is a and outer shell is b. Find the potential in between the cylindrical shells. Solution The potential depends on r alone. Hence 1 @ @V r ¼0 r @r @r )

ð3:240Þ

@V A ¼ @r r

ð3:241Þ

) V ¼ A loge r þ B

ð3:242Þ

where A and B are constants. The boundary conditions for the problem is at

ðiÞ r ¼ a ðiiÞ r ¼ b

V ¼ Va V ¼ Vb

ð3:243Þ

232

3 Electric Charges at Rest—Part II

Using boundary conditions Eq. 3.243(i), (ii) in Eq. 3.242 Va ¼ A loge a þ B

ð3:244Þ

Vb ¼ A loge b þ B

ð3:245Þ

Solving Eqs. 3.244, 3.245 Va Vb

loge ba

ð3:246Þ

ðVa loge b Vb loge aÞ

loge ba

ð3:247Þ

A¼

B¼

Substituting Eqs. 3.246, 3.247 in Eq. 3.242 V¼

3.13

Va Vb ðV log b V log aÞ

loge r a e b e a loge b loge ba

ð3:248Þ

Summary

In Chap. 1 we discussed in detail about vectors. In Sect. 1.1 we saw that vectors are difﬁcult to deal with as compared to scalars. In Chap. 2 our primary aim was to calculate the electric ﬁeld, for a given charge distribution. We discussed in detail about Coulomb’s law, Gauss’s law and potential formulation. Section 2.18 compares about this three methods and we hinted that the above three methods cannot be used to calculate the electric ﬁeld if the charge distribution is not known. In such situations we calculate the electric ﬁeld using method of images (or) Laplace’s equations. Section 3.5 discussed in detail about method of images. In Sect. 3.7 we saw that method of images cannot be used if there are no isolated free charges. In such cases potential can be obtained using Laplace’s equation by separation of variables. However the usefulness of Laplace’s equation in obtaining the potential depends on whether boundaries of the given problem coincides with the coordinate surfaces. In three dimensions if one coordinate is held constant and other coordinates are varied then the surface obtained is called coordinate surface. As an example consider Fig. 3.14. Consider the coordinate surface at y = b. Clearly it is a plane. One of the semi inﬁnite electrode kept at V = 0 potential, coincides with the coordinate surface at y = b. Similarly another semi inﬁnite electrode kept at V = 0 potential, coincides with the coordinate surface at y = 0. The electrode kept at potential V = Vo coincides with x = 0 coordinate surface.

3.13

Summary

233

Obtaining solution in the above case was easy as the boundaries coincide with coordinate surfaces. However if the boundaries are complex obtaining solution using Laplace’s equation will not be easy. In such situations we can obtain solutions using analytical methods which will not be discussed in this book.

3.14

Dielectrics

In Sect. 2.24 we saw that materials in general can be classiﬁed into conductors, semiconductors and insulators. We discussed in detail about conductors. In this following sections we will discuss in detail about insulators also known as dielectrics. In the case of conductors the valence electrons are free to move, however in the case of dielectrics the valence electrons are tightly bound to the atoms. When a conductor is placed in an external electric ﬁeld the electrons move in a direction opposite to the ﬁeld having a net positive charge on the other side of the conductor, as we have seen in Fig. 2.64. However when a dielectric is subjected to an electric ﬁeld the electrons are slightly displaced from their original position as they are tightly bound to the atoms. The above state of the dielectric is called polarization of dielectric under the action of the applied electric ﬁeld. If applied electric ﬁeld is very strong the dielectric will breakdown—it will lose its insulating property. In the following sections we will discuss about polarization and dielectric breakdown.

3.15

Dielectric in an Electric Field

A dielectric atom is shown in Fig. 3.20a. N is the nucleus surrounded by an electron cloud. The center of the electron cloud O coincides with the center of the nucleus. Suppose the dielectric atom is placed in the electric ﬁeld Eo. The electrons are attracted in the direction opposite to the ﬁeld as shown in Fig. 3.20b. The center of

Fig. 3.20 a In a given atom the center of the electron cloud O coincides with the center of the nucleus N. b Under the action of the external electric ﬁeld the electron cloud gets displaced. The center of the electron cloud O doesn’t coincide with the center of the nucleus N

234

3 Electric Charges at Rest—Part II

the electron cloud O is thus displaced from the center of the nucleus. The displacement however is restricted by strong restoring forces due to the change in the charge conﬁguration of the atom. Thus as we see in Fig. 3.20b the center O of the electron cloud and center of the nucleus are separated by a small distance and we know that two equal and opposite charges separated by a small distance form a electric dipole. We call this created dipole as “induced dipole” because the dipole is induced by the electric ﬁeld in the dielectric atom. The electric ﬁeld thus “polarizes” the dielectric atom leaving a net dipole moment. Thus if a dielectric is placed in the external electric ﬁeld all the atoms in the dielectric gets polarized. Applying an external ﬁeld creates an set of “induced electric dipoles” which in turn produce an own electric ﬁeld. Thus there are two distinct electric ﬁelds—one is the external electric ﬁeld which we applied and the another electric ﬁeld produced by the set of induced electric dipoles. In Sect. 3.17 we will calculate the electric ﬁeld due to this induced electric dipoles.

3.16

Polar and Non-Polar Molecules

In the previous sections we considered atoms. In Fig. 3.20a the center of positive and negative charges coincide with each other leaving no dipole moment. However if such an atom is placed in a electric ﬁeld as shown in Fig. 3.20b an induced dipole is produced. Now instead of atoms let us consider molecules. Molecules in general can be classiﬁed as polar and non-polar molecules. Consider the example, H2, N2 molecules. In this molecule the center of positive and negative charges coincide with each other leaving no net dipole moment. If such molecules are placed in external electric ﬁeld the centers of positive and negative charges get separated by a distance giving rise to induced dipole moment. These molecules are known as non-polar molecules. Other examples of non-polar molecules are Cl2, CCl4, CH4 etc. On the other hand in molecules like water shown in Fig. 3.21 the centers of positive and negative charges do not coincide with each other even in the absence of electric ﬁeld. That is they have permanent dipole moment. The dipole is built in the material. These molecules are called polar molecules. In a dielectric consisting

Fig. 3.21 A water molecule

3.16

Polar and Non-Polar Molecules

235

Fig. 3.22 A dielectric material is placed in an external electric ﬁeld. The net result will be a production of set of dipoles aligned in the direction of the ﬁeld

of polar molecules the dipoles are randomly oriented thereby the net dipole moment is zero. If a dielectric consisting of polar molecules is placed in a electric ﬁeld then the electric ﬁeld tries to rotate the “permanent dipoles” in the direction of the ﬁeld and align the dipoles in its direction. Thus either it be a non-polar dielectric or polar dielectric when placed in an external electric ﬁeld the net result a set of dipoles aligned in the direction of the electric ﬁeld, as shown in Fig. 3.22 and the entire dielectric gets “polarized”. We deﬁne polarization P as dipole moment per unit volume. Considering a small volume element ds0 P¼

p ds0

ð3:249Þ

where p is the dipole moment.

3.17

Potential Produced by the Polarized Dielectric

As we already noted in the case of dielectric placed in an electric ﬁeld there are two distinct electric ﬁelds. The “external” electric ﬁeld which caused the polarization and the electric ﬁeld produced by the induced or permanent dipoles. We will calculate the ﬁeld produced by the dipoles (i.e.,) due to polarization itself. The potential due to a single dipole is (Eq. 2.256) Vdip ¼

1 ^r p 4peo r2

ð3:250Þ

236

3 Electric Charges at Rest—Part II

(In Eq. 2.256 we have used ^r instead of ^r because in Fig. 2.48 the dipole lies in the origin. In general if the dipole doesn’t lie in the origin we use ^r instead of ^r as in Eq. 3.250). When a dielectric material is placed in an electric ﬁeld we have number of dipoles aligned in the direction of electric ﬁeld as shown in Fig. 3.22b. In order to obtain the potential produced by all the dipoles we integrate Eq. 3.250 over the entire volume s0 of the entire dielectric V¼

1 4peo

Z

^r p

r2

s0

ð3:251Þ

From Eq. 3.249 p ¼ P ds0

ð3:252Þ

Substituting Eq. 3.252 in Eq. 3.251 1 V¼ 4peo

Z s0

^r P 0 ds 2

r

ð3:253Þ

The above equation gives the potential produced by the dipoles in the entire volume s0 of the dielectric material at a point whose coordinates are (x, y, z) as shown in Fig. 3.23. The coordinate (x, y, z) and coordinate of volume element ds0 ; ðx0 ; y0 ; z0 Þ is with respect to some arbitrary origin. From Fig. 3.23

r 2 ¼ ð x x0 Þ 2 þ ð y y 0 Þ 2 þ ð z z 0 Þ 2

Fig. 3.23 Illustration to show the potential produced by polarization in the dielectric material at a point (x, y, z/Emphasis>)

ð3:254Þ

3.17

Potential Produced by the Polarized Dielectric

237

From the above equation r0

1

r

¼

^r

ð3:255Þ

r2

Substituting Eq. 3.255 in Eq. 3.253 V¼

1 4peo

Z

r0

1

s0

r

P ds0

ð3:256Þ

Now using vector identity 10 from Sect. 1.20 0

r

P

1

0

0

1

¼ r PþP r r r 1 P 1 ) P r0 ¼ r0 r0 P

r

r

r

ð3:257Þ

r

Substituting Eq. 3.257 in Eq. 3.256 V¼

1 4peo

Z

r0

P

r

s0

ds0

1 4peo

Z s0

1

r

r0 P ds0

Using divergence theorem I I 1 1 1 1 P ds0 ðr0 PÞds0 V¼ 4peo 4peo r r S0

)V¼

1 4peo

ð3:259Þ

s0

I S

ð3:258Þ

0

1

r

^Þds0 ðP n

1 4peo

I s0

1

r

ðr0 PÞ ds0

ð3:260Þ

^ is the unit vector in the direction of ds0 : Here n Suppose if we replace ^ rp ¼ P n

ð3:261Þ

qp ¼ r0 P

ð3:262Þ

then V¼

1 4peo

I S

0

rp ds0

r

þ

1 4peo

Z s0

qp ds0

r

ð3:263Þ

238

3 Electric Charges at Rest—Part II

The above equation shows that the potential due to the polarized dielectric in same as that of sum of the potential of ‘bound charges’ rp and qp . Instead of integrating over the entire volume s0 to ﬁnd the potential and electric ﬁeld produced by the dipoles as in Eq. 3.253 we can calculate the potential and hence electric ﬁeld of the entire dielectric by summing up the potential of rp and qp . The potential produced by rp and qp can be calculated by the methods we discussed in Chap. 2.

3.18

Bound Charges rp and qp

In Sect. 3.17 we have seen that the potential due to the polarized object can be calculated by calculating potential of the bound charges rp and qp . In this section we will see the physical interpretation of those charges. Let us redraw Fig. 3.22 in Fig. 3.24. We see that there is a set of aligned dipoles produced inside the dielectric because of the application of electric ﬁeld. At points A, B in Fig. 3.24a the positive and negative charges cancel out each other. Similarly at points like C, F positive and negative charges cancel out each other. Along the given line of dipoles the positive and negtive charges cancel out each other and the result is net negative charges on one side and net positive charge on the other side of the dielectric as shown in Fig. 3.24b. Thus in the case of “uniform” polarization the application of electric ﬁeld produces a surface charge density rp . If the polarization is “non uniform” we get volume bound charge density qp in addition of rp . With the above physical interpretation of the bound charges we will not derive expressions for these charges. Consider a imaginary surface S inside a polarized dielectric. Let an electric ﬁeld Eo be applied normal to the surface S. In the process of polarization for each charge q crossing the surface S a equal negative charge –q crosses the surface S in opposite

Fig. 3.24 a A dielectric material is placed in an external electric ﬁeld. The net result will be a production of set of dipoles aligned in the direction of the ﬁeld. Adjacent positive and negative charges cancel out. b In an uniformly polarized dielectric cancellation of adjacent positive and negative charges results in a surface charge density

3.18

Bound Charges rp and qp

239

Fig. 3.25 a In a polarized dielectric placed in a external electric ﬁeld equal number of positive and negative charges cross a imaginary surface S. b In general the applied electric ﬁeld can be inclined at an angle to the surface S

direction as shown in Fig. 3.25a. The positive charge +q and the negative charge –q move a distance of d=2 in opposite direction. Clearly the volume of the parallelopiped shown in Fig. 3.25a is S d. Suppose the electric ﬁeld Eo is inclined at an angle h as shown in Fig. 3.25b then the effective area shown by S to the Eo is S cos h (see Sect. 1.12). Hence the total volume is then ðS cos hÞd ¼ S d

ð3:264Þ

If the dielectric contains N molecules per unit volume then the total number of molecules within the volume S d is N(S d). If qnet is the net total charge that crosses the surface S then qnet ¼ NðS dÞq

ð3:265Þ

But qd is the dipole moment p. Hence qnet ¼ NðS pÞ

ð3:266Þ

As p is the dipole moment of single dipole, Np is the number of dipoles per unit volume and by the deﬁnition of polarization P ¼ Np: Hence qnet ¼ S P ¼ P S

ð3:267Þ

^S ) qnet ¼ P n

ð3:268Þ

^ is the unit vector normal to S as shown in Fig. 3.25b. The bound surface where n charge density is rp ¼

qnet ^ ¼Pn S

ð3:269Þ

240

3 Electric Charges at Rest—Part II

Fig. 3.26 A non - uniform polarization produces a volume bound charge density

which is what we guessed in Eqs. 3.261, 3.263 and proved in Eq. 3.269. Thus a surface bound charge rp appears when the polarization is uniform. However when the polarization is non uniform in addition to bound charges rp appearing at the surface a bound charge qp appears in the volume. Consider an imaginary volume s inside a dielectric whose surface is S as shown in Fig. 3.26. As positive charges are pushed out of the surface S an equal number of negative charges lie within volume s. From Eq. 3.267 the total charges that cross out of surface S in Fig. 3.26 is I P ds ð3:270Þ S

An equal amount of negative charge remains within volume s and this negative charge is given by I ð3:271Þ qnet ¼ P ds S

(Negative value of Eq. 3.270). Applying divergence theorem Z qnet ¼

r P ds

ð3:272Þ

s

From Eq. 2.22 Z qnet ¼

qp ds s

Comparing Eqs. 3.272, 3.273a

ð3:273aÞ

3.18

Bound Charges rp and qp

241

qp ¼ r P

ð3:273bÞ

In Sect. 3.17 we saw that potential produced by the polarized object can be calculated by mathematical ﬁctitious charges rp and qp . However in this section we see that rp and qp are real accumulation of charges in the dielectric.

3.19

Electric Displacement Vector and Gauss Law in Dielectrics

In chapter two we calculated electric ﬁeld for the given charge distribution by H using Gauss’s law. We constructed appropriate Gaussian surface, then evaluated E ds and ﬁnally obtained E. When constructing Gaussian surface we made sure that required charges are enclosed in the Gaussian surface. Whenever we apply Gauss law for dielectrics we must make sure that all charges including bound charges produced by polarization is also included in the Gaussian surface. Previously we noted there are two distinct electric ﬁelds in action in a dielectric. One the external electric ﬁeld which we applied and the other true electric ﬁeld produced due to polarization. The electric ﬁeld produced by the polarization is due to bound charges- rp and qp . Let all the charges responsible for external electric ﬁeld be represented by qf : Let us call qf , the free charges, in opposition to bound charges. qf the free charge might consist of some static charges inside the dielectric… etc., which are not due to bound charges. From Gauss law rE¼

q eo

In the case of dielectrics q ¼ q p þ qf

ð3:274Þ

r E ¼ qp þ qf =eo

ð3:275Þ

Thus

Substituting Eq. 3.273b in Eq. 3.275 r ðeo EÞ ¼ r P þ qf

ð3:276Þ

) r ðeo E þ PÞ ¼ qf

ð3:277Þ

The quantity D ¼ eo E þ P is known as electric displacement (or) electric flux density

242

3 Electric Charges at Rest—Part II

r D ¼ qf is the differential form of Gauss’s law in dielectrics. In integral form I D ds ¼ qf enc

ð3:278Þ

ð3:279Þ

S

Here qf enc is the free charges enclosed in the volume bounded by surface S. The above equation refers to only to free charges. Free charges are in our control and the electric ﬁeld produced by these free charges is responsible for bound charges. Once we know free charges, then we calculate D as we did in usual method of Gauss law in Chap. 2.

3.20

Linear Dielectrics

If the applied electric ﬁeld is not too strong then for most substances P ¼ eo ve E

ð3:280Þ

Here ve is called electric susceptibility of the material and the dielectrics which obey Eq. 3.280 are called linear dielectrics. Now

D ¼ eo E þ P

ð3:281Þ

For linear dielectrics D ¼ eo E þ eo ve E

ð3:282Þ

D ¼ eo ð1 þ ve ÞE

ð3:283Þ

Let e ¼ eo ð1 þ ve Þ

ð3:284Þ

D ¼ eE

ð3:285Þ

e ¼ ð1 þ v e Þ eo

ð3:286Þ

Then

and er ¼

er is called relative permittivity or dielectric constant of the material and e is called absolute permittivity or simply permittivity of the medium.

3.21

3.21

Dielectric Breakdown

243

Dielectric Breakdown

We saw that whenever a dielectric is subjected to an electric ﬁeld the dielectric gets polarized. The electrons are displaced from their original position but do not leave the atom. But however if the applied electric ﬁeld is very strong the electrons can be permanently removed from the atom and gets accelerated by the action of the applied electric ﬁeld. The accelerated electrons collide with other atoms in the dielectric thereby ionizing these atoms. The result is large currents and the dielectric starts conducting. The dielectric has lost its insulating property and the phenomena is called dielectric breakdown. The ﬁeld at which the dielectric breaks down is the maximum ﬁeld the dielectric can withstand and is known as dielectric strength of the material.

3.22

Boundary Conditions in the Presence of Dielectrics

The boundary conditions which we derived in Sect. 2.25 must be modiﬁed in the presence of dielectrics. In Figs. 2.66, 2.67 if the media 1, 2 are two dielectrics with different dielectric primitivities e1, e2 noting down that D1, D2 point in the same direction as E1 and E2 we have (i) For tangential component E1t ¼ E2t )

D1t D2t ¼ e1 e2

ð3:287Þ ð3:288Þ

as D ¼ e E (ii) For normal component For medium 1 and medium 2 in Fig. 2.66, the boundary conditions in terms of D is D1n D2n ¼ rf

ð3:289Þ

If medium 1, medium 2 are dielectrics with no free charge in between then D1n ¼ D2n

ð3:290Þ

) e1 E1n ¼ e2 E2n

ð3:291Þ

Example 3.9 Consider boundary between two dielectrics with permittivity e1 and e2. The electric ﬁeld makes an angle h1 and h2 with the normal as shown in Fig. 3.27. Assume there is no charge in the boundary. Show that

244

3 Electric Charges at Rest—Part II

Fig. 3.27 Boundary between two dielectrics with different permivitties

tan h1 ðer Þ1 ¼ tan h2 ðer Þ2 where er is the relative permittivity of the dielectric media. Solution Because there is no charges in the boundary from Eq. 3.290 D1n ¼ D2n From Fig. 3.27 D1n ¼ D1 cos h1

ð3:292Þ

D2n ¼ D2 cos h2

ð3:293Þ

) D1 cos h1 ¼ D2 cos h2

ð3:294Þ

) e1 E1 cos h1 ¼ e2 E2 cos h2

ð3:295Þ

As D1n ¼ D2n

From Eq. 3.287 E1t ¼ E2t From Fig. 3.28 E1t ¼ E1 sin h1

ð3:296Þ

3.22

Boundary Conditions in the Presence of Dielectrics

245

Fig. 3.28 A linear dielectric sphere of radius R placed in an external electric ﬁeld

E2t ¼ E2 sin h2

ð3:297Þ

E1 sin h1 ¼ E2 sin h2

ð3:298Þ

tan h1 e1 ðer Þ1 ¼ ¼ tan h2 e2 ðer Þ2

ð3:299Þ

As E1t ¼ E2t

From Eqs. 3.295, 3.298

Example 3.10 A flat linear dielectric surface whose relative permittivity is 5 is subjected to a electric ﬁeld with a electric flux density of 2 C per m2 and the dielectric is uniformly polarized. The electric ﬁeld is perpendicular to the slab. The volume of the dielectric is 1 m3. Calculate E, P, p for the dielectric. Solution D ¼ e E ¼ eo er E D 2 )E¼ ¼ eo er ð8:85 1012 Þ ð5Þ ) E ¼ 4:52 1010 V/m The polarisation D ¼ eo E þ P

246

3 Electric Charges at Rest—Part II

) P ¼ D eo E D ) P ¼ D eo eo er 1 )P¼D 1 er 1 C )P¼2 1 ) P ¼ 1:6 3 5 m The dipole moment

1 p ¼ P volume ¼ 2 1 ð1Þ 5 p ¼ 1:6 c m:

Example 3.11 Consider a linear dielectric sphere of radius R and dielectric constant er be placed in a electric ﬁeld Eo as shown in Fig. 3.28. Find the electric ﬁeld inside and outside the sphere. Also calculate the dipole moment and polarisation of the sphere. Solution When the dielectric sphere is placed in an electric ﬁeld the sphere becomes polarised and bound charges appear on the surface. Inside the sphere thus the electric ﬁeld will be reduced although not completely zero as in the case of conducting sphere (Example 3.5). The potential of the sphere is given by (Eq. 3.164) Vout ¼

X

An rn Pn ðcos hÞ þ

X Bn Pn ðcos hÞ rn þ 1 n

ð3:300Þ

Cn rn Pn ðcos hÞ þ

X Fn Pn ðcos hÞ rn þ 1 n

ð3:301Þ

n

Vin ¼

X n

The boundary conditions to be satisﬁed are 9 ðiÞ Vin ¼ Vout at r ¼ R > > > > ðiiÞ V ¼ Eo r cos h at r R > > > ðiiiÞ D1n ¼ D2n as there are no free charges on the surface = ) ðE 1n Þr¼R¼ er ðE2n Þr¼R >

> > @ Vin out > ) @V ¼ e > r > @r @ r > r¼R r¼R ; ðivÞ V must be finite as r ! 0

ð3:302Þ

3.22

Boundary Conditions in the Presence of Dielectrics

247

The ﬁrst boundary condition is required on the physical grounds that the discontinuous potential will produce a inﬁnite electric ﬁeld. Applying boundary condition Eq. 3.302(ii) in Eq. 3.300 and remembering that as r R (i.e.,) at large r the second term can be neglected X Vout ¼ An rn Pn ðcos hÞ X ð3:303Þ Eo r cos h ¼ An rn Pn ðcos hÞ n

) Eo r P1 ðcos hÞ ¼ A0 P0 ðcos hÞ þ A1 r P1 ðcos hÞ þ A1 r2 P2 ðcos hÞ þ ð3:304Þ Comparing the coefﬁcients of r on both sides A1 ¼ Eo r A0 ¼ A2 ¼ A3 ¼ ¼ 0

ð3:305Þ

Substituting Eq. 3.305 in Eq. 3.300 Vout ¼ Eo r P1 ðcos hÞ þ

X Bn Pn ðcos hÞ rn þ 1 n

ð3:306Þ

Applying boundary condition Eq. 3.302(iv) in Eq. 3.301 Fn ¼ 0 Hence Eq. 3.301 is Vin ¼

X

Cn rn Pn ðcos hÞ

ð3:307Þ

n

The boundary condition Eq. 3.302(i) requires Vout ðR; hÞ ¼ Vin ðR; hÞ From Eqs. 3.306, 3.307 B0 P1 ðcos hÞ þ B1 þ R R2 ¼ Co þ C1 RP1 ðcos hÞ þ

Eo R P1 ðcos hÞ þ

ð3:308Þ

From the boundary condition Eq. 3.302(iii) and Eqs. 3.306, 3.307 B0 2B1 3B2 þ 3 P1 ðcos hÞ þ 4 P2 ðcos hÞ þ R2 R R ¼ er ½C1 P1 ðcos hÞ 2C2 R P2 ðcos hÞ. . . ) Eo P1 ðcos hÞ þ

ð3:309Þ

248

3 Electric Charges at Rest—Part II

From Eqs. 3.308, 3.309 equating the coefﬁcients of Legendre polynamials 9 B0 ¼ C > > 0 > R > = B 1 Eo R þ 2 ¼ C1 R R > B2 ¼ C R2 > > 2 > ; R2 ... 9 B0 ¼ 0 > > 2 > R > = 2B1 Eo þ R ¼ er C1 > 3B2 > ¼ 2er C2 R > > ; R4 ...

ð3:310Þ

ð3:311Þ

From Eqs. 3.310, 3.311 9 Bo ¼ C0 ¼ 0 > > 3> > > B1 ¼ eerr 1 þ 1 Eo R > > = 2Eo C 1 ¼ er þ 2 > > B2 ¼ C2 ¼ 0 > > > > B3 ¼ C3 ¼ 0 > ; ... ... ...

ð3:312Þ

Substituting Eqs. 3.312 in Eqs. 3.306, 3.307 Vout ¼ Eo r cos h þ Vin ¼

er 1 Eo R3 cos h er þ 2 r2

3 Eo r cos h er þ 2

ð3:313Þ ð3:314Þ

The electric ﬁeld outside the sphere is given by @Vout er 1 2R3 ¼ Eo 1 þ ðEr Þout ¼ cos h er þ 1 r 3 @r

ð3:315Þ

1 @Vout er 1 R3 ¼ Eo 1 ðEh Þout ¼ sin h r @h er þ 1 r 3

ð3:316Þ

Field inside the sphere is given by @Vin ¼ ðEr Þin ¼ @r

3 Eo cos h er þ 2

ð3:317Þ

3.22

Boundary Conditions in the Presence of Dielectrics

1 @Vin ¼ ðEh Þin ¼ r @h

3 Eo sin h er þ 2

249

ð3:318Þ

Thus qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 3 2 2 Ein ¼ Er þ Eh ¼ Eo er þ 2

ð3:319Þ

The dipole moment can be calculated as follows. From Eq. 3.313 the potential produced by polarised charges is Vout ¼

er 1 Eo R3 cos h er þ 2 r2

ð3:320Þ

The term Eo r cos h in Eq. 3.313 is due to the external ﬁeld. We know that the potential due to the dipole is given by V¼

1 p cos h 4peo r2

ð3:321Þ

Comparing Eqs. 3.320, 3.321 1 p cos h er 1 Eo R3 ) ¼ cos h 4peo r2 er þ 1 r2 er 1 )p¼ Eo R3 4peo er þ 2

ð3:322Þ

Now the polarisation is

er 1 er þ 2

Eo R3 4peo

p ¼ volume 4=3pR3 er 1 P¼3 eo Eo er þ 2 P¼

ð3:323Þ

Equation 3.319 is

3 Ein ¼ Eo er þ 2 er 1 ) Ein ¼ Eo Eo er þ 2 P Ein ¼ Eo 3eo

ð3:324Þ

250

3 Electric Charges at Rest—Part II

Example 3.12 Consider an inﬁnite line charge which lies at the center of the cylindrical shell region as shown in the Fig. 3.29. The inﬁnite line charge carries a uniform line charge density k. The cylindrical shell region contains a dielectric and the inner and outer radii of the dielectric is Ri and Ro. The cylindrical shell region extends to inﬁnity. Calculate the electric ﬁeld in the regions (i) r < Ri (ii) Ri < r < Ro (iii) r > Ro Solution Case (i) r < Ri. In this region Gauss’s law reads I E ds ¼

qenc eo

Applying Gauss’s law in usual manner E¼

k ^r 2peo r

Case (ii) Ri < r < Ro. In this region Gauss’s law reads I D ds ¼ qf enc Constructing a cylindrical Gaussian surface within the dielectric region with radii r and length L as shown in Fig. 3.30 and noting down that the charge enclosed by the Gaussian surface is kL we get

Fig. 3.29 Illustration showing an inﬁnite line charge lying along the center of the cylindrical shell. The inﬁnite line charge carries an uniform line charge density. The region between the cylindrical shell contains a dielectric

3.22

Boundary Conditions in the Presence of Dielectrics

251

I Dds cos 0 ¼ kL I ) D ds ¼ kL ) D 2prL ¼ kL k ^r )D¼ 2pr Inside the dielectric the electric ﬁeld cannot be determined because we do not know the polarisation. Case (iii) For r > Ro. In this region Gauss law reads I D ds ¼

qf enc eo

Following as in case (ii) we get )D¼

k ^r 2pr

In this region r > Ro there is no polarisation and hence D ¼ eo E þ P ¼ eo E D k ^r )E¼ ¼ eo 2peo r Example 3.13 Repeat Example 3.12 but this time the dielectric is a linear dielectric with dielectric constant er and electric susceptibility ve. Also ﬁnd the polarisation in the dielectric.

Fig. 3.30 Construction of a Gaussian surface in the cylindrical shell region

252

3 Electric Charges at Rest—Part II

Solution Case (i) r < Ri. Following case (i) in Example 3.12 we get E¼

k ^r 2peo r

Case (ii) Ri < r < Ro. Following case (ii) in Example 3.12 we get D¼

k ^r 2pr

Because the dielectric is linear the electric ﬁeld can be calculated using the relation D ¼ eE )E¼

k ^r 2peo r

The polarisation is given by P ¼ D eo E k k eo ¼ 2pr 2per k h eo i 1 P¼ 2pr e k 1 1 )P¼ 2pr er k er 1 ) P¼ 2pr er Alternatively the relation P ¼ eo ve E can be directly used to calculate polarisation. Case (iii) r > Ro. Following case (iii) in Example 3.12 we get E¼

k ^r 2peo r

3.23

3.23

Capacitance and Capacitors

253

Capacitance and Capacitors

A capacitor is a device used for storing charges. It consists of two conductors separated by a dielectric medium. The conductors carry equal and opposite charges. The capacitance C of the capacitor is deﬁned by the relation C¼

q V

ð3:325Þ

Here q is the charge in coulomb in each conductor, V is the potential difference between the conductors due to equal and opposite charges on them. The unit of capacitance is farads.

3.24

Principle of a Capacitor

Consider two plates P, R as shown in Fig. 3.31a. The plate P is maintained at potential V and carries a charge q. Due to induction negative and positive charges get induced in plate R. However negative charges lie close to plate P and positive charges lie far away from plate P. While the negative charges tend to reduce the potential of plate P the positive charges tend to raise the potential of plate P. However as the induced positive charges on plate R are far away than the induced negative charges the effect of negative charges in lowering the potential of plate P than the effect of induced positive charges in raising the potential of plate P is more. Thus the net effect is the potential of plate P is slightly lowered. In order to bring back the plate P to potential V additional charges must be given. Suppose the plate R is earthed as shown in Fig. 3.31b. The positive charges get grounded and now only the negative induced charges remain on plate R. Thus the potential of plate P is further reduced. Additional charges must be given to retain the potential V of plate P. Thus the stored charge in P can be increased. In other words the capacity of plate P has been increased. The arrangement in Fig. 3.31b thus forms a capacitor.

Fig. 3.31 A parallel plate capacitor. a The plate P is maintained at potential V. Negative and positive charges are induced in plate R. b Illustration showing plate R being earthed. The positive charges are grounded

254

3 Electric Charges at Rest—Part II

3.24.1 Capacity of a Parallel Plate Capacitor Consider two parallel plates P and R separated by a distance d as shown in Fig. 3.32. Let the plate P carry a charge of q and plate R carry a charge of –q. Let the area of plates be A. Except in the edges the electric ﬁeld E is uniform between plates and is given by (Eq. 2.301) E1n ¼ E ¼

r eo

ð3:326Þ

and tangential component is zero from Eq. 2.305. In the above equation E is the ﬁeld between the plates, r is the surface charge density on the plates. Hence E¼

r q ¼ e o eo A

ð3:327Þ

The potential difference between the plates is equal to the work done in moving a unit positive charge form plate R to plate P ZP V¼

E dl

ð3:328Þ

R

From Fig. 3.32 dl ¼ dx and we have E dl ¼ E dl cos 180 ¼ E dl E dl ¼ Edx

Fig. 3.32 Illustration showing the direction of electric ﬁeld vector between the plates of the parallel plate capacitor

ð3:329Þ

3.24

Principle of a Capacitor

255

RP Now V ¼ E dl R

Z0 )V¼

E dx

ð3:330Þ

d

)V¼ þ

qd eo A

ð3:331Þ

The capacitance is given by C¼

q eo A ¼ V d

ð3:332Þ

If the space between the capacitor is ﬁlled with a dielectric of dielectric constant er than the capacitance is given by C¼

eo er A d

ð3:333Þ

In the above derivation we have neglected the fringing ﬁelds at the edges of the plates.

3.24.2 Capacity of a Parallel Plate Capacitor with Two Dielectrics Consider a parallel plate capacitor consisting of two dielectrics shown in Fig. 3.33. The distance between the two plates is d and the thickness of the dielectric is t. The potential difference between the two plates P and R is given by V¼

q q ðd tÞ þ t ðFrom Eq. 3:331Þ A eo A er eo

q The ﬁrst term A e ðd tÞ corresponds to air and second term corresponds to o dielectric. Hence V¼

q t d tþ A eo er

ð3:334Þ

256

3 Electric Charges at Rest—Part II

Fig. 3.33 A parallel plate capacitor ﬁlled with two dielectrics

The capacitance is given by C¼

q eo A ¼ V d t þ et

ð3:335Þ

r

3.25

Capacitance of a Spherical Capacitor

Consider a spherical capacitor consisting of two concentric spherical shells separated by air as shown in Fig. 3.34. Let R1, R2 be the radii of inner and outer shells respectively. The outer shell B is earthed. If a charge q is distributed over the outer surface of the inner shell A then an equal and opposite charge –q will be induced in the inner surface of the outer shell.

Fig. 3.34 A spherical capacitor consisting of two concentric spherical shells separated by air

3.25

Capacitance of a Spherical Capacitor

257

The electric ﬁeld between the spheres is given by E¼

1 q 4peo r2

ð3:336Þ

from Example 2.11. The potential difference between the spheres is the work done in moving a unit positive charge from B to A ZR1 VAB ¼

E dl

ð3:337Þ

R2

Now E dl ¼ E dl cos 180 ¼ E dl From Fig. 3.34 dl ¼ dr: Hence E dl ¼ Edl ¼ Edr

ð3:338Þ

Now ZR1 VAB ¼

ZR1 E dl ¼

R2

E dr

ð3:339Þ

R2

ZR1 ) VAB ¼

q dr 4peo r2

ð3:340Þ

R2

) VAB ¼

q R2 R1 4peo R1 R2

ð3:341Þ

Capacitance is given by q VAB

ð3:342Þ

4peo R1 R2 R2 R1

ð3:343Þ

C¼ C¼

If the air between two spheres is replaced by another medium of dielectric constant er then C¼

4per eo R1 R2 ðR2 R1 Þ

ð3:344Þ

258

3.26

3 Electric Charges at Rest—Part II

Capacitance of a Cylindrical Capacitor

The cylindrical capacitors consists of two conducting cylinders lying coaxially as shown in Fig. 3.35. The radius of inner and outer cylinders being R1 and R2 respectively. The outer cylinder is earthed and the length of capacitor is L. The inner cylinder A carries a charge +q and hence a negative charge –q is induced on the inner wall of the outer cylinder B (neglecting the fringing ﬁelds). Construct a cylindrical Gaussian surface of radius r and length L. Applying Gauss’s law to the Gaussian surface I E ds ¼

qenc eo

I )

E ds cos 0 ¼

ð3:345Þ qenc eo

ð3:346Þ

As E is constant over the Gaussian surface I )E

ds ¼

qenc eo

ð3:347Þ

q eo

ð3:348Þ

I From Fig. 3.35

ds ¼ 2prL; qenc ¼ q Eð2prLÞ ¼ )E¼

Fig. 3.35 A cylindrical capacitor consisting of two coaxial conducting cylinders. A Gaussian surface to calculate the electric ﬁeld is also shown in the ﬁgure

q 2peo r L

ð3:349Þ

3.26

Capacitance of a Cylindrical Capacitor

259

The potential difference between the two cylinders which is the work done in moving a unit positive charge from B to A is ZR1 VAB ¼

E dl

ð3:350Þ

R2

From Fig. 3.35 E dl ¼ Edl cos 180 ¼ Edl and dl ¼ dr: Hence E dl ¼ Edl ¼ Edr Thus ZR1 VAB ¼

ZR1 E dl ¼

R2

E dr

ð3:351Þ

R2

ZR1

q dr 2peo r L

ð3:352Þ

q R2 loge 2peo L R1

ð3:353Þ

VAB ¼ R2

VAB ¼ The capacitance is given by C¼

q 2peo L

¼ VAB log R2 e R1

ð3:354Þ

Eq. 3.354 is based on the assumption that charge q is uniformly distributed over the inner cylinder and the space between the two cylinders is ﬁlled with air. If the space between the two cylinders contains a medium of dielectric constant er then C¼

4per eo L

2 2 loge R R1

ð3:355Þ

260

3.27

3 Electric Charges at Rest—Part II

Capacitors in Parallel and Series

Let us calculate the equivalent capacity of a certain number of capacitors joined in either parallel or series. Equivalent capacitance means replacing a group of capacitors with a single capacitor such that the single capacitor has same charge for the same applied potential difference as that of the group of capacitors (Fig. 3.36). (i) Capacitors joined in series Consider N number of capacitors C1, C2, C3 … CN joined in series and voltage V is applied across the capacitor (Fig. 3.36). Let the left side of the plate C1 is given a charge of q. As explained in Sect. 3.24 a equal negative charge and positive charge is induced on the right side plate of C1. However the induced positive charge q on the right side plate of C1 is given to the left side plate, of C2. The entire process continues up to capacitor CN and hence each capacitor is charged to a charge q. Because the capacitors are connected in series the sum of potential differences V1, V2, V3 … VN across the capacitors C1, C2, C3 … CN is equal to the total voltage V. Now V1 ¼

q q q ; V2 ¼ . . .VN ¼ C1 C2 CN

V ¼ V1 þ V2 þ V2 þ þ VN 1 1 1 )V¼q þ þ þ C1 C2 CN

ð3:356Þ ð3:357Þ ð3:358Þ

C, the equivalent capacitor must carry a charge q if a voltage V is applied across it. )V¼

q C

Fig. 3.36 N number of capacitors connected in series

ð3:359Þ

3.27

Capacitors in Parallel and Series

261

Comparing Eqs. 3.358 and 3.359 q 1 1 1 ¼q þ þ þ C C1 C2 CN

ð3:360Þ

1 1 1 1 ¼ þ þ þ C C1 C2 CN

ð3:361Þ

N 1 X 1 ¼ C C i¼1 i

ð3:362Þ

)

)

(ii) Capacitors in parallel Consider N number of capacitors connected in parallel as shown in Fig. 3.37. Let a potential difference of V be applied across the capacitors. Each capacitor C1, C2, C3 … CN carries a charge of q1, q2, q3 … qN. “The potential difference between two points is independent of the path travelled (Sect. 2.21 Eq. 2.284). Thus if we select a path between points A, B in Fig. 3.37 either via C1 or C2 or C3 or … CN the potential difference is same. Hence all the capacitors are maintained at same potential difference V.” Therefore

Fig. 3.37 N number of capacitors joined in parallel

262

3 Electric Charges at Rest—Part II

q1 ¼ VC1 ; q2 ¼ VC2 ; q3 ¼ VC3 . . .qN ¼ VCN

ð3:363Þ

The total charge on the capacitors is qtot ¼ q1 þ q2 þ q3 þ þ qN

ð3:364Þ

qtot ¼ VC1 þ VC2 þ VC3 þ þ VCN

ð3:365Þ

qtot ¼ V½C1 þ C2 þ C3 þ þ CN

ð3:366Þ

C, the equivalent capacitor must carry a charge qtot if a voltage V is applied across it qtot ¼ VC

ð3:367Þ

VC ¼ V½C1 þ C2 þ C3 þ þ CN

ð3:368Þ

) C ¼ C1 þ C2 þ C3 þ þ CN

ð3:369Þ

Comparing Eqs. 3.366, 3.367

3.28

Energy Stored in a Capacitor

Consider a parallel plate capacitor holding equal and opposite charges whose plates are kept close to each other. A force of attraction will exist between opposite charges and if we are interested in moving the plates a ﬁnite distance away work must be done against this force of attraction. This work done is stored as electrical potential energy between the plates. At any instant of time let the capacitor be charged to a small charge dq by applying a potential difference of V. The small work done dW is dW ¼ Vdq ¼

q dq C

ð3:370Þ

If the capacitor is charged to a total charge of q then Z W¼

Zq dW ¼

qdq C

ð3:371Þ

0

)W¼

1 q2 2 C

ð3:372Þ

3.28

Energy Stored in a Capacitor

263

as q = CV 1 ) W ¼ C V2 2

ð3:373Þ

This work done is stored as the electrical potential energy of the system. For a parallel plate capacitor C¼

eo A ; V ¼ Ed d

Substituting the above values in Eq. 3.373 W¼

1 eo A 2 2 1 E d ¼ eo E2 Ad 2 d 2

ð3:374Þ

Energy stored per unit volume is then ¼

1 eo E2 2

ð3:375Þ

Example 3.14 A parallel plate capacitor ﬁlled with air has a capacitance of 0.01 nF. The distance of separation between the plates is 0.2 mm. What is the plate area? Solution For a parallel plate capacitor C¼

eo A d

Cd ð0:01Þ 109 0:2 103 ¼ eo 8:85 1012 4 2 A ¼ 2:25 10 m )A¼

Example 3.15 A parallel plate capacitor consists of air and the distance between the plates is 4 cm. A dielectric slab of thickness 2 cm and dielectric constant 5 is introduced between plates. The distance between the two plates is so charged that the capacity of the capacitor remains unchanged. Calculate the new distance between the plates.

264

3 Electric Charges at Rest—Part II

Solution In the absence of dielectric the capacity of the capacitor is given by C¼

eo A eo A ¼ d 0:04

In the presence of dielectric let the new distance be d0 then eo A eo A ¼ 0 C¼ 0 t d 0:02 þ d t þ er

0:02 5

Because the capacitance remains unchanged eo A eo A ¼ 0 0:04 d 0:02 þ

) d0 ¼ 5:6 cm

0:02 5

Example 3.16 Consider two spherical shells forming a spherical capacitor as shown in Fig. 3.34. The radius of the inner and outer sphere’s are R1 and R2 respectively. Consider a point P between the inner and outer spheres whose distance is r from the center of the sphere. The space between the two shells is ﬁlled with two dielectrics such that from R1 to r the dielectric constant is er1 and from r to b the dielectric constant is er2. Calculate the capacitance of the capacitor. The charge in the inner sphere is q. Solution From Eq. 3.339 ZR1 VAB ¼

ZR2 E dr ¼

R2

E dr R1

Zr ) VAB ¼

E dr þ R1

Zr

) VAB ) VAB

ZR2 E dr r

ZR2 q q ¼ dr þ dr 2 4peo er1 r 4peo er2 r2 r R1 q 1 1 1 1 1 1 ¼ þ 4peo er1 R1 r er2 r R2

3.28

Energy Stored in a Capacitor

265

Fig. 3.38 a, b Two parallel plate capacitors ﬁlled with two dielectrics in two different conﬁgurations

The capacitance is C¼

q VAB

C¼

h

1 er1

4pe i o h i 1 1 1 1 1 þ e r r r2 R1 R2

Example 3.17 A parallel plate capacitor is ﬁlled with two dielectrics as shown in Fig. 3.38a, b. The two dielectrics are of same dimensions with dielectric constants er1, er2 respectively. Find the capacitance. Solution In Fig. 3.38a the arrangement is simply two capacitors in parallel–each capacitor having area A/2. Hence total capacitance C ¼ C1 þ C2 Hence eo A=2 er1 eo A=2 e2 þ d d eo A )C¼ ð er1 þ er2 Þ 2d

C¼

266

3 Electric Charges at Rest—Part II

For Fig. 3.38b the arrangement can be regarded as two capacitors in series. Each capacitor will have a area A and separation distance d/2. Thus the total capacitance is 1 1 1 d 1 1 ¼ þ ¼ þ C C1 C2 2 eo A er1 eo A er2 : 2 eo A er1 er2 C¼ d er1 þ er2

Exercises Problem 3.1 Consider a equilateral triangle of side 12 cm as shown in Fig. 3.39. Three charges are situated at the corner’s of the triangle. Calculate the work done is assembling those charges. q = 2 nC. Problem 3.2 Consider a cube of side a. Eight charges each of magnitude q are placed at the eight corners of the cube. What is the electric potential energy stored in the system. Problem 3.3 A hollow sphere consists of a uniform surface charge r and the total charge on the sphere is q. Find the energy in the system if the radius of the sphere is R. Problem 3.4 A sphere whose radius is R carries a volume charge density q ¼ C r2 where C is a constant. Calculate the work done is assembling the charge distribution. Problem 3.5 Calculate the work done in moving a point charge q = 7 nC from (0, 0, 0) to ð1m; p=2; pÞ: The electric ﬁeld in spherical polar coordinates is given by E ¼ 5 r ^r þ Choose your own path. Fig. 3.39 Three charges situated at the corners of the equilateral triangle. Illustration for problem 3.1

7^ h r

Exercises

267

Fig. 3.40 Illustration for problem 3.6

Fig. 3.41 Illustration for problem 3.7

Problem 3.6 A point charge q is situated at distance a, b from two perpendicular half planes. The planes are grounded as shown in Fig. 3.40. Fix the image charge and calculate the electric ﬁeld in the region x, y > 0. Problem 3.7 In Fig. 3.41 ﬁnd the potential in the region z > 0. Problem 3.8 Find the potential between the plates of a parallel plate capacitor using Laplace’s equation. One of the plate is situated at x = 0 and the other at x = d along the x-axis. The plates are maintained at potential V1 and V2.

268

3 Electric Charges at Rest—Part II

Fig. 3.42 A conducting block having a deep narrow slot. Illustration for problem 3.9

Fig. 3.43 Illustration for problem 3.10

Problem 3.9 A conducting block which is grounded has a very deep narrow slot as shown in Fig. 3.42. The dimensions of the slot are such that b a. The slot is covered with a plate insulated from the slot and maintained at potential V ¼ Vo sin

npx a

Determine the potential inside the slot. Problem 3.10 In Fig. 3.43 the plate at z = 0 is maintained at zero potential, but the plate at z = a is maintained at potential V. Calculate the potential within the hollow cube.

Exercises

269

Fig. 3.44 A spherical shell. Illustration for problem 3.12

Problem 3.11 In Fig. 3.43 both the plates at z = 0, z = a are maintained at potential V. Calculate the potential within the hollow cube. Problem 3.12 Using Laplace’s equation calculate the potential between the plates of two spherical shells shown in Fig. 3.44. The inner shell is maintained at a potential of Va while the outer shell is maintained at a potential of Vb. The radii of inner and outer shells are R1 and R2 respectively and Va > Vb. Problem 3.13 Consider a charged conducting sphere holding charge q, radius R placed in a ^ By using Laplace’s equation calculate the electric ﬁeld uniform electric ﬁeld Eo k: inside and outside the sphere. Calculate the dipole moment, induced charge density of the sphere. Problem 3.14 Consider a circular ring of radius R carrying a total charge q. The total charge q is uniformly distributed over the ring. Calculate the potential at any point P as shown in Fig. 3.45 by using Laplace’s equation. Problem 3.15 Consider a hollow sphere of radius R. The potential on the sphere is given by VðhÞ ¼ 2 cos2 h=2: Calculate the potential inside and outside the sphere using Laplace’s equation.

½Hint : Use the relation 2 cos2 h=2 ¼ 1 þ cos h ¼ Po ðcos hÞ þ P1 ðcos hÞ

270

3 Electric Charges at Rest—Part II

Fig. 3.45 A circular ring of radius R. Illustration for problem 3.14

Problem 3.16 For Example 3.5 calculate the potential and ﬁeld inside the sphere by using Laplace’s equation. Problem 3.17 Consider two concentric sphere’s with radii R1, R2 as shown in Fig. 3.46. The lower hemisphere of the sphere with radius R2 and upper hemisphere of the sphere with radius R1 is maintained at potential Vo. The other hemisphere’s are maintained at zero potential. Calculate the potential within the region R1 r R2 . Problem 3.18 Consider a cylindrical shell of radius R as shown in Fig. 3.47. Find the potential at point P if the shell is maintained at potential Vo. The cylindrical shell is of inﬁnite extent.

Fig. 3.46 Two concentric spheres. Illustration for problem 3.17

Exercises

271

Fig. 3.47 A cylindrical shell of radius R. Illustration for problem 3.18

Fig. 3.48 A cylindrical tube of radius R. Illustration for problem 3.20

Problem 3.19 For Example 3.7 calculate the dipole moment, electric ﬁeld and surface charge density. Problem 3.20 A cylindrical tube of radius R with the upper half of the tube insulated from the lower half is shown in Fig. 3.48. The upper half is maintained at potential –Va and the lower half at Vb. Find the potential inside and outside the tube. Problem 3.21 Equations 2.11, 2.58 are E ¼ E1 þ E2 þ þ En Equation 2.77 is V ¼ V1 þ V2 þ þ Vn Show that for the case of work done W 6¼ W1 þ W2 þ þ Wn

272

3 Electric Charges at Rest—Part II

Problem 3.22 In Sect. 3.16 we hinted that whenever a dipole is placed in an electric ﬁeld, the electric ﬁeld tries to rotate the dipole in its direction. Calculate the torque acting on the dipole and the potential energy stored in the system. Problem 3.23 A dielectric sphere of radius R is uniformly polarized with polarization P. Find the bound charge densities, potential and ﬁeld of the sphere. Problem 3.24 A inﬁnitely long linear dielectric cylinder with dielectric constant er and radius R is placed in a uniform electric ﬁeld Eo. The ﬁeld Eo is perpendicular to the length of cylinder. Calculate the potential ﬁeld inside and outside the cylinder. Also ﬁnd the dipole moment of the cylinder, polarization and bound charge densities. Problem 3.25 Consider a parallel plate capacitor with air as medium. The plates are separated by 4 cm and the potential difference between them is 200 kV. Calculate the electric ﬁeld intensity and check whether the value is within the breakdown of air. What should be the minimum breakdown strength of an alternate dielectric so that the dielectric will not breakdown. Problem 3.26 (a) Calculate the capacitance of a parallel plate capacitor of area 400 cm2 and a separation distance of 6 mm. The dielectric medium is air. The potential difference between the plates is 400 V. Calculate the total energy stored in the capacitor and the energy density. (b) Calculate the equivalent capacitance for the conﬁguration shown in Fig. 3.49. Problem 3.27 One of the plates of a parallel plate capacitor is maintained at zero potential. The separation distance between the plates is d. The space between the plates is ﬁlled with a dielectric. The dielectric constant of the dielectric varies uniformly from one plate to the other. Calculate the capacitance of the capacitor if the values of the dielectric constant at the two plates is er1 and er2 respectively. The area of the plates is A.

Fig. 3.49 Illustration for problem 3.26

Exercises

273

Problem 3.28 For a parallel plate capacitor (with air as dielectric) a potential difference of 200 V is applied across the plates. A dielectric slab of thickness 0.7 cm and dielectric constant 6 is introduced between the plates. Calculate (i) Capacitance with and without dielectric. (ii) Free Charge. (iii) Electric ﬁeld in the air gap and dielectric.

A ¼ 200 cm2 d ¼ 1:5 cm: Problem 3.29 A cylindrical capacitor having inner and outer radii 4.5 and 6.5 cm consists of two dielectrics between them. The dielectric having contact with inner cylinder (radii 4.5 m) is having a dielectric constant of 5 and the other one is having a dielectric constant of 7. Calculate the capacitance per unit length. Problem 3.30 Consider a parallel plate capacitor whose capacitance is 0.5 lF. It is connected to the voltage source so that the voltage across the capacitor is 200 V. It is then connected to a second capacitor in parallel mode. The area of the plates of the second capacitor is 5 times that of the ﬁrst capacitor. Calculate the charge distribution on each capacitor. Also calculate the loss in energy.

Chapter 4

Magnetostatics

4.1

Introduction

In the previous chapters we calculated the electric ﬁeld produced by static charges. In this chapter we will see that the moving charges produce not only the electric ﬁeld but also the magnetic ﬁeld. In 1820 Oersted observed that whenever a compass needle is brought near a current-carrying wire the compass needle showed deflection. The direction of the deflection was reversed when the direction of the current in the wire was reversed. “We deﬁne the space surrounding the current-carrying conductor as containing magnetic ﬁeld”. As we represented electric ﬁelds by ﬁeld lines we represent magnetic ﬁeld by magnetic ﬁeld lines. Consider two current-carrying wires as shown in Fig. 4.1. The magnetic ﬁelds due to the current-carrying wires are shown in the ﬁgure. The magnetic lines of force are circular and are concentric. The direction of the magnetic lines of force can be obtained by right-hand rule as shown in the same ﬁgure. Let the thumb in the right hand point in the direction of the current and other ﬁngers of the hand curl around. The direction of the ﬁngers curling around gives the direction of the magnetic ﬁeld. We represent the cross section of a conductor (in which the current is flowing) on the paper by a circle. As shown in Fig. 4.2a the dot in the circle represents the current flowing out of the plane of the paper. The cross in the circle as shown in Fig. 4.2b represents the current flowing inside the plane of the paper. Our approach in magnetostatics will be similar to electrostatics. In the case of electrostatics we wrote the electric force as F ¼ QE. Similarly we will write an expression for force in magnetostatics. Then as we went in search for easier methods like Gauss’s law and potential in electrostatics, similarly we will go in search of easier methods in magnetostatics. Before discussing these methods we will see in detail about Lorentz force law. © Springer Nature Singapore Pte Ltd. 2020 S. Balaji, Electromagnetics Made Easy, https://doi.org/10.1007/978-981-15-2658-9_4

275

276

4 Magnetostatics

Fig. 4.1 a A wire carrying current in the upward direction. b A wire carrying current in the downward direction (The direction of the magnetic ﬁeld obtained by using right hand rule is depicted in the ﬁgure)

Fig. 4.2 Representation of current flow. a A dot along with circle represents current flowing out of the paper. b A cross along with the circle represents current flowing inside the paper (The direction of the magnetic flux density is also shown in the ﬁgure)

4.2

Lorentz Force Law

Suppose assume that a charge Q is moving, in a region of magnetic ﬁeld whose magnetic flux density is B. If the velocity of the charge is v then the magnetic force acting on the charge Q is given by Fmag ¼ Qðv BÞ

ð4:1Þ

The above equation is known as Lorentz force law. The total force acting on the charge Q in the presence of electric ﬁeld E and magnetic flux density is given by Ftot ¼ QðE þ ðv BÞÞ

ð4:2Þ

Equation 4.1 is a postulate and cannot be derived from any fundamental law. Its validity has been established through experiments. The charged particle trajectories in cyclotron, velocity selector, etc., is determined by Lorentz force equation.

4.2 Lorentz Force Law

277

Fig. 4.3 A charged particle moving in a circular path under the action of the magnetic ﬁeld

Suppose assume that a particle with charge Q is projected in a uniform magnetic ﬁeld with flux density B with velocity v as shown in Fig. 4.3. As Fmag ¼ Qðv BÞ, the magnetic force is given by cross product of v, B the force is perpendicular to the plane containing v, B. The particle moves in a circular path. Let us calculate the work done by magnetic force dW ¼ Fmag dl ¼ Qðv BÞ vdt dW ¼ 0

ð4:3Þ

as in a scalar triple product if two vectors are equal, the scalar triple product amounts to zero. Thus magnetic forces don’t do any work. Example 6.16 discusses more about this point.

4.3

Applications of Lorentz Force—Hall Effect

If a conductor (or) a semiconductor material carrying a current is placed in a magnetic ﬁeld that is acting perpendicular to the direction of the flow, a voltage is developed across the material in a direction perpendicular to both the direction of the current and that of the magnetic ﬁeld. The voltage developed is known as Hall voltage, and the phenomenon is called Hall effect. In Fig. 4.4 a metal or n-type semiconductor (Fig. 4.4a) and a p-type semiconductor (Fig. 4.4b) carrying current I is placed in a ﬁeld of flux density B such that B is perpendicular to I.

278

4 Magnetostatics

Fig. 4.4 a Figure showing a) a metal or n - type semiconductor b) a p - type semiconductor carrying a current I placed in a magnetic ﬁeld acting perpendicular to the direction of flow of current

In Fig. 4.4a we see that the electrons (majority carriers in n-type semiconductor) flowing with a drift velocity v experience a Lorentz force upward and collect upward on the top of the material leaving a net positive charge on the bottom of the material. In Fig. 4.4b we see that the holes (majority carriers in p-type semiconductor) flowing with a velocity v experience a Lorentz force upward and collect upward on the top of the material leaving a net negative charge on the bottom of the material. As shown in Fig. 4.4a, b the polarity of Hall voltage VH is different for n-type and p-type semiconductor and serves as a means to identify the type of the semiconductor.

4.3 Applications of Lorentz Force—Hall Effect

279

The Lorentz force acting on the particle of charge q is given by F ¼ qðv BÞ But from Fig. 4.4 B is perpendicular to v hence F ¼ qvB sin 90 ¼ qvB

ð4:4Þ

In Fig. 4.4a, b application of magnetic ﬁeld to a current-carrying conductor or semiconductor ðB?IÞ results in a separation of charges. Hence an electric ﬁeld EH is set up within the material. The separation of charges will continue until force due to EH balances the Lorentz force qEH ¼ qvB

ð4:5Þ

EH B

ð4:6Þ

)v¼

Let n be the number of atoms per unit volume of the material and A its area of cross section. Let q be the charge moving with velocity v. The current flowing through the conductor is given by charge flowing per second I ¼ ðnvAÞq

ð4:7Þ

But A = Lb where L, b are the dimensions of the face of the material as shown in Fig. 4.4. Substituting the value of A and Eq. 4.6 in Eq. 4.7 I¼n

EH ðLbÞq B

ð4:8Þ

IB nqLb

ð4:9Þ

EH ¼

The voltage VH established between the top and bottom surfaces of the material because of EH is VH ¼ EH L

ð4:10Þ

Substituting Eq. 4.9 in Eq. 4.10 ) VH ¼

IB IB L¼ nqLb nqb

ð4:11Þ

IB b

ð4:12Þ

VH ¼ RH

1 where RH ¼ nq is called Hall coefﬁcient and is characteristic of the material.

280

4 Magnetostatics

As mentioned previously Hall effect can be used to distinguish between n-type and p-type semiconductors and also ﬁnds applications in magnetic ﬁeld measurement.

4.4

Sources of Magnetic Field

A magnetic ﬁeld can be produced by a bar magnet or moving electrical charges (currents). For the time being we will not discuss the magnetic ﬁeld’s produced by a bar magnet. We will conﬁne our attention to the magnetic ﬁelds produced by the currents (i.e. moving electrical charges). In “electrostatics” we saw that “stationary charges” produce “electric ﬁelds” that are constant in time. Now we will deal with “magnetostatics” in which “steady currents” produce “magnetic ﬁelds” that are constant in time. Suppose assume a “steady” current I flows in a wire (situated in free space) as shown in Fig. 4.5. The element Idl0 is known as current element. We will be interested in calculating the magnetic ﬁeld produced by the current element and the entire wire at a far away point.

4.5

Magnetic Force Between Two Current Elements

Coulomb’s law gives an expression for the “electrical force” between two “electrical charges”. Similarly we can write an expression for “magnetic force” between two “current elements”. Suppose assume that there are two wires carrying current Ia and Ib as shown in Fig. 4.6. Consider two current elements Ia dl0a and Ib dl0b in the two wires. The distance between the two elements is r. The force between current elements Ia dl0a and Ib dl0b (i) varies directly as the product of magnitudes of current (ii) depends upon the nature of the medium

Fig. 4.5 A wire carrying a steady current I

4.5 Magnetic Force Between Two Current Elements

281

Fig. 4.6 Two wires a, b carrying currents Ia, Ib

(iii) varies inversely as the square of distance between the two current elements (iv) depends upon the lengths and orientations of the two current elements (v) is attractive if the current flow is in the same direction and repulsive if they flow in opposite direction. The force exerted by the current element Ia dl0a on Ib dl0b is 0 0 dlb dla ^r lo dFab ¼ Ia Ib r2 4p I I 0 0 dlb dla r lo Ia Ib Fab ¼ r3 4p a

ð4:13Þ ð4:14Þ

b

Here r is a vector pointing from current element Ia dl0a to Ib dl0b . The line integrals are evaluated over two wires. The constant lo is called permeability of free space and has a value of 4 p 107 N/A2.

4.6

Biot–Savart Law

Under electrostatic conditions the force between two charges q and Q separated by distance r is given by F¼

1 qQ ^r 4 p eo r 2

282

4 Magnetostatics

We have rewritten the above expression as F ¼ QE where E ¼

1 q ^r 4 peo r2

ð4:15Þ

is the electric ﬁeld produced at the site of Q by the charge q. As we wrote the electrostatic force F ¼ QE in terms of electric ﬁeld E, we will rewrite Eq. 4.14 in terms of magnetic flux density B. Equation 4.14 is I Fab ¼

l o 4p

Ib dl0b

b

I

Ia dl0a ^r

a

r2

ð4:16Þ

The above equation can be written as I ) Fab ¼

Ib dl0b Ba

ð4:17Þ

Ia dl0a ^r r2

ð4:18Þ

b

where Ba ¼

lo 4p

I a

Ba is the magnetic flux density. In general for Fig. 4.7 the magnetic flux density produced at point P is B¼

lo 4p

I

Idl0 ^r r2

ð4:19Þ

Equation 4.17 is the magnetic analogue of F = Q E, and Eq. 4.19 is the magnetic analogue of Eq. 2.6. B is a vector, and its direction can be obtained by right-hand rule as shown in Fig. 4.1. Equation 4.19 is called Biot–Savart law.

Fig. 4.7 Illustration to show calculation of magnetic flux density at point P for a wire carrying current I

4.7 Current Distributions

4.7

283

Current Distributions

In Figs. 4.5, 4.6, 4.7 and 4.8a we have shown currents which are distributed in a line, i.e. charge flows over a line. However charge can also flow over a surface or a volume giving rise to surface currents or volume currents. In Fig. 4.8b we show a surface carrying a current which flows at a velocity v. The surface current density denoted as K is deﬁned as current per unit width perpendicular to flow K¼

dI dlper

ð4:20Þ

where dlper is the width (perpendicular width to the flow) a small inﬁnitesimal element running parallel to the flow in Fig. 4.8b, and I is the current in the element.

Fig. 4.8 Figure depicting a) a current flowing along a line b) a current flowing along a surface c) a current flowing along a volume

284

4 Magnetostatics

If the surface charge density of the moving charges in Fig. 4.8b is r then K ¼ rv

ð4:21Þ

Suppose assume that the charges are moving in a volume as shown Fig. 4.8c with a velocity v. The volume current density denoted as J is deﬁned as current per unit area perpendicular to the flow J¼

dI dsper

ð4:22Þ

where dsper is a small area (perpendicular area to the flow) of a small inﬁnitesimal element running parallel to the flow in Fig. 4.8c, and dI is the current in the element. The volume charge density of the moving charges in Fig. 4.8c is q then J ¼ qv

ð4:23Þ

From Eq. 4.1 the force experienced by a small charge q in the presence of magnetic flux density B is (dq is small portion of the charge q) Z Fmag ¼

ðv BÞdq

ð4:24Þ

For a line charge dq ¼ kdl0 Z

ðv BÞkdl0

) Fmag ¼

ð4:25Þ

But current I ¼ charge flowing per unit time I ¼ kv

ð4:26Þ

Hence Z Fmag ¼

ðkv BÞdl0 Z

) Fmag ¼

ðI BÞdl0

For surface charge dq ¼ rds0 Z Z ) Fmag ¼ ðv BÞrds0 ¼ ðK BÞds0

ð4:27Þ ð4:28Þ

ð4:29Þ

4.7 Current Distributions

285

For volume charge dq ¼ qds0 Z Z Fmag ¼ ðv BÞqds0 ¼ ðJ BÞds0

ð4:30Þ

The Biot–Savart law for the line, surface and volume currents as shown in Fig. 4.8a, b, c can be written as BðrÞ ¼

lo 4p

Z l0

ðI ^rÞ 0 l dl ¼ o I 2 r 4p

BðrÞ ¼

BðrÞ ¼

lo 4p lo 4p

Z S0

Z s0

Z

dl0 ^r r2

ð4:31Þ

Kðr0 Þ ^r 0 ds r2

ð4:32Þ

Jðr0 Þ ^r 0 ds r2

ð4:33Þ

In Eq. 4.31 we have used the fact Idl0 ¼ Idll In Fig. 4.9a we see that the vectors I and dl0 point in the same direction. In Sect. 1.1b we have seen that two vectors are equal to each other, only if magnitude and direction of both the vectors are equal. The vector Idl0 points in the direction of I and Idl0 points in the direction of the dl0 . Because as shown in Fig. 4.9a I, dl0 point in same direction the vectors Idl0 , Idl0

Fig. 4.9 Figure depicting current elements and their directions for a) a line current density b) a surface current density

286

4 Magnetostatics

have same direction. Thus the magnitude and direction of Idl0 and Idl0 are same and hence Idl0 ¼ Idl0

ð4:34Þ

and we have used Eq. 4.34 in Eq. 4.31. Now let us consider Fig. 4.9b. The vector K in Fig. 4.9b is parallel to the surface and points along the surface. Hence the vector Kds0 having the same direction of K points along the surface. The vector ds0 is perpendicular to the surface (we have selected upward normal for the surface element ds0 for the sake of explanation). The vector Kds0 points in the direction of ds′ and is perpendicular to the surface. Thus the vector Kds0 is parallel to the surface, and the vector Kds0 is perpendicular to the surface. Hence, although vectors Kds0 , Kds0 have same magnitude but they have different directions and hence Kds0 6¼ Kds0

ð4:35Þ

This issue doesn’t arise with Jds0 because ds0 being volume element is a scalar quantity and doesn’t have direction. The vector Jds0 points in the direction of J. If there are collection of currents I1, I2, I3 … In each producing its flux density B1, B2, B3 … Bn then similar to Eq. 2.11 the net magnetic flux density B is given by B ¼ B1 þ B2 þ B3 þ þ Bn )B¼

n X

Bi

ð4:36Þ ð4:37Þ

i¼1

If one knows the current distributions one can substitute them in Eqs. 4.31, 4.32 and 4.33 and calculate B. However like Coulomb’s law, Biot–Savart law calculates B directly in terms of vector. As we have seen in Sect. 1.1a, Example 1.10, vectors are very difﬁcult to deal with. Hence like in electrostatics we will go in search of easier methods for calculating B. We will see that like Gauss’s law in electrostatics we will be able to calculate B easily using Ampere’s circuital law or Ampere’s law. Then as in electrostatics we will go in search of magnetic potential. Before discussing these methods we will calculate the B for an inﬁnitely long straight wire using Biot–Savart law.

4.8 Magnetic Flux Density Due to a Steady Current …

4.8

287

Magnetic Flux Density Due to a Steady Current in a Inﬁnitely Long Straight Wire

Consider an inﬁnitely long straight wire carrying a steady current I as shown in Fig. 4.10. We are interested in calculating the B at point P at a distance r from the element dl0 . The point P lies at a distance of r from the wire. Using Biot–Savart law dB, the flux density at point P due to the small element dl0 is lo I dl0 ^r dB ¼ 4p r2

ð4:38Þ

lo Idl0 sin h 4p r2

ð4:39Þ

) dB ¼

The B at point P due to the entire wire is given by Z B¼

l I dB ¼ o 4p

0 lZ ¼1

l0 ¼1

sin hdl0 r2

ð4:40Þ

from Fig. 4.10 h i1=2 2 r ¼ ðl 0 Þ þ r 2

ð4:41Þ

and r r sin h ¼ sinðp hÞ ¼ ¼ h i1=2 r ðl 0 Þ2 þ r 2

Fig. 4.10 An inﬁnitely long straight wire carrying a steady current I

ð4:42Þ

288

4 Magnetostatics

Substituting Eqs. 4.42 and 4.41 in Eq. 4.40 lI B¼ o 4p

Z1 h 1

r dl0 r2 þ ðl0 Þ2

i3=2

ð4:43Þ

From Fig. 4.10

When

l0 ¼ 1 l0 ¼ 1

l0 ¼ r tan a

ð4:44Þ

) dl0 ¼ r sec2 ada

ð4:45Þ

a ¼ p=2 a ¼ p=2

Hence lI B¼ o 4pr

Zp=2 cos a da

ð4:46Þ

p=2

)B¼

lo I 2pr

ð4:47Þ

is the magnetic flux density at point P due to the entire wire. The direction of B is given by right-hand rule as explained in Sect. 4.1. We have sketched the B around the wire in Fig. 4.11. In cylindrical coordinates B¼

Fig. 4.11 Lines of magnetic flux density around a wire carrying a steady current I

l0 I ^ / 2pr

ð4:48Þ

4.9 Ampere’s Circuital Law

4.9

289

Ampere’s Circuital Law

In this section we will discuss about Ampere’s circuital law (or) alternatively known as Ampere’s law. Suppose assume that an inﬁnitely long straight wire carries current as shown in Fig. 4.12a. B is sketched in the ﬁgure.

Fig. 4.12 a B of a inﬁnitely long straight wire. A circular path has been selected to prove Amperes law. b A non circular path has been selected to prove Amperes law

290

4 Magnetostatics

Let us consider a circular path of radius r and a small element dl as shown in Fig. 4.12a. Let us calculate the line integral of B around the circular path. Because B, dl, point in the same direction along the circular path I I B dl ¼ Bdl ð4:49Þ From Eq. 4.47 I

I Bdl ¼

lo I l I dl ¼ o 2pr 2pr

I B dl ¼

I ð4:50Þ

dl

lo I 2 p r ¼ lo I 2pr

ð4:51Þ

H where dl ¼ 2 p r is the circumference of the circular path. If the path selected is not circular then (Fig. 4.12b) from Eq. 4.48 I I lo I ^ ^ þ dz ^z B dl ¼ / drc rc þ rc d// 2pr

ð4:52Þ

As r = rc I

l I B dl ¼ o 2pr

Z2p

l I rc d/ ¼ o 2p

0

I B dl ¼

Z2p ð4:53Þ

d/ 0

lo I 2 p ¼ lo I 2p

ð4:54Þ

I B dl ¼ lo I

ð4:55Þ

In Eq. 4.52 we have used the form of dl from cylindrical coordinates. Suppose if the path does not enclose current as shown in Fig. 4.13a draw two lines Aa and Ab. Let the lines Aa, Ab cut the arcs cd and ab on the path. The angle between the lines Aa, Ab is d/. Let Ab = r1, Ad = r2 The line integral of B along the path abcd is I Z Z Z Z B dl ¼ B dl þ B dl þ B dl þ B dl ð4:56Þ abcd

ab

bc

cd

da

4.9 Ampere’s Circuital Law

291

Fig. 4.13 a A non - circular path that doesn't enclose current I. b, c, d The path abcd along with various vector components are shown in the ﬁgure

The path abcd along with dl is shown in Fig. 4.13b. Because B by Biot–Savart law is given by I ^r the direction of B with respect to line bc and da is shown in Fig. 4.13c, d. Thus the angle between dl and B along the path bc and da is 90°. Hence Z Z B dl ¼ B dl ¼ 0 ð4:57Þ bc

da

292

4 Magnetostatics

Substituting Eq. 4.57 in Eq. 4.56 Z Z Z B dl ¼ B dl þ B dl abcd

ab

ð4:58Þ

cd

In Fig. 4.13b we have resolved dl into perpendicular and parallel components ðdl? Þcd and dljj cd along the path cd. We see that ðdl? Þcd is perpendicular to B and hence B. ðdl? Þcd ¼ 0. Thus the contribution to the integral comes only from the dljj cd . Similarly along the path ab we have resolved dl into perpendicular and parallel components ðdl? Þab and dljj ab in Fig. 4.13b. We see that ðdl? Þab is perpendicular R to B and hence B. ðdl? Þab ¼ 0. Thus the contribution to the integral ab B dl comes from the component dljj ab . Hence Eq. 4.58 is Z

Z

B dljj ab þ

B dl ¼ abcd

As B dljj

cd

ab

Z

B dljj cd

ð4:59Þ

cd

have opposite directions in Fig. 4.13b Z

Z B dl ¼

abcd

B dljj ab

ab

Z

B dljj cd

ð4:60Þ

cd

Along the path ab B¼

lo I 2 p r1

ð4:61Þ

B¼

lo I 2 p r2

ð4:62Þ

and at path cd

Substituting Eqs. 4.61 and 4.62 in Eq. 4.60 Z

Z B dl ¼

abcd

lo I dljj ab 2pr1

ab

B dl ¼

) abcd

lo I dljj cd 2pr2

ð4:63Þ

cd

Z

Z

Z

ab

lo I dljj 2 p r1

Z ab

cd

lo I dljj cd 2p r2

ð4:64Þ

4.9 Ampere’s Circuital Law

293

From Fig. 4.13a

dljj ab dljj cd ¼ ¼ d/ r1 r2

ð4:65Þ

Hence Z

Z B dl ¼

abcd

lo I d/ 2p

Z

lo I d/ 2p

ð4:66Þ

Z B dl ¼ 0

ð4:67Þ

abcd

Thus Ampere’s law can be summarized as

I B dl ¼

lo I 0

if the path encloses current if the path does not enclose current

ð4:68Þ

Suppose as shown in Fig. 4.14 if the path encloses large number of currents then

Fig. 4.14 A non - circular path enclosing number of currents

294

4 Magnetostatics

I B d ¼ lo Ienc

ð4:69Þ

where Ienc is the total current enclosed by the entire path. From Eq. 4.22 ) dIenc ¼ Jdsper Z ) Ienc ¼ Jdsper

ð4:70Þ ð4:71Þ

From Sect. 1.12 dsper ¼ ds cos h. Hence Z Ienc ¼

J ds cos h

ð4:72Þ

Z Ienc ¼

J ds

ð4:73Þ

Substituting Eq. 4.73 in 4.69 I

Z B dl ¼ lo

J ds

Applying Stokes theorem Z

Z ðr BÞ ds ¼ lo

S

ð4:74Þ

J ds

ð4:75Þ

S

Because the above equation holds good for small element ds the integration can be dropped ðr BÞ ds ¼ lo J ds

ð4:76Þ

) r B ¼ lo J

ð4:77Þ

Equation 4.69 is the integral form of Ampere’s law while Eq. 4.77 is the differential form of Ampere’s law. Like Gauss’s law is used to calculate E easily in electrostatics similarly Ampere’s law is used to calculate B easily in magnetostatics. In Gauss’s law we construct a Gaussian surface over which E ds ¼ E ds cos h is evaluated easily.

4.9 Ampere’s Circuital Law

295

Once this has been done then after ensuring H E cos h is constant over the Gaussian surface we take E cos h out of the integral E ds and ﬁnally calculate E from the equation I q E ds ¼ enc eo Similarly in Ampere’s law we construct a Amperian loop such that over the Amperian loop B dl ¼ B dl cos h is easily evaluated. Also we select the Amperian loop such that over the Amperian loop B cos h is constant so that B cos h can be taken out of the integral I

I B dl ¼

I Bdl cos h ¼ B cos h

dl

ð4:78Þ

Once this has been done then B can be easily calculated using the integral I B dl ¼ lo Ienc

ð4:79Þ

I ) B cos h )B¼

dl ¼ lo Ienc lo Ienc H cos h dl

ð4:80Þ ð4:81Þ

In electrostatics we found that Gauss’s law is the easiest method, and successful application of Gauss’s law requires E ds to be easily evaluated over the Gaussian surface and E cos h to be constant over the Gaussian surface so that E cos h can be pulled out of the integral. For the given problem when the above conditions are not satisﬁed (e.g. see Sects. 2.14 and 2.15) we went in search of new methods like potential formulation to calculate the electric ﬁeld. Similarly in magnetostatics Ampere’s law is the easiest method, and successful application of Ampere’s law requires B dl to be easily evaluated and B cos h to be constant over the Amperian loop so that B cos h can be pulled out of the integral. For theHgiven problem when the above conditions are not satisﬁed it with be still true that B dl ¼ lo Ienc but the quantity which we require B is inside the integral and in such situations Ampere’s law cannot be easily used to calculate B (see Sect. 4.21). So we will go in search of new methods like potential formulation to calculate B. But before that we will calculate the divergence of B from Biot–Savart law directly.

296

4.10

4 Magnetostatics

Equation of Continuity

In a closed system it has been experimentally veriﬁed that the net amount of electric charge remains constant. Thus in a certain region, if the net charge decreases with time it implies that same quantity of charge appears in some other region. Suppose assume that a quantity of charge located within volume s decreases with time. The transport of charge constitutes a current dq dt

I=

ð4:82Þ

Negative sign indicates that charge contained in the speciﬁed volume decreases with time. From Eq. 4.73 I J ds

I=

ð4:83Þ

Let S be the surface bounded by volume s. As charge is conserved whatever charge is flowing out of surface S must come from the volume s. Hence from Eqs. 4.82 to 4.83 I dq ð4:84Þ J ds ¼ dt S

Applying Gauss’s divergence theorem and noting down that q ¼ Z r Jds ¼ s

d dt

r J ds ¼ s

Z ) s

q ds

q ds

ð4:85Þ

@q ds @t

ð4:86Þ

s

Z

s

s

Z

Z )

R

@q ds r Jþ @t

¼0

ð4:87Þ

The above equation must hold good for a small volume ds and hence dropping the integral r Jþ As ds 6¼ 0

@q ds ¼ 0 @t

ð4:88Þ

4.10

Equation of Continuity

297

r Jþ

@q ¼0 @t

ð4:89Þ

The above equation is known as equation of continuity and is a theoretical expression of the experimental fact that charge is conserved. In magnetostatics we deal with steady currents. Under steady currents charge doesn’t pile up anywhere and hence charge density is constant with time leading to @q ¼ 0. @t Hence from equation of continuity for magnetostatics rJ¼0 That is divergence of J for steady current is zero.

4.11

The Divergence of B

In this section we will derive an expression for divergence of B. That is we will calculate r B. A volume current density J (r') along with a volume element is shown in Fig. 4.15. The Biot–Savart law for volume current density J is given by Eq. 4.23 l BðrÞ ¼ o 4p

Fig. 4.15 A volume current density J(r') along with a volume element in Cartesian coordinate system

Z s

Jðr0 ) ^r 0 ds r2

ð4:90Þ

298

4 Magnetostatics

The above expression gives the B at point P(x, y, z), i.e. B (r) is a function of (x, y, z). Here (x, y, z) are called ﬁeld coordinates (see Sect. 2.4). As we are interested in calculating r B the divergence must be carried out with respect to (x, y, z) the unprimed ﬁeld coordinates. The volume s0 contains the current density and hence is a function of ðx0 ; y0 ; z0 Þ. Thus the integration in Eq. 4.90 must be carried out with respect to primed coordinates ðx0 ; y0 ; z0 Þ. Here ðx0 ; y0 ; z0 Þ coordinates are the primed source coordinates. Keeping the above points in the mind about “unprimed ﬁeld coordinates (x, y, z)” and “primed source coordinates ðx0 ; y0 ; z0 Þ” we will proceed to calculate r B. Using the results of Example 1.14 Eq. 4.90 can be written as B¼

lo 4p

Z s0

1 J r ds0 r

ð4:91Þ

1 r J r ds0 r

ð4:92Þ

Thus l rB¼ o 4p

Z s0

From Sect. 1.20 using the vector identity 8 with A ¼ J and B ¼ r

1 1 1 r Jr ¼r ðr JÞ J r r r r r

1 r

ð4:93Þ

But we have seen that J is a function of ðx0 ; y0 ; z0 Þ and in curl r J; r is carried out in unprimed coordinates x, y, z. As J is not a function of (x, y, z) we then have rJ¼0

ð4:94Þ

1 rr ¼0 r

ð4:95Þ

Also

as from Example 1.9 curl of a gradient of a scalar function is zero. Substituting Eqs. 4.94 and 4.95, in Eq. 4.93

1 r Jr ¼0 r

ð4:96Þ

4.11

The Divergence of B

299

and hence from Eq. 4.92 rB¼

lo 4p

1 r J r ds0 ¼ 0 r

Z s0

rB¼0

ð4:97Þ ð4:98Þ

And from Eq. 4.77 we know that r BðrÞ ¼ lo Jðr0 Þ

ð4:99Þ

We have speciﬁcally noted that B is the function of unprimed ﬁeld coordinate r and J is the function of primed source coordinate r′.

4.12

Magnetic Monopoles

(For better understanding reader is advised to go through Sects. 1.14 and 1.15 before continuing the section) Thus far we have derived four important expressions from electrostatics and magnetostatics. ðiÞ

rE¼

q eo

ð4:100aÞ

ðiiÞ

rE¼0

ð4:100bÞ

ðiiiÞ

rB¼0

ð4:100cÞ

ðivÞ

r B ¼ lo J

ð4:100dÞ

Recall the physical signiﬁcance of divergence and curl from Sects. 1.14 and 1.15 . Divergence is the measure of how much the given vector diverges or spreads out from the point in question. Curl is the measure of how much the vector rotates about the given point in question. From electrostatics we see that r E ¼ eqo and r E ¼ 0, i.e. in the presence of static charges r E 6¼ 0; r E ¼ 0. r E 6¼ 0 implies that the vector E diverges (or) spreads out and r E ¼ 0 implies that the electrostatic ﬁeld E doesn’t curl around. “Thus the electrostatic ﬁeld E diverges ðr E 6¼ 0Þ and doesn’t curl around ðr E ¼ 0Þ”. This has been shown pictorially in Fig. 4.16a. From point charge q the electrostatic ﬁeld E diverges or spreads out without any curl.

300

4 Magnetostatics

Fig. 4.16 a Under electrostatic conditions the electric ﬁeld diverges without any curl. bUnder magnetostatic conditions the magnetic flux density curls around without any divergence

From magnetostatics we see that r B ¼ 0 and r B ¼ l0 J, i.e. in the presence of steady currents r B ¼ 0 and r B 6¼ 0 r B ¼ 0 implies that the vector B doesn’t diverge or spread out. r B 6¼ 0 implies that the magnetic flux density B curls around (or) rotates around”. “Thus the B curls around (or) rotates around ðr B 6¼ 0Þ and doesn’t diverge r B ¼ 0”. This has been shown pictorially in 4.16b. From the steady current I the B curls around without any divergence. “Thus from four Eqs. 4.100a–4.100d we get an interesting result. Nature has created E and B such that E diverges and doesn’t curl around while B curls around and doesn’t diverge”. The above result is shown pictorially in Fig. 4.16. Assume a vector G (say). In Sect. 1.14 we saw that for a given vector G if r G [ 0 it means that there is a source from which the vector G emanates from, r G\0 means that there is a sink into which the vector G flows in and if r G ¼ 0 then G doesn’t have source or sink. Now considering electrostatic and magnetostatic cases we see that for E, B rE¼

q ) r E 6¼ 0 and e0 rB¼0

r E 6¼ 0 implies that there is a source or sink for E. For a positive charge r E [ 0. It means that a positive charge q is acting like a source from which the E emanates from. This is shown in Fig. 4.17a. For a negative charge r B ¼ 0. It means that a negative charge −q is acting like a sink into which E flows in. This is

4.12

Magnetic Monopoles

301

Fig. 4.17 a Point charge q acting like a source for electric ﬁeld vector. b Point charge -q acting like a sink for electric ﬁeld vector

shown pictorially in Fig. 4.17b (see Figs. 1.22 and 1.23 and discussion therein). Just pipe at point p, q in Fig. 1.23a, b acts as source and sink for water, similarly charges q, −q in Fig. 4.17a, b act as source and sink for E. “Thus in conclusion r E 6¼ 0 provides us with a very interesting result that “electric monopoles q(−q)” exist in nature from which (into which) the E emanates from (flows in)”. Considering equation r B ¼ 0 there is no source or sink from which or into which the vector B emanates from or flows into (see Figs. 1.22 and 1.23 and discussion therein). If we see Fig. 4.16b there is no starting point (or) ending point for B. B curls around I without any beginning or end point. The vectorial ﬁelds with no starting point or ending point are called solenoidal ﬁelds. Because there is no source or sink for B from which or into which the vector B emanates from or flows into, there are no magnetic monopoles. “Thus in conclusion the equation r B ¼ 0 provides us with a very interesting result-magnetic monopoles doesn’t exist in nature”. Take a bar magnet as shown in 4.18a with north pole and south pole. In an attempt to create magnetic monopoles if the magnet is broken then we get two bar magnets as shown in Fig. 4.18b instead of isolated north and south poles. Thus magnetic monopoles shown in Fig. 4.19 do not exist in nature. It is the circulation of electric charges which produces B and not the magnetic monopoles.

Fig. 4.18 a A bar magnet with north and south poles. b A broken bar magnet results in two bar magnets with no isolated north and south poles

302

4 Magnetostatics

Fig. 4.19 a A south pole. b A north pole

4.13

Magnetic Vector Potential

We hinted in Sect. 4.9 that Ampere’s law cannot be always used to calculate B. When Ampere’s circuital law cannot be used to calculate B then we need alternate methods to calculate B. In electrostatics we saw that when Gauss’s law cannot be used to calculate E then we can use potential V to calculate E½E ¼ rV, instead of using Coulomb’s law to calculate E because Coulomb’s law calculates E directly in vectorial form involving direction. V being a scalar quantity doesn’t involve direction. Similarly in magnetostatics when Ampere’s circuital law is not useful in calculating B then we are left with Biot–Savart law to calculate B which calculates B directly in vectorial form involving direction. Hence we require an easy method to calculate B without using Biot–Savart law when Ampere’s law is not useful. Using r E ¼ 0 in electrostatics we deﬁned an electric potential E ¼ rV (see Sect. 2.11). Similarly in magnetostatics rB¼0

ð4:101Þ

From Example 1.8 divergence curl of any vector is zero r ðr AÞ ¼ 0

ð4:102Þ

where A is a vector quantity. Comparing Eqs. 4.101 and 4.102 B¼rA

ð4:103Þ

In analogy to electrostatics we call A in Eq. 4.103 as “magnetic vector potential”.

4.13

Magnetic Vector Potential

303

Equation 4.91 is B¼

lo 4p

1 J r ds0 r

Z s0

ð4:104Þ

From Sect. 1.20 using vector identity 6 we see J 1 1 r ¼ r Jþr J r r r

ð4:105Þ

However as we previously noted the curl operation is being carried out with respect to ﬁeld coordinates (x, y, z) while J depends on source coordinates ðx0 ; y0 ; z0 Þ. Hence rJ¼0

ð4:106Þ

Substituting Eq. 4.106 in 4.105

J 1 r ¼r J r r

1 J ) Jr ¼r r r

ð4:107Þ ð4:108Þ

Substituting Eq. 4.108 in 4.103 l B¼ o 4p

Z s0

J 0 r ds r

ð4:109Þ

Because the integration is carried out with respect to source coordinates ðx0 ; y0 ; z0 Þ and curl is being carried out with respect to ﬁeld coordinates (x, y, z) both the operations can be interchanged.

l B¼r o 4p

Z

Jds0 ds0 r

ð4:110Þ

Comparing Eqs. 4.103 and 4.110 l A¼ o 4p

Z s0

Jds0 r

ð4:111Þ

304

4 Magnetostatics

For line and surface currents l A¼ o 4p

Z l0

Idl0 lo ¼ r 4p

Z l0

Idl0 r

ð4:112Þ

and A¼

lo 4p

Z S

0

Kdhboxs0 r

ð4:113Þ

A, the magnetic vector potential is not as much useful as electric potential V in electrostatics. Because A is still a vector we need to remain confused about components and directions. However the integration involved in Eqs. 4.101–4.103 is comparatively easier than the integrations in Biot–Savart law (4.31–4.33). Thus if we are interested in calculating B, ﬁrst we use Eqs. 4.101–4.103 to calculate A, then substitute A in Eq. 4.103 to calculate B. One of the interesting features of A is that it satisﬁes the Poisson’s equation. Equation 4.77 is r B ¼ lo J Substituting Eq. 4.103 in the above equation r r A ¼ lo J

ð4:114Þ

From Sect. 1.20 using vector identity 7 we see that rðr AÞ r2 A ¼ lo J

ð4:115Þ

We have seen in electrostatics that potential V doesn’t have signiﬁcance, only the electric ﬁeld E has signiﬁcance. The above fact we have seen in detail in Sect. 2.22 and Eqs. 2.285 and 2.286. For example altering the reference point in Fig. 2.55, Eqs. 2.285 and 2.286, changes the potential V, but doesn’t change the electric ﬁeld E. So the actual quantity of physical signiﬁcance is E and not the potential V. Similarly in magnetostatics the actual quantity of physical signiﬁcance is B not the vector potential A. The value of A can change without altering the value of B as explained below. From Eq. 4.103 B¼rA

ð4:116Þ

4.13

Magnetic Vector Potential

305

Because curl gradient of any scalar function is zero (Example 1.9) the above equation can be written as B ¼ r A þ r ru

ð4:117Þ

where u is a scalar function. That is by adding curl gradient of scalar function u to B, B remains unchanged. Equation 4.116 is B ¼ r ½A þ ru

ð4:118Þ

A1 ¼ A þ ru

ð4:119Þ

B ¼ r A1

ð4:120Þ

Let

Hence

Thus adding r ru to B in Eqs. 4.116 and 4.117 doesn’t alter B because r ru ¼ 0 but alters A as given by Eq. 4.119. This freedom can be exploited to make divergence of A to be zero. That is we keep on adding any number of ru's to A which will change the value of A such that rA¼0

ð4:121Þ

Or in other words we can select the value of A such that its divergence is zero by adding any number of ru’s, without altering the value of B. Substituting Eq. 4.121 in 4.115 r2 A ¼ l0 J

ð4:122Þ

Comparing the above equation with its electrostatic counterpart Poisson’s equation r2 V ¼

q eo

ð4:123Þ

We see that Eq. 4.122 is the Poisson’s equation in magnetostatics and A satisﬁes Poisson’s equation. Equation 4.121 is called the Coulomb gauge.

306

4.14

4 Magnetostatics

Magnetic Scalar Potential

As we have noted in Sect. 4.13, A, the magnetic vector potential is still a vector. We have already seen in Chaps. 1 and 2 that scalars are very easy to work with. So instead of magnetic vector potential A, it will be useful if we are able to ﬁnd a magnetic scalar potential Vm. Because the actual quantity which we are interested in calculating is B, which happens to be a vector, it will be easy to calculate Vm ﬁrst because it is a scalar and then calculate B. We know from Eq. 4.77 Ampere’s law is r B ¼ lo J In Fig. 4.8a, b, c, at points like P there is no current. In Fig. 4.8a current is restricted to a line, in Fig. 4.8b current is restricted to a plane, and in Fig. 4.8c current is restricted to a volume. Hence at point P in Fig. 4.8a, b, c, there is no current. Therefore at point P, the current-free region J ¼ 0. Hence in region where there is no current, Ampere’s law is rB¼0

ð4:124Þ

From Example 1.9 curl of a gradient of a scalar function is zero and hence we can deﬁne ) r ðlo rVm Þ ¼ 0

ð4:125Þ

Comparing Eqs. 4.124 and 4.125 we get B ¼ lo rVm

ð4:126Þ

for current-free regions. An important note must be made about Vm the magnetic scalar potential. It is not a single-valued function like electric scalar potential V. Once the zero of the reference point is ﬁxed electric scalar potential V has a unique or single value at a given point. To make the point clear let us calculate the magnetic scalar potential Vm of the current-carrying inﬁnitely long straight wire of Sect. 4.8. Integrating Eq. 4.126 1 Vm ¼ lo

Z B dl

With ^ þ dz^z dl ¼ drc^rc þ rc d//

ð4:127Þ

4.14

Magnetic Scalar Potential

307

Fig. 4.20 A circular path ACFG around a inﬁnitely long straight wire carrying current I

in cylindrical coordinates and from Eq. 4.48 we can write Vm ¼

1 lo

Z

lo I ^ ^ þ dz^z / drc^rc þ rd/ / 2pr Z I d/ ) Vm ¼ 2p ) Vm ¼

I / 2p

ð4:128Þ ð4:129Þ ð4:130Þ

Suppose assume that we traverse a circular path ACFG around the long straight wire as shown in Fig. 4.20. During the ﬁrst complete round or when we traverse path ACFG for the ﬁrst time starting from point A At point A angle / ¼ 0; hence Vm ¼ 0 At point C angle / ¼ p=2; hence Vm ¼ At point F; angle / ¼ p; hence Vm ¼

I 4

I 2

At point G angle / ¼ 3p=2; hence Vm ¼

3I 4

After completing the ﬁrst round, let us traverse ACFG for the second time, i.e. path ACFG is traversed for the second time At point A; angle / ¼ 2p; hence Vm ¼ I

308

4 Magnetostatics

At point C; angle / ¼

5p 5I ; hence Vm ¼ 2 4

At point F; angle / ¼ 3p; hence Vm ¼ At point G; angle / ¼

3I 2

7p 7I ; hence Vm ¼ 2 4

Thus Vm has multiple values at a given point. During the ﬁrst round the value of Vm at point C is I/4, but during second round the value of Vm at point C is 5I/4. And when we traverse the path for third time the value of Vm will change accordingly, and for each time we traverse the path the value of Vm changes at a given point. The reason for multi-valuedness Hof Vm can be explained as below. In the case of electrostatics r E ¼ 0 and E dl ¼ 0 (Eq. 2.280). The integral Zb Va Vb ¼

E dl a

where Va and Vb are potentials at points a and b are independent of the Hpath (Sect. 2.21). But in magnetostatics r B ¼ 0 (For current-free regions) and B dl ¼ lo I even if the path we selected is free of current. At the path ACFG in Fig. 4.20 there is no current, but still as we traverse the path for the second time the result of integration increases by I. For example at point C during the ﬁrst round the potential is I/4 and during the second round the potential is 5I/4. The difference I 5I in the potential is þ ¼ I. Similarly the difference between the potentials at 4 4 3I 7I point G is þ ¼ I. Hence each time we traverse the path the result of inte4 4 gration increases by I. This results in multiple values of Vm at a given point. Still we can manage to get a single value of Vm at a given point by restricting the path. For example in Fig. 4.21 if we traverse the circular path from A to A1 along Fig. 4.21 A circular path around a inﬁnitely long straight wire carrying current I which doesn't enclose the current

4.14

Magnetic Scalar Potential

309

the path ACFG A1 then we do not enclose current I and hence we get a single value of Vm. The magnetic potential difference between any two points a and b is given by (similar to Eq. 2.284 and from Eq. 4.127) 1 ðVm Þa ðVm Þb ¼ lo

4.15

Zb B dl

ð4:131Þ

a

Comments on Magnetic Vector Potential A and Magnetic Scalar Potential Vm

(Reader is advised to go through Sects. 2.21 and 2.22 before continuing this section). We have seen that the magnetic vector potential happens to be a vector A. So if we want to calculate B using the relation B ¼ r A then we need to remain confused about the components of A and the direction of A. Hence we went in search of magnetic scalar potential Vm and found one. But as we will see below Vm is not that much useful as electric potential V. In Fig. 4.22 we have redrawn Fig. 2.56b. In Fig. 4.22 we are interested in measuring the length of the thread A1 A2. The length of the thread is 5 cm. We have set the zero of the reference point as 2 cm. When we measure the length of the thread for the ﬁrst time we see that initial reading is 2 cm and ﬁnal reading is 7 cm. Hence the length of the thread is 7 cm 2 cm ¼ 5 cm. Suppose we repeat the measurement for the second time, that is we measure the length thread for the second time then once again initial reading is 2 cm and ﬁnal reading is 7 cm and hence the length of the thread is 7 cm 2 cm ¼ 5 cm. If we repeat the measurement for the third time then once again 7 cm 2 cm ¼ 5 cm is the length of the thread. “Thus the conclusion is once we ﬁx the zero of the reference point and make repeated measurements, the actual physical quantity the length of thread remains unchanged”. This is physically expected. The length of the thread cannot change

Fig. 4.22 Measuring the length of the thread using a scale by setting the reference point at 2 cm

310

4 Magnetostatics

Fig. 4.23 Moving from point a to point b along two paths Path 1 and Path 2

when we make repeated measurements. Now consider the electrostatic case. We have redrawn Fig. 2.25 in Fig. 4.23 along with reference point R. The potential difference between the points a and b is given by (Eq. 2.284) Zb Va Vb ¼

E dl

ð4:132Þ

a

We have seen in Sect. 2.21 Eq. 4.132 is independent of the path. That is we can travel from point a to b shown in Fig. 4.23 through path 1 or path 2 and the integral in Eq. 4.132 remains unchanged. Or in other words the potential Va and Vb of points a and b have a single or unique value irrespective of the path travelled. Because Va and Vb don’t depend on the path we get a single unique value of E for the given point from Eq. 4.132. “Thus the conclusion is once we ﬁx the zero of the reference point the single valuedness of electric potential V ensures that we get a single value of the actual physical quantity E at a given point”. This is physically expected. Just as the length of the thread cannot change when we make repeated measurements similarly E at a given point cannot change, when we travel along different paths. Now consider the magnetostatic case, Eq. 4.131 is 1 ðVm Þa ðVm Þb ¼ lo

Zb B dl a

4.15

Comments on Magnetic Vector Potential …

311

Let us rewrite the above equation for points C, G in Fig. 4.20 1 ðVm ÞC ðVm ÞG ¼ lo

ZG B dl

ð4:133Þ

C

During the ﬁrst round in Fig. 4.20, after substituting (Vm)C and (Vm)G in Eq. 4.133 and evaluating the integral we can get B. During the second round in Fig. 4.20 as explained in Sect. 4.14 (Vm)C and (Vm)G would have changed and substituting the new value of (Vm)C and (Vm)G in Eq. 4.133 we will get another value for B. Thus each time the path ACFG is traversed in Fig. 4.20 the value of Vm will keep changing. Because Vm changes its value, for each time the path is traversed, we will get new value of B for each time the path is traversed from Eq. 4.133 which is physically unacceptable. The actual physical quantity B must not change its value even if we make measurements after travelling the path each time. But because Vm is the path-dependent function each time we travel the path we get a new value of B which is not acceptable. In analogy to our thread example if we measure the length of the thread repeatedly number of times and ﬁrst time we get the length of the thread 5 cm, second time we get 27 cm, third time we get 101 cm, Is that acceptable? No. The length of the thread is 5 cm and any number of repeated measurements must give the same value for the length of the thread. Similarly we must get a single value of B (the actual physical quantity) irrespective of the path travelled. However Vm is a path-dependent function and hence we get different values of B for each path travelled which is physically unacceptable. V, the electric potential is a scalar quantity and is independent of path. E is a vector quantity. Because V is a scalar quantity and is independent of path ﬁrst we calculate V and then from V we calculate E by using the relation E ¼ rV. Vm the magnetic scalar potential is a scalar quantity, but is a path-dependent function. Because Vm is path-dependent we get different values of B for each path travelled which is not acceptable. Hence Vm rarely ﬁnds application in calculating B. As we have said in Fig. 4.21 we can restrict the path to avoid path dependency of Vm. A more detailed picture is shown in Fig. 4.24. Paths a1ba2, a1ca2 and a1da2 don’t enclose current I. We have restricted the path between a1 and a2. Hence for all these paths we get same value of B from Eq. 4.131. However the path a1ba2 ea1 encloses current I. Hence each time we travel the path a1ba2 ea1 the potential gets incremented by current I as explained in Sect. 4.14. Thus the potential will have different values at each given point each time the path a1ba2 ea1 is traversed. This will give rise to different values of B each time the path a1ba2 ea1 is traversed when Eq. 4.131 is used to calculate B along the path a1ba2 ea1.

312

4 Magnetostatics

Fig. 4.24 A inﬁnitely long straight wire carrying current I. Figure illustrates various paths around the wire. Paths a1ba2, a1ca2, a1da2 doesn't enclose the current. While the path a1ba2ea1 encloses the current

Another point to be noted is similar to electrostatics, in magnetostatics we cannot write the following relation between Vm and magnetic volume charge density qm Z lo qm 0 ds Vm ¼ 4p r s0

as magnetic monopoles do not exist. As electrostatic forces can do work we were able to deﬁne the potential difference between two points as the work done per unit charge. However magnetic forces do not do any work and hence similar physical interpretation of A or Vm is not possible. And A is still a vector and hence we need to remain confused about direction. The above said reasons restrict the usage of A and Vm. It must be said that A (or) Vm is not as much useful as electric scalar potential V. However as mentioned previously, as A reduces the labour involved in integration to calculate B, it will be used to calculate B for the given current distribution. H Because Eq. 4.132 is independent of path and E dl ¼ 0 (Eq. 2.280) E is called conservative ﬁeld. H However Eq. 4.131 is dependent on path and B dl ¼ lo I (Ampere’s law) and hence B is non-conservative ﬁeld.

4.16

4.16

B of a Current-Carrying Inﬁnitely Long Straight Conductor

313

B of a Current-Carrying Inﬁnitely Long Straight Conductor

(a) Using Biot–Savart law. The B of an inﬁnitely long straight conductor carrying current I has been calculated using Biot–Savart law in Sect. 4.8. (b) Using Ampere’s circuital law (or) Ampere’s law. Consider Fig. 4.25. By right-hand rule we know that B is circumferential. We can select dl such that B and dl have same direction. Because of symmetry, B is constant over the Amperian loop in Fig. 4.25 and hence applying Ampere’s law I B dl ¼ lo Ienc I Bdl cos 0 ¼ lo I

ð4:134Þ

I )

Bdl ¼ lo I

ð4:135Þ

Because B is constant over the Amperian loop in Fig. 4.25 I B

dl ¼ lo I

) B2 pr ¼ lo I )B¼

lo I 2 pr

ð4:136Þ ð4:137Þ ð4:138Þ

By Ampere’s law we calculated B in just few steps. Compare the above derivation with the method of calculating B using Biot–Savart law in Sect. 4.8.

Fig. 4.25 An Amperian loop for an inﬁnitely long straight wire carrying current I

314

4 Magnetostatics

Biot–Savart law is laborious method because it calculates B directly in terms of vector. However Ampere’s law calculates B easily by exploiting symmetry. More about this point is in Sect. 4.21.

4.17

B of a Current-Carrying Finite Straight Conductor

Consider a straight wire of length 2L carrying a current I. We are interested in calculating the B at point P at a distance r from the current element dl0 . The point P lies at a distance of r from the wire as shown in Fig. 4.26. Let us calculate B at point P using Biot–Savart law. The dB at point P due to the small element dl0 is dB ¼

lo Idl0 sin h 4 p r2

ð4:139Þ

The magnetic flux density at point P due to the entire wire is given by Z B¼

dB ¼

lo I 4p

ZL L

sin hdl0 r2

ð4:140Þ

From Fig. 4.26 h i1=2 2 r ¼ ðl 0 Þ þ r 2

Fig. 4.26 A current I flowing in a ﬁnite straight conductor of length 2L

ð4:141Þ

4.17

B of a Current-Carrying Finite Straight Conductor

315

and r r sin h ¼ sinðp hÞ ¼ ¼ h i1=2 r ðl 0 Þ2 þ r 2

ð4:142Þ

Substituting Eqs. 4.141 and 4.142 in Eq. 4.140 ZL

l I B¼ o 4p

L

ZL

l Ir B¼ o 4p

L

rdl0 h i3=2 ðl 0 Þ2 þ r 2

ð4:143Þ

dl0 h i3=2 ðl 0 Þ2 þ r 2

ð4:144Þ

2 B¼

3L 0

lo Ir 6 l 7 4 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ5 4p 2 2 2 r r þ ðl 0 Þ "

)B¼

4.18

lo I L pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 2 2 pr r þ L2

ð4:145Þ L

# ð4:146Þ

B Along the Axis of the Current-Carrying Circular Loop

Consider a circular coil of radius R carrying a current I as shown in Fig. 4.27. We are interested in calculating B at point P. Initially consider a small element dl0 as shown in Fig. 4.27a. The dB at point P due to the single element dl0 is dB ¼

lo I dl0 ^r 4 p r2

ð4:147Þ

In this case at point P the angle between dl0 and r is 90° hence dB ¼

lo I dl0 sin 90 4p r2

ð4:148Þ

lo I dl0 4p r2

ð4:149Þ

dB ¼

316

4 Magnetostatics

Fig. 4.27 a, b A circular coil carrying current I. Two opposing current elements and their magnetic flux densities are shown in the ﬁgure. c, d Resolution of magnetic flux densities into its components

Let us now consider another exactly opposite element dl0 as shown in Fig. 4.27b. The dB0 produced by element’s dl0 in Fig. 4.27a, b is shown in Fig. 4.27c. Clearly the horizontal sin a components cancel out while the vertical cos a components add up. Similarly the entire circular loop can be divided into small elements such that for opposite elements horizontal sin a components cancel out while all the vertical cos a components add up as shown in Fig. 4.27d. Hence the B at point P due to the entire loop is Z B ¼ dB cos a ð4:150Þ Z B¼

lo I dl0 cos a 4p r2

ð4:151Þ

4.18

B Along the Axis of the Current-Carrying Circular Loop

317

From Fig. 4.27a cos a ¼

R R ¼ 1=2 r 2 r þ R2

r2 ¼ r2 þ R2

ð4:152Þ ð4:153Þ

Substituting Eq. 4.152 and 4.153 in Eq. 4.151 lI B¼ o 4p B¼ But

R

Z

dl0 R

2 2 2 r þ R r þ R2 1=2

lo I R

4p r2 þ R2 3=2

Z

dl0

ð4:154Þ

ð4:155Þ

dl0 ¼ 2 pR the circumference of the circular loop B¼

lo I R

2 pR 4 p r2 þ R2 3=2

ð4:156Þ

lo I R2

2 r2 þ R2 3=2

ð4:157Þ

B¼

4.19

B Inside a Long Solenoid

(a) Using Biot–Savart law A solenoid consists of a long cylindrical coil of an insulated copper wire shown in Fig. 4.28a. The cross-sectional view of the current-carrying solenoid is shown in Fig. 4.28b. From Eq. 4.157 the B along the axis of the single circular loop is B¼

lo I R2 2 r2 þ R2 3=2

If the circular loop has N1 turns then B ¼ N1

lo I R2

2 r2 þ R2 3=2

ð4:158Þ

318

4 Magnetostatics

Fig. 4.28 a A solenoid carrying current I. b A cross - sectional view of a solenoid

Consider a axial length L of the solenoid which consists of N number of turns. The number of turns per unit length of the coil is N/L. Let P be the point in Fig. 4.28b where B is to be determined. The edges of the solenoid G1 and G2 make an angle a2 and a1 with the lines G1P and G2P as shown in Fig. 4.28b. Consider the small element CD of width dx situated at a distance of x from point P. The number of turns contained in the width dx is ¼

N Ldx

ð4:159Þ

Substituting for N1 ¼ N L dx in Eq. 4.158 we get the dB at point P due to the small element CD with x ¼ r dB ¼

N lo I R2 dx

L 2 x2 þ R2 3=2

ð4:160Þ

4.19

B Inside a Long Solenoid

319

where R is the radius of the solenoid as shown in Fig. 4.28. From Fig. 4.28b tan a ¼

R x

) x ¼ R cot a

ð4:161Þ

) dx ¼ R cosec2 a da

ð4:162Þ

R 2 þ x 2 ¼ b2

ð4:163Þ

and

sin a ¼

R b

ð4:164Þ

With the aid of Eqs. 4.161–4.164 Eq. 4.160 can be written as N lo IR2 R cosec2 a da L 2 b3

ð4:165Þ

N lo I R3 cosec2 a da L 2 b3

ð4:166Þ

N lo I 3 sin a cosec2 a da L 2

ð4:167Þ

lo IN sin a da 2L

ð4:168Þ

dB ¼

dB ¼ dB ¼

dB ¼

Thus B at point P due to the entire length of the solenoid is given by l0 IN B¼ 2L B¼

Za1 sin a da

ð4:169Þ

a2

lo IN ½cos a1 cos a2 2L

ð4:170Þ

Equation 4.170 is the value of B at point P in Fig. 4.28b. Let us consider a special case. For an inﬁnitely long solenoid when the point P lies well within the solenoid and far away from either end a1 ¼ 0 and a2 ¼ p

320

4 Magnetostatics

Fig. 4.29 An inﬁnitely long solenoid along with Amperian loops

then from Eq. 4.170 B¼

lo IN ½cos 0 cos p 2L )B¼

ð4:171Þ

lo IN L

ð4:172Þ

(b) Using Ampere’s law Consider an inﬁnite solenoid as shown in Fig. 4.29. In order to calculate the ﬁeld inside and outside the solenoid we construct two Amperian loops R1 and R2. In the loop R2 for path 2 and 4 the B and dl are perpendicular to each other. Hence Z

I B dl ¼ loopR2

Z B dl þ

Pathl

Path2

I

Z B dl þ

Path3

Z B dl ¼

loopR2

Z B dl þ

Path4

Z B dl þ 0 þ

Path1

B dl ð4:173Þ

B dl þ 0 Path3

ð4:174Þ

4.19

B Inside a Long Solenoid

I

Z B dl ¼

loopR2

321

Z B dl þ

Path1

B dl

ð4:175Þ

Path3

Because loop R2 doesn’t enclose current by Ampere’s law for loop R2 Z B dl ¼ lo Ienc ¼ 0 ð4:176Þ loopR2

From Eqs. 4.175 to 4.176 Z

Z B dl þ

Path1

B dl ¼ 0

ð4:177Þ

Path3

In Fig. 4.29 path 1 and path 3 lie at distances e and f, respectively, from one end of the solenoid. Hence we can write Eq. 4.177 as ½BðeÞ BðfÞL1 ¼ 0

ð4:178Þ

where L1 is the path length of path 1 and 3, respectively As L1 6¼ 0 we have BðeÞ ¼ BðfÞ

ð4:179Þ

We know that at a point far away from the solenoid B is zero. If we construct loop R2 such that the end f of loop R2 is far away from the solenoid then Bðf Þ ¼ 0 Hence from Eq. 4.179 BðeÞ ¼ BðfÞ ¼ 0

ð4:180Þ

Thus for all points outside the solenoid B is zero. For the loop R1 we have I B dl ¼ lo Ienc

ð4:181Þ

loopR1

If N is the total number of turns enclosed by the loop R1 in the solenoid then we can write Ienc ¼ NI

ð4:182Þ

Equation 4.182 can be written as I B dl ¼ lo NI loopR1

From Fig. 4.29

ð4:183Þ

322

4 Magnetostatics

Z

Z

Z

B dl þ ab

Z

B dl þ bc

B dl þ cd

B dl ¼ lo NI

ð4:184Þ

da

But for the path da and bc B is perpendicular to dl Z

Z B dl þ 0 þ

ab

B dl þ 0 ¼ lo NI

ð4:185Þ

B dl ¼ lo NI

ð4:186Þ

cd

Z )

Z B dl þ

ab

cd

For the path cd B = 0 because the path cd is outside the solenoid. Hence Z ) B dl þ 0 ¼ lo IN ð4:187Þ ab

Z )

B dl ¼ lo NI

ð4:188Þ

ab

Along the path ab B is parallel to dl. If the length of the path ab is L then Eq. 4.188 is BL ¼ lo NI )B¼

lo NI L

ð4:189Þ ð4:190Þ

The above problem demonstrates the advantage of using Ampere’s law as compared to Biot–Savart law. Ampere’s law uses symmetry to calculate B easily, while Biot–Savart law calculates B directly in terms of vector.

4.20

B of a Toroid

A toroid is solenoid in the form of a circle with the ends joined as shown in Fig. 4.30. Consider a circular Amperian loop P as shown in Fig. 4.30b. The B is circular within the cylindrical wire of the toroid. Let N be the number of turns in the toroid then Ienc ¼ N I where I is the current in each turn. Applying Ampere’s law to path P

4.20

B of a Toroid

323

Fig. 4.30 a A toroid carrying current I. b A toroid along with an Amperian loop is shown in the ﬁgure

I I

B dl ¼ lo Ienc

ð4:191Þ

B dl ¼ lo NI because B and dl are in same direction, B dl ¼ Bdl cos 0 ¼ B dl and B is constant over the Amperian loop so Eq. 4.191 is I ð4:192Þ B dl ¼ lo NI ) B 2pr ¼ lo NI )B¼

l0 NI 2 pr

ð4:193Þ ð4:194Þ

For points like P1 in Fig. 4.30b the Amperian loop doesn’t enclose any current and hence at point P1 I B dl ¼ lo Ienc ¼ 0 )B¼0 In the case of solenoid in Sect. 4.19 we have seen that for external points the B is zero. Hence at points like P2, B ¼ 0 Thus within the toroid B is given by

324

4 Magnetostatics

B¼

4.21

lo NI ^ / 2 pr

ð4:195Þ

Summary

In electrostatics we rewrote the Coulomb’s law as F R ¼ QE and deﬁned the electric ﬁeld E. Similarly in magnetostatics we wrote F ¼ I dl B and deﬁned magnetic flux density B. We noted that E is a vector and calculating E directly in terms of vector using Coulomb’s law is difﬁcult. Coulomb’s law calculates E in terms of vector and because vectors involve direction calculating E in terms of Coulomb’s law is tedious. So we went in search of new easier methods like Gauss’s law and potential to calculate E. Similarly in magnetostatics calculating B using Biot–Savart law is a difﬁcult process, because Biot–Savart law calculates B directly in terms of vector. So we went in search of new easier methods to calculate B. Like Gauss’s law in electrostatics we found Ampere’s law in magnetostatics. H q Gauss’s law reads E ds ¼ eenc . Successful and easy application of Gauss’s o law requires construction of a Gaussian surface over which E ds ¼ E ds cos h can be easily evaluated and to be constant over the Gaussian surface so that H

H H H E ds ¼ Eds cos h ¼ E cos h ds and ds, and the calculation of area is simple mathematics. Then I q E ds ¼ enc eo )E¼

qenc H cos h eo ds

ð4:196Þ

Hence E can be easily evaluated. When the above conditions are not satisﬁed it H q for the given Gaussian surface, but the quantity E we is still true that E ds ¼ eenc o are interested in calculating is inside the integral and hence E cannot be easily obtained using Gauss’s law. H Similarly in magnetostatics we found Ampere’s law as B dl ¼ lo Ienc : Successful and easy application of Ampere’s law requires construction of an H Amperian loop in which B dl ¼ B dl cos h can be easily evaluated and B cos h to be constant over the Amperian loop so that H

H H H B dl ¼ B dl cos h ¼ B cos h dl and dl, and the calculation of line integral is simple mathematics. Then

4.21

Summary

325

I B dl ¼ lo Ienc )B¼

lo Ienc H cos h dl

ð4:197Þ

Hence B can be easily evaluated. H When the above conditions are not satisﬁed it is still true that B dl ¼ lo Ienc but the quantity B which we require is inside the integral and hence B cannot be easily obtained using Ampere’s law. Symmetry is crucial for easy application of Gauss’s law for stationary charges and Ampere’s law for steady currents. In electrostatics three types of symmetry permit us to calculate E ds easily over the Gaussian surface and E cos H h to be constant over to Gaussian surface so that can be taken out of the integral E ds. The symmetries (which helps in calculating H E ds easier) are (1) Plane symmetry (Sect. 2.16) (2) Cylindrical symmetry (Sect. 2.13, Problems 2.1, 2.10, 2.11 and 2.9) (3) Spherical symmetry (Sect. 2.17, Problem 2.8). Similarly in magnetostatics four types of symmetry enable us to calculate B H dl easily over the Amperian loop so that B cos h can be taken out of the integral B dl (1) (2) (3) (4)

Inﬁnite straight line currents (Sect. 4.16) Inﬁnite solenoid (Sect. 4.19) Toroids (Sect. 4.20) Inﬁnite planes (Problem 4.1).

When Gauss’s law is not useful in calculating E we went in search of other easy methods to calculate E and we found one method—potential. Potential V is a scalar quantity, and working with scalar quantities is much easier as compared to working with vector quantities. Once V has been found for the given charge distribution it is easy to calculate E because E ¼ rV. Similarly when Ampere’s law is not useful we went in search of other methods. Our attempt to ﬁnd easy way of calculating B using potential was not so much successful. The magnetic vector potential A still happens to be a vector. Our attempt to calculate B using magnetic scalar potential Vm was also not completely successful, because Vm is a path-dependent function although it is a scalar quantity. And for other reasons mentioned in Sect. 4.15 both A and Vm are not as much useful as electric potential V.

326

4.22

4 Magnetostatics

Magnetic Dipole

Consider a current-carrying loop as shown in Fig. 4.31. Let the current in the loop be I and we are interested in calculating the vector potential A at point P which is situated at a distance of r from the centre of the loop. The vector potential at point P due to the entire loop is A¼

lo I 4p

I C

0

dl0 r

ð4:198Þ

From Fig. 4.31 r¼ 1 ¼ r

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ r2 þ ðr0 Þ2 2rr0 cos h 1 r2 þ ðr0 Þ2 2 r r0 cos h

1=2

" #1=2 0 2 1 1 r 2r0 ¼ 1þ cos h r r r r

Fig. 4.31 A magnetic dipole - A current carrying loop

ð4:199Þ

ð4:200Þ

4.22

Magnetic Dipole

327

Suppose the point P is very far away from the loop then r2 ðr0 Þ2 and hence r 0 2 r term can be neglected

1=2 1 1 r0 ’ 1 2 cos h r r r

1 1 r0 ) ’ 1 þ cos h r r r

ð4:201Þ ð4:202Þ

Substituting Eq. 4.202 in Eq. 4.198 A¼

lo I 4p

I C0

1 r0 1 þ cos h dl0 r r

ð4:203Þ

Noting down that integration is carried out with primed source coordinates we can write Eq. 4.203 2 A¼

lo I 41 4p r

I

dl0 þ

C0

1 r2

I

3 r0 cos h dl05

ð4:204Þ

C0

H

dl0 represents net vector displacement. Because the initial and ﬁnal points are same the net vector displacement is zero. I dl0 ¼ 0 C0

C0

Hence Eq. 4.204 is l I1 A¼ o 2 4p r

I C

A¼

lo I 4pr2

r0 cos h dl0

ð4:205Þ

ð^r r0 Þdl0

ð4:206Þ

0

I C0

From Eq. 1.135 in Example 1.17 with Q ¼ ^r we get I Z 0 0 ð^r r Þdl ¼ ^r ds0 C

0

S

0

ð4:207Þ

328

4 Magnetostatics

Substituting Eq. 4.207 in Eq. 4.206 2 3 Z lo I 4 ^r ds0 5 A¼ 4 pr2 S

ð4:208Þ

0

We deﬁne the magnetic dipole moment m as Z ^ ¼ IS0 k ^ ¼ I pR2 k ^ m ¼ I ds0 k

ð4:209Þ

S0

Hence Eq. 4.208 is lo ½^r m 4 p r2

l m ^r )A¼ o r2 4p A¼

ð4:210Þ ð4:211Þ

The magnitude of m is the product of the current and the area of the loop. The direction of m is given by right-hand rule when the ﬁngers of right hand curl around the direction in which the current I is flowing then the thumb of right hand gives the direction of m. From Eq. 2.256 the potential of an electric dipole is given by V¼

1 p ^r 4 p e0 r 2

Comparing Eq. 4.211 with Eq. 2.256 we see that A is analogous to V. Thus we see that a small current-carrying loop is a magnetic dipole. Now let us calculate B of the magnetic dipole. We know that B¼rA Substituting Eq. 4.211 in B

lo m ^r B¼r 4p ^r2 m r l B¼ o r r3 4p

ð4:212Þ ð4:213Þ

4.22

Magnetic Dipole

329

Because m is a constant vector setting Q ¼ m and R ¼ rr3 in Eq. 1.134 from Example 1.16 Eq. 4.213 can be written as B¼

lo h r ri m r 3 ðm rÞ 3 r r 4p

ð4:214Þ

^ from Fig. 4.31 r ¼ x^i þ y^j þ zk We have mx

@ r mx^i r ¼ 3 3mx x 5 3 @x r r r

ð4:215Þ

my

@ r my^j r ¼ 3 3my y 5 3 @y r r r

ð4:216Þ

mz

^ @ r mz k r ¼ 3 3mz z 5 3 r @z r r

ð4:217Þ

Using Eqs. 4.215, 4.216, 4.217 we can write ðm rÞ

r m 3ðm rÞr ¼ r3 r3 r5

ð4:218Þ

and we have

r 3 3r 3 3 m r 3 ¼m 3r 5 ¼m 3 3 ¼0 r r r r r

ð4:219Þ

Substituting Eqs. 4.219 and 4.218 in Eq. 4.214 B¼

lo 3ðm rÞr m r5 r3 4p

ð4:220Þ

In Fig. 4.32a we plot the electric ﬁeld of an electric dipole and in Fig. 4.32b we plot B of a magnetic dipole.

4.23

Magnetic Boundary Conditions

In electrostatics we saw that at the interface between two media the electric ﬁeld suffers a discontinuity. Similarly at the interface between two magnetic media B suffers a discontinuity. The two magnetic media may be two magnetic materials with different magnetic properties or a magnetic media and vacuum. Consider an interface between two magnetic media carrying a surface current K as shown in Fig. 4.33a.

330

4 Magnetostatics

Fig. 4.32 a Plot of electric ﬁeld lines of a electric dipole. b Plot of magnetic flux density of a magnetic dipole

Fig. 4.33 Interface between two magnetic media carrying a surface current density shown along. a with a pill box to evaluate the relationship between normal components of magnetic flux density. b with a Amperian loop to evaluate the relationship between tangential components of magnetic flux density

4.23

Magnetic Boundary Conditions

331

We know that r B ¼ 0 Selecting an arbitrary volume s and integrating the above equation over the volume s Z r Bds ¼ 0 ð4:221Þ s

Applying Gauss’s divergence theorem I B ds ¼ 0

ð4:222Þ

In Fig. 4.33a let us construct a pillbox of negligible sides, and the area of the top and bottom surface of the pillbox is S. Because the sides of the pillbox are negligible, applying Eq. 4.222 to the pill box B1n S B2n S ¼ 0

ð4:223Þ

) ðB1n B2n ÞS ¼ 0

ð4:224Þ

As S 6¼ 0 B1n B2n ¼ 0 ) B1n ¼ B2n

ð4:225Þ

In order to ﬁnd the relation between tangential components construct an Amperian loop as shown in Fig. 4.33b. The Amperian loop is running perpendicular to the current. The Amperian loop is so constructed that the side of the loop a is negligible. Then applying Ampere’s law I B dl ¼ lo Ienc Because the side a is negligible for the Amperian loop B1t L B2t L ¼ lo KL

ð4:226Þ

) B1t B2t ¼ lo K

ð4:227Þ

Thus when there is a surface current then the tangential component of B is discontinuous by the amount given by Eq. 4.227. If K is zero across the boundary between two magnetic media then B1t ¼ B2t

ð4:228Þ

332

4.24

4 Magnetostatics

Force Between Two Parallel Current-Carrying Conductors

Consider two long straight conductors a and b carrying currents Ia and Ib. Let the distance of separation between the two conductors be d. The value of B at the site of conductor b produced by conductor a is given by (Eq. 4.47) Ba ¼

lo Ia 2 pd

ð4:229Þ

By right-hand rule Ba points into the page as shown in Fig. 4.34. The length L of the conductor b experiences a force given by Fa ¼ Ib LBa

ð4:230Þ

Substituting Eq. 4.229 in 4.230 lo Ia 2 pd

ð4:231Þ

lo L Ia Ib 2 pd

ð4:232Þ

Fa ¼ Ib L Fa ¼

Fig. 4.34 Two long straight conductors a, b carrying currents Ia and Ib

4.24

Force Between Two Parallel Current-Carrying Conductors

333

Because the magnetic force is given by F ¼ Iðdl BÞ the direction of Fa is towards the conductor a as shown in Fig. 4.34. The value of B at the site of conductor a produced by conductor b is given by Bb ¼

lo Ib 2 pd

ð4:233Þ

By right-hand rule Bb points out of the page as shown in Fig. 4.34. The length L of the conductor a experiences a force Fb given by Fb ¼ Ia LBb lo Ib 2 pd

ð4:234Þ

lo L Ia Ib 2 pd

ð4:235Þ

Fb ¼ Ia L ) Fb ¼

The direction of Fb is shown in Fig. 4.34. Thus Fa and Fb are equal in magnitude but opposite to each other. Therefore two parallel straight conductors carrying currents in same direction attract each other. Similarly it can be shown that two parallel straight conductors carrying currents in opposite direction repel each other.

4.25

Torque on a Current Loop in a Uniform Magnetic Field

Consider a rectangular loop PQRS placed in a uniform magnetic flux density B as shown in Fig. 4.35. In Fig. 4.35 PS ¼ QR ¼ b and PQ ¼ SR ¼ a. The loop is situated in the B such that the axis (shown as dotted line in Fig. 4.35) is perpendicular to B, while the plane of the loop makes an angle h with the direction of B. The loop carries a current I. The force F1 acting on the limb Q R is given by F1 ¼ Iðb BÞ

ð4:236Þ

) F1 ¼ Ib B sin h

ð4:237Þ

The direction of F1 is downward. Similarly F2 given by Ib B sin h is acting upward and hence F1 and F2 cancel out each other. Now the force F4 acting on limb PQ is given by F4 ¼ Iða BÞ ¼ IaB sin 90 ¼ IaB

ð4:238Þ

334

4 Magnetostatics

Fig. 4.35 A rectangular loop PQRS placed in a uniform magnetic ﬁeld

The direction of F4 is shown in Fig. 4.35. The force F3 acting on limb SR is IaB but has a direction opposite to F4 as shown in Fig. 4.35. Thus F3 and F4 are opposite to each other and hence cancel out. However the line of action of F3 and F4 is different and constitute a torque. The top cross-sectional view of the rectangular loop placed in B is shown in Fig. 4.36. The torque acting on the loop is given by

Fig. 4.36 Top cross sectional view of the rectangular loop placed in the uniform magnetic ﬁeld

T ¼ F4 ðGSÞ ¼ F4 b cos h

ð4:239Þ

T ¼ IaBb cos h

ð4:240Þ

T ¼ IBðabÞ cos h

ð4:241Þ

4.25

Torque on a Current Loop in a Uniform Magnetic Field

335

) T ¼ IBS1 cos h

ð4:242Þ

) T ¼ IðS1 BÞ

ð4:243Þ

where S1 = ab is the area of the coil. If there are N turns in the loop then the torque is T ¼ NIS1 B

ð4:244Þ

From Eq. 4.209 m ¼ I S1 is the magnetic moment and hence T ¼ Nm B

4.26

ð4:245Þ

Magnetic Flux

Magnetic flux is deﬁned as the total number of lines of magnetic flux density B passing through the given surface and is denoted by U. Let B be the magnetic flux density. In Fig. 4.37 we see that B makes an angle h with the normal to the given surface whose area is S. From Sect. 1.12 the flux passing through the surface S in Fig. 4.37 is then I U ¼ B ds ð4:246Þ The unit of U is weber. Example 4.1 Find the current distribution that produces a magnetic flux density of B ¼ j cos x^j where j is a constant. Fig. 4.37 Lines of magnetic flux density passing through a surface area S - Magnetic flux

336

4 Magnetostatics

Solution We know that r B ¼ l0 J 1 rB lo ^j ^ ^i k 1 @ @ @ ¼ lo @x @y @z 0 j cos x 0

J¼

J¼

j sin x ^ k l0

Example 4.2 (a) Consider a cylinder of circular cross section of radius R which carries a uniform current I perpendicular to the circular cross section. Find J. (b) If the current density is such that J ¼ jr2c then ﬁnd the total current in the wire. Solution dI where ds is the area perpendicular to the flow. Here (a) We know that J ¼ ds ? ? 2 I ds? ¼ pR Hence J ¼ 2 as the current I is uniformly distributed. pR R (b) Here I ¼ J ds? because J varies with rc. Then R I ¼ j r2c rc drc d/ as ds? ¼ rc drc d/ is the area perpendicular to the flow. Z2p

ZR I¼

j r3c drc 0

d/ ¼

p j R4 2

0

Example 4.3 A disc carries a uniform static surface charge density. The disc is rotating with an angular velocity x. Find K at a distance r from the centre. Solution We know that K ¼ rv but v ¼ xr. Hence K ¼ rxr. Example 4.4 Consider two loops each of radius R situated as shown in Fig. 4.38. Calculate B at point P. Each loop is carrying current I.

4.26

Magnetic Flux

337

Fig. 4.38 Two loops - loop a and loop b each of radius R carrying a current I

Solution We have seen in Eq. 4.157 B due to a current-carrying circular loop along its axis is given by Ba ¼

lo I R2 ^

k 2 z2 þ R2 3=2

B at point P due to the loop b is Bb ¼

lo I R2 ^ k as z ¼ 0 for loop b: 2 R3 Bb ¼

lo I 1 ^ k 2 R

Hence the net B at point P is B ¼ Ba þ Bb " # lo I 1 R2 ^ þ B¼

3=2 k 2 R z2 þ R 2

338

4 Magnetostatics

Fig. 4.39 Two circular coils P and Q having common axis and are carrying current

Example 4.5 Two circular coils P and Q and have common axis as shown in Fig. 4.39. Coil P has 10 turns and has radius R. The current passing through coil P is 12 A. Coil Q has 2 turns with a radius of 7 cm, and the current passing through it is 10 A. What should be the value of R so that B be zero at the centre of coil P. The distance between the coils is 5 cm. Solution From Eq. 4.157 B¼

l0 IN R2

2 z2 þ R2 3=2

For coil P, the value of B (I ¼ 12 A, R ¼ R, z ¼ 0; N ¼ 10)

at

the

centre

4p 107 R2 ð12Þð10Þ 3 2 R 4 p 107 1 BP ¼ ð10Þð12Þ R 2 BP ¼

The value of B due to coil Q at the centre of coil P is (I ¼ 10 A, N ¼ 2; R ¼ 7 cm, z ¼ 5 cm). BQ ¼

4 p 107 ð10Þð2Þ 72 2 ½ 52 þ 72

For the value of B at the centre of coil P to be zero

of

coil

P

is

4.26

Magnetic Flux

339

BP ¼ BQ (Noting down that BP and BQ are in opposite directions) 4 p 107 1 4p 107 72 ð10Þð12Þ ¼ ð10Þð2Þ 2 R 2 2 ð 5 þ 72 Þ ) R ¼ 9:06 cm: Example 4.6 Consider a long straight wire carrying current I. Calculate the magnetic vector potential A and hence calculate B. Solution Let L be the length of the wire. Let it carry current I as shown in Fig. 4.40. Let us solve the problem in Cartesian coordinates. At point P the vector potential dA due to element dx is given by l I dx ^ dA ¼ o i 4p r The magnetic vector potential at P due to the whole wire is then l I A¼ o 4p

lI A¼ o 4p A¼

Fig. 4.40 A long straight wire carrying a current I

ZL=2 L=2

ZL=2 L=2

dx^ i r

dx pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ^i x2 þ y2

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃiiL=2 lo I h h ^i log x þ x2 þ y2 L=2 4p

340

4 Magnetostatics

2

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ﬃ3 2 L 2 l I 6 L=2 þ 4 þ y 7^ ﬃ5i qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ A ¼ o log4 2 4p L=2 þ L4 þ y2 If the wire is very long so that L2 y2 then 1=2 4y2 1 þ 1 þ 2 l I ^ L A ¼ o log i 1=2 2 4p 1 þ 1 þ 4y2 L 3 2 4y2 1 þ 1 þ 2 l I 2L 2 5^i A ’ o log4 4y 4p 1 þ 1 þ 2 2L

l I L2 A ¼ o log 1 þ 2 ^i 4p y As L2 y2 A¼

lo I L 2 log ^i 4p y

As B ¼ r A ^i lo @ 2I @x B¼ 4p log L

^j

@ @y

y

0

^ k @ @z 0

lo ^ @ log L 2I k @y y 4p

l @ ^ ½log L log y k B ¼ 0 2I @y 4p B¼

As L is a constant. B¼

lo 2I ^ k 4p y

B¼

lo I ^ k 2py

4.26

Magnetic Flux

341

Example 4.7 A vector potential is given by ^ in Cartesian coordinates (a) A ¼ y2 sin x^i þ e2x k ^ (b) A ¼ j cos h h in spherical polar coordinates. Calculate B. Solution B¼rA ðaÞ

^ A ¼ y2 sin x^i þ e2x k ^i ^j ^ k @ @ @ B ¼ @y @z @x y2 sin x 0 e2x ^ B ¼ ze2x^j 2y sin xk

ðbÞ

B¼rA¼ B¼

[email protected] ^ ðjr cos hÞ/ r @r

j cos h ^ / r

Example 4.8 A wire of conﬁguration shown in Fig. 4.41 (wire PQRS) carries a current I. Calculate the B at point P1. Solution The point P1 lies along the line PQ and along the line SR. Hence the B produced by PQ and SR at the site of P1 is zero [dl;^r in Biot–Savart law have either same or opposite direction]. The flux density at the centre of a circular loop is given by (Eq. 4.157) B¼

lo I 2r

where r is the radius of the loop. The flux density at point P1 due to the segment PS is then BPS ¼

h lo I 2p 2r1

342

4 Magnetostatics

Fig. 4.41 A wire PQRS carrying a current I

The flux density at P due to the segment BC is then BQR ¼

h lo I 2p 2r2

Applying right-hand rule we see that the direction of BPS is out of the plane and the direction of BQR is into the plane of the paper. From Eq. 4.36 the value of net B at point P1 is B ¼ BPS BQR B¼

lo I ðr2 r1 Þ 4p r2 r1

As r1 is shorter than r2 BPS > BQR. Hence B at point P1 will be directed out of the plane of paper. Example 4.9 Consider a long wire carrying current I. The small portion of the wire is bent into a semicircular loop as shown in Fig. 4.42. The radius of the circular loop is R. Calculate the magnetic flux density at point P due to the entire wire. Solution The B due to the straight portion of the wire is zero because B¼

lo I 4p

Z

dl ^r r2

4.26

Magnetic Flux

343

Fig. 4.42 A long wire carrying current I. A small portion of the wire is bent into a semicircle of radius R

For the straight portion of the wire angle between dl and ^r is zero. Hence for the straight portion B¼ B¼

lo I 4p

lo I 4p

Z

Z

dl ^r r2 dl0 sin 0 r2

B ¼ 0: For the curved portion B¼

lo I 4p

Z

dl ^r

r2

From Fig. 4.42 r ¼ R and dl ¼ Rdh ^ h l I B¼ o 4p

Zp 0

Rdhð^ h ^rÞ

r2

344

4 Magnetostatics

l I B¼ o 4 pR B¼

Zp dhbz 0

lo I l I pbz ¼ o bz 4 pR 4R B¼

lo I bz 4R

In Fig. 4.42 we see that ^z always points in the z-direction. Being a unit vector it has unit magnitude. Thus ^z has constant magnitude and constant direction. However consider ^ h at points P1 and P2 in Fig. 4.43. We see that ^h being a unit vector has unit magnitude. At points P1 and P2 it has different directions. Similarly in cylindrical coordinates ^z always has constant magnitude and con^ ^rC depends on the point in the consideration. stant direction. But the direction of /, ^ ^ The direction of ^r, h, / depends on the point in consideration in spherical polar ^ have different directions depending upon the point in coordinates. Thus ^r, ^ h and / consideration in spherical polar coordinates. ^ ^z in cylindrical coordinates, ð^rc ; h; ^ in ^ /Þ Hence whenever working with ^rc ; /; spherical polar coordinates better to express these unit vectors in terms of cartesian ^ using the following equations. coordinates ^i; ^j; k In cylindrical coordinates 9 ^rc ¼ cos /^i þ sin /^j = ^ ¼ sin /^i þ cos /^j / ; ^ ^z ¼ k In spherical polar coordinates

Fig. 4.43 Figure depicts unit vectors at points P1 and P2 have different directions

ð4:247Þ

4.26

Magnetic Flux

345

9 ^= ^r ¼ sin h cos /^i þ sin h sin /^j þ cos hk ^ ^ h ¼ cos h cos /^i þ cos h sin /^j sin hk ; ^ ¼ sin h^i þ cos /^j /

ð4:248Þ

^ have constant magnitude and direction. The unit vectors The unit vectors ^i; ^j; k always point in x-, y- and z-direction, respectively, and hence are easy to work with. Example 4.10 If B is given by B ¼ jr cos /^rc Calculate the magnetic flux in cylindrical coordinates for the surface 0 z 2m and p=2 / p=2 (j is a constant). Solution We know that Z U¼

B ds

Z2 Zp=2 )U¼ 0

p=2

j cos /^rc rd/dz^rc r Zp=2

)U¼2

cos / d/ p=2 p=2

) U ¼ 2½sin /p=2 ) U ¼ 4 webers: Example 4.11 Consider a regular hexagon of side a carrying current I. Find the expression for B at the centre of the hexagon. Solution Consider a hexagon of side a as shown in Fig. 4.44. The B produced by side A1 A2 at point is given by (Eq. 4.146 with L ¼ a/2) BA1 A2 From Fig. 4.44

" # lo I a=2 pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ¼ 2 pr r2 þ a2 =4

346

4 Magnetostatics

Fig. 4.44 A regular hexagon of side a carrying current I

a2 ¼

a2 þ r2 4

3a2 ; )r ¼ 4 2

pﬃﬃﬃ 3a : r¼ 2

Substituting for r2, r in BA1 A2 2

BA1 A2

3 a lI6 7 2 ¼ o 4qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 5 2 2 pr 3 a2 a þ 4 4 BA1 A2 ¼

BA1 A2 ¼

lo I 4 pr

lo I lo I pﬃﬃ ¼ pﬃﬃﬃ 3a 2 p 3a 4p 2

The hexagon is having six sides and hence the total B at point P is Btot

6lo I pﬃﬃﬃ ¼ ¼ 2 p 3a

pﬃﬃﬃ 3lo I pa

At point P the B points inward perpendicular to the plane of paper. Example 4.12 Consider a long solenoid of length 2 m having 2000 turns of wire. The mean radius is 20 cm. A current of 30 A flows through the wire. Calculate B at the centre of the solenoid along its axis. Solution Consider Fig. 4.45. From Eq. 4.170

4.26

Magnetic Flux

347

Fig. 4.45 A long solenoid carrying current I

B¼ B¼

lo IN ½cos a1 cos a2 2L

lo In n ½cos a1 cos a2 with N ¼ 2 L

Given I ¼ 30 A, N ¼ 1000 turns/m. lo ¼ 4p 107

weber Am

1 cos a1 ¼ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 1 þ ð0:2Þ2 1 cos a1 ¼ pﬃﬃﬃﬃﬃﬃﬃ 1:4 1 cos a2 ¼ cosðp a1 Þ ¼ pﬃﬃﬃﬃﬃﬃﬃ 1:4

ð4 p 107 Þð30Þð1000Þ 2 pﬃﬃﬃﬃﬃﬃﬃ B¼ 2 1:4 B ¼ 3:18 102 weber=m2 :

348

4 Magnetostatics

Exercises Problems 4.1 An inﬁnite sheet of current flows in the x−y plane as shown in Fig. 4.46. Find the magnetic flux density. Problems 4.2 Show that for Fig. 4.47a resolving into parallel and perpendicular components B1n ¼ B2n and B1t B2t ¼ lo K. Similarly for Fig. 4.47b show that B1n ¼ B2n ; B1t ¼ B2t .

Fig. 4.46 An inﬁnite sheet of current carrying a surface current density

Fig. 4.47 An inﬁnite sheet of current. Illustration for problem 4.2

Exercises

349

Fig. 4.48 Two inﬁnite wires each carrying current I. Illustration for problem 4.4

Problem 4.3 Consider a cylindrical conductor of radius R. The current density within the cylindrical region is given by J ¼ Cejr ^z where C and j are positive constants. Determine magnetic flux density everywhere. Problem 4.4 Consider two inﬁnite wires each carrying current I as shown in Fig. (4.48). Calculate B at point P in Fig. (4.48). Problem 4.5 A solid inﬁnite cylindrical conductor with radius R is carrying a current I. Find the B both inside and outside the cylinder. Sketch the variation of B as a function of distance from the conductor axis. Problem 4.6 Consider a wire of radius R. A current I flows through the wire. Find (a) The volume current density if the current distribution is such that it is directly proportional to distance from the axis (b) If the charges are moving such that its volume charge density is q then calculate their velocity. Problem 4.7 Consider circular coils carrying current I1 and I2 and placed apart by a distance r. Calculate the force caused by current element I1dl1 on I2dl2 and by I2dl2 on I1dl1. Then calculate the force exerted by coil 1 on coil 2 and coil 2 on coil 1. Problem 4.8 Consider two loops of radius 15 cm situated as shown in Fig. 4.49. Each loop is carrying a current of 10A. The distance between the centre of loop is 20 cm. Calculate the B at (a) z ¼ 10 cm (b) z ¼ 20 cm. (c) z ¼ 30 cm. Problem 4.9 If the current in Fig. 4.27 is reversed how will the answer to Sect. 4.18 change. Plot the magnetic flux density for Fig. 4.50.

350

4 Magnetostatics

Fig. 4.49 Two loops carrying current. Illustration for problem 4.8

Fig. 4.50 A circular loop carrying current I

Problem 4.10 Find the value of A and B for two long parallel wires separated by a distance h. Both the wires carry current I in opposite direction as shown in Fig. 4.51. Problem 4.11 Consider a spherical shell of radius R. It carries a uniform surface charge. If it is rotating about its own axis ﬁnd the vector potential at a point outside the spherical shell. Problem 4.12 Two inﬁnitely long wires each carrying current of 15 A are situated as shown in Fig. 4.52. Find the value of B at the points P1, P2 and P3. Clearly mention the direction of B at each of these points. Both the wires are carrying currents in the same direction. Problem 4.13 In Problem 4.12 if the wires are carrying current in opposite direction, how the answer to Problem 4.12 will get modiﬁed.

Exercises

351

Fig. 4.51 Two long parallel wires carrying current. Illustration for problem 4.10

Fig. 4.52 Illustration for problem 4.12

Problem 4.14 Find the J which will produce a vector potential (a) A ¼ C sin h ^ h in spherical polar coordinates. ^ in Cartesian coordinates. (b) A ¼ y2 cos x ^i þ ex k ^ in cylindrical coordinates. (c) A ¼ Cr2 / where C is a constant. Problem 4.15 Three inﬁnitely long wires a, b and c each carrying current I1, I2 and I3 are isolated from each other as shown in Fig. 4.53. Calculate the magnetic force on dl2 and dl3 due to I1. Problem 4.16 Find the vector potential above and below the plane in Problem 4.1

352

4 Magnetostatics

Fig. 4.53 Three inﬁnitely long wires carrying current. Illustration for problem 4.15

Fig. 4.54 A wire in the form of equilateral triangle carrying current and placed in a magnetic ﬁeld. Illustration for problem 4.17.

Problem 4.17 A wire is bent in the form of an equilateral triangle. The side of the triangle is 5 cm. A current of 10 A flows in the triangle. The entire triangle is placed in a magnetic ﬁeld as shown in Fig. 4.54. Find the forces on the three sides of the triangle. Problem 4.18 Calculate the magnetic dipole moment of spinning spherical shell in Problem 4.11.

Exercises

353

Fig. 4.55 A cylindrical conductor placed in a cylindrical coaxial cable. Illustration for problem 4.19

Problem 4.19 A cylindrical conductor is placed in a cylindrical coaxial cable as shown in Fig. 4.55. The inner conductor carries a current I uniformly distributed over the conductor. The coaxial cable carries an equal current in opposite direction with uniform current density. Find the value of B at a distance r from the centre where (a) r < a (b) a < r < b (c) b < r < c (d) r > c.

Chapter 5

Magnetic Fields in Materials

5.1

Introduction

In the previous chapter we calculated the magnetic flux density B for a given current distribution and the current distribution was assumed to be present in vacuum. For example we calculated the magnetic flux density B of a solenoid. Inside the solenoid it was assumed that there is a free space or vacuum. Now instead of free space or vacuum inside the solenoid if the free space is ﬁlled with a material, will the magnetic flux density inside the solenoid get altered? The answer is yes. To be precise, suppose assume that we place a material in a magnetic ﬁeld. How then the magnetic ﬁeld gets changed? Our purpose is this chapter is to answer the above questions.

5.2

Diamagnetism, Paramagnetism and Ferromagnetism

In the previous chapter we saw that magnetic monopoles do not exist in nature. It is the circulation of electric charges which produce magnetic ﬁeld. However if we take a bar magnet we will see it has nothing to do with currents but still it produces a magnetic ﬁeld. If you examine the bar magnet at a microscopic level you will still observe very small currents, i.e. electrons orbiting about the nuclei.1 As we will see this electron motion gives rise to macroscopic magnetic properties of materials. The electronic current loops are small enough so that they can be treated as magnetic dipoles and in Sect. 4.22 we have seen that the magnetic dipole posses a magnetic dipole moment.

1

Strictly speaking the motion of electrons must be studied using the theories of quantum mechanics. © Springer Nature Singapore Pte Ltd. 2020 S. Balaji, Electromagnetics Made Easy, https://doi.org/10.1007/978-981-15-2658-9_5

355

356

5 Magnetic Fields in Materials

In the case of diamagnetic materials say for example hydrogen molecule, a number of electrons present in each atom tend to pair up. The result is magnetic effects of electrons both due to spin as well as orbital motion exactly cancel out each other and the atom as a whole is devoid of any permanent magnetic moment. “When such materials are placed in a magnetic ﬁeld there will be a change in the magnetic moment and the change will be such that it is opposite to the direction of the applied magnetic ﬁeld” (see Problem 5.1). These substances are repelled away from the applied magnetic ﬁeld. In the case of paramagnetic materials say for example copper, there will be at least are electron which doesn’t pair up with other electrons. In such cases the atom has a permanent magnetic moment. If we take a bulk of such material due to random alignments of such atomic magnetic dipoles net magnetic moment is zero as shown in Fig. 5.1a. However when such materials are placed in an external magnetic ﬁeld a “torque” (see Problem 5.2) acts on the atomic dipoles tending to align it in the direction of the magnetic ﬁeld as shown in Fig. 5.1b. These substances known as paramagnetic substances are feebly attracted by external magnetic ﬁeld. Ferromagnetic materials in comparison to paramagnetic materials are distinguished by high degree of alignment of atomic dipoles. Ferromagnetic substances exhibit almost all the properties of paramagnetic materials but to a much larger extent. A strong coupling force exists between atoms in a small region of the ferromagnetic material which makes the magnetic dipole moments of all the atoms contained in the small region to orient in the same direction. This small region is called domain. A ferromagnetic material contains large number of such domains as shown in Fig. 5.2. In the absence of an external applied magnetic ﬁeld the net magnetic moments of domain’s point in different directions as shown in Fig. 5.2 and hence as a whole the magnetic moment of the entire material is zero. However when a ferromagnetic material is placed in a external magnetic ﬁeld and the magnetic ﬁeld is increased domains whose magnetization are parallel to the applied ﬁeld grow and the others gets shrinked. At very high ﬁelds one domain grows over completely and then the ferromagnetic material is said to be saturated.

Fig. 5.1 a Randomly oriented atomic magnetic dipoles in a paramagnetic material. b Atomic magnetic dipoles aligned in the direction of the externally applied magnetic ﬁeld

5.2 Diamagnetism, Paramagnetism and Ferromagnetism

357

Fig. 5.2 Domains in a ferromagnetic material

Let us now switch off the external magnetic ﬁeld. It is observed that the domain growth process is irreversible. There is some return of the domains to the randomly oriented state, but the process is incomplete. Hence now the ferromagnetic material turns out to be a permanent magnet. Now suppose you put a ferromagnetic material is a magnetic ﬁeld and slowly increase the ﬁeld. The magnetization and the flux density B increases slowly as represented by the part OA of the curve is Fig. 5.3 and reaches a maximum value at A. The ferromagnetic material is said to have reached the saturation. If we reduce the magnetic ﬁeld to zero the flux density is the magnetic material decreases at a lessor rate and does not become zero at H = 0. The flux density OC left is the specimen is called residual magnetism (or) remanence (or) retentivity. If we now reverse the direction of the ﬁeld H, the value of B decreases and ﬁnally becomes zero for a particular value of −H. This decrease is represented as CD in Fig. 5.3. The value of H corresponding to OD is Fig. 5.3 which renders B to zero is called coercive force or coercivity of the material. With further increase in H in the reverse direction the part DE of the curve is obtained. Here magnetic saturation occurs at E, symmetrical to that of A. If we now reduce the magnetizing ﬁeld H to zero and then increase H in positive direction we reach back A along the path EFGA. Fig. 5.3 Hysteresis loop for a ferromagnetic material

358

5 Magnetic Fields in Materials

Fig. 5.4 Hysteresis loop for a soft and hard ferromagnetic material

From Fig. 5.3 it is clear that the flux density B always lag behind the magnetizing ﬁeld H, when the ferromagnetic material is taken through a complete cycle of magnetization. This loss of magnetization behind the magnetizing ﬁeld is called hysteresis and the curve shown in Fig. 5.3 is called hysteresis loop or curve. The area under the curve is the energy lost per unit volume per cycle. This loss is also known as hysterestis loss and is the energy used to overcome friction encountered during domain wall motion and domain rotation. Ferromagnetic materials in general can be classiﬁed into two types: hard magnetic materials and soft magnetic materials. The typical hysteresis loop for hard and soft ferromagnetic material is shown in Fig. 5.4. Hard magnetic materials have high coercivity and show high resistance to magnetization and demagnetization. Hence these materials are used in permanent magnets, recording media where high coercivity is needed. Soft magnetic materials are those which can be easily magnetized and demagnetized because their domain walls can move easily. They are characterized by high saturation magnetization and low coercivity. Because their magnetization is high they are used in motors, transformers and electric generators. We have already seen that coupling between neighbouring atoms gives rise to ferromagnetism. If the temperature of the ferromagnetic materials is raised above a critical temperature, the thermal energy exceeds the coupling energy and the coupling between atoms is broken. The critical temperature is known as Curie temperature and the material above the Curie temperature behaves like paramagnetic material. In ferromagnetic materials coupling between atoms is such that the alignments are in the same direction leading to formation of domains. However in some elements like chromium instead of parallel alignment atomic dipoles align in an antiparallel manner as shown in Fig. 5.5a. The net magnetic moment is thus zero and these materials are called antiferromagnetic material.

5.2 Diamagnetism, Paramagnetism and Ferromagnetism

359

Fig. 5.5 a Atomic magnetic dipoles of equal magnitude having anti-parallel arrangement in a antiferromagnetic material. b Atomic magnetic dipoles of unequal magnitude having anti-parallel arrangement in a ferrimagnetic material

There are another class of materials known as ferrimagnetic materials. In these materials as shown in Fig. 5.5b the magnetic moments are antiparallel but are of unequal magnitude. Hence the net magnetic moment will not be zero and a spontaneous magnetization remains. For example if the adjacent moments correspond to different ions [say Fe2+ and Fe3+] then they will have unequal magnetic moments leading to incomplete cancellation of dipole moments. Example 5.1 Prove that $

1

r

¼ $

0

1

r

Solution From Fig. 4.15

r ¼ ðx x0 Þ^i þ ðy y0 Þ^j þ ðz z0 Þk^ r2 ¼ ðx x0 Þ2 þ ðy y0 Þ2 þ ðz z0 Þ2

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ðx x0 Þ2 þ ðy y0 Þ2 þ ðz z0 Þ2 1 @ 1 @ 1 @ 1 ^ ^ ^ $ ¼i þj þk @x r @y r @z r r 0 @ 1 ðx x Þ ¼ @x r r3 @ 1 ð y y0 Þ ¼ @y r r3 @ 1 ðz z0 Þ ¼ @z r r3

r¼

360

5 Magnetic Fields in Materials

Hence 1 ðx x0 Þ ^ ðy y0 Þ ^ ðz z0 Þ r ^r $ j k ¼ 3 ¼ 2 ¼ ^i r r3 r3 r3 r r 1 @ 1 @ 1 ^ @ 1 and $0 ¼ ^i 0 þ ^j 0 þk @x r @y r @z0 r r @ 1 ð x x0 Þ ¼ 0 @x r r @ 1 y y0 ¼ 0 @y r r @ 1 z z0 ¼ 0 @z r r So that r

0

1

r

¼

^ r

r2

1 1 Thus $ ¼ $0 .

r

5.3

r

Magnetization: Bound Currents

Whenever we place a material in a magnetic ﬁeld the material becomes magnetized. Two mechanisms are important for the magnetization of material: (i) The change in orbital dipole moment due to applied magnetic ﬁeld is such that it has opposite direction of the applied ﬁeld—Diamagnetism. (ii) The dipoles (or) domains in the presence of applied magnetic ﬁeld experience a torque tending to align them in the direction of applied ﬁeld— Paramagnetism and Ferromagnetism. We deﬁne the magnetization M as “Magnetic moment per unit volume”. We will not bother about the external magnetic ﬁeld which produces the magnetization, but we will now calculate the magnetic ﬁeld produced by the magnetization M. Consider a magnetized material whose magnetization M is given. In order to calculate the magnetic ﬁeld produced by M we will ﬁrst write the expression for vector potential A for a single dipole [from Eq. 4.211]. A¼

lo m r^ 4p r2

ð5:1Þ

5.3 Magnetization: Bound Currents

361

Fig. 5.6 A magnetized material along with a small volume element is shown in the ﬁgure

A small volume element ds0 in the magnetized material carries a dipole moment M ds0 , so that the vector potential due to the entire volume is s0 (Fig. 5.6). A¼

lo 4p

Z

M r^

r2

ds0

ð5:2Þ

Note that M here is a function of source coordinates while A is a function of ﬁeld coordinaties. We know that $

0

1

r

¼

^r

r2

From Example 5.1. Hence l A¼ o 4p

Z 0 1 M$ ds0

ð5:3Þ

r

From Sect. 1.20 using vector identity (5.6) we get M$

0

1

r

1

0

0

¼ $ M$

M

r

r

ð5:4Þ

Hence substituting Eq. 5.4 in 5.3 A¼

lo 4p

Z s0

$0 M

r

ds0

lo 4p

Z s0

$0

M

r

ds0

ð5:5Þ

362

5 Magnetic Fields in Materials

From Problem (1.28) we can write for any vector G Z I $0 Gds0 ¼ G ds0 s0

S0

Hence using the above result in A we can write Z I lo 1 0 lo 1 0 ð$ MÞds þ ðM ds0 Þ A¼ 4p r 4p r s0

S

ð5:6Þ

0

The ﬁrst term appears like the potential of volume current Jp, and the second term looks like the potential of surface current Kp. Hence with ^ Jp ¼ $ M, Kp ¼ M n Z I lo Jp 0 lo Kp 0 ds þ ds A¼ 4p r 4p r s0

ð5:7Þ ð5:8Þ

S0

^ is the normal unit vector to the surface. where n The above expression shows that in order to calculate A and in turn magnetic flux density produced by magnetization M, instead of integrating the contributions from all dipoles we ﬁnd the ﬁelds produced by Jp and Kp, the bound currents and sum them up.

5.4

Physical Interpretation of Bound Currents

We have seen that whenever a material is placed in an external magnetic ﬁeld the material gets magnetized. In Sect. 5.3 we calculated the magnetic ﬁeld produced by these magnetized material. Finally we found the magnetic ﬁeld produced by the magnetized material is same as that of produced by bound currents Jp and Kp. Now let us see how these currents arise physically inside the magnetized material. Consider a thin slab of uniformly magnetized material as shown in Fig. 5.7a. The dipoles are shown by tiny current loops. In the interior of the material for two neighbouring current loops the currents in the adjacent side are in opposite direction and cancel out each other. However at the edges no such cancellation is possible. Hence the result is a surface current as shown in Fig. 5.7b. When the magnetization is non-uniform the internal currents do not completely cancel out each other. In that case we get bound currents inside the volume.

5.4 Physical Interpretation of Bound Currents

363

Fig. 5.7 A thin slab of uniformly magnetized material shown in Fig. a. Currents in adjacent dipoles depicted as tiny current loops cancel out each other resulting in a surface current as shown in Fig. b

Example 5.2 Consider an inﬁnitely long circular cylinder which is carrying a uniform magnetization M parallel to its axis. Calculate the magnetic flux density inside and outside the cylinder due to the magnetization M. Solution Consider the cylinder as shown in the Fig. 5.8. The volume bound current is given by Jp ¼ $ M ¼ 0 because the magnetization is uniform. ^ The surface bound current is Kp ¼ M n As shown in Fig. 5.8 if we rewrite the above expression in terms of cylindrical coordinates. b Kp ¼ M ^rc ¼ ðMÞ^z ^rc ¼ M / So ﬁnally we have the surface current over the inﬁnite circular cylinder as shown in Fig. 5.9. Fig. 5.8 An inﬁnitely long circular cylinder carrying a uniform magnetization

364

5 Magnetic Fields in Materials

Fig. 5.9 An inﬁnitely long circular cylinder having a surface current

Fig. 5.10 An Amperian loop around a inﬁnitely long circular cylinder

Now let us apply Ampere’s circuital law to calculate B inside and outside the cylinder in Fig. 5.9. First let us see whether there can be any circumferential components. In Fig. 5.10 we see that B/ is constant around the Amperian loop concentric with the cylinder and the net current enclosed by the loop is zero. Hence for the Amperian loop shown in Fig. 5.10. I B dl ¼ B/ ð2prc Þ ¼ lo Ienc ¼ 0 ) B/ ¼ 0 So there are no circumferential components. For similar reasons circumferential components inside the solenoid is zero. Next let us check whether there are Br components. We know that $ B ¼ 0 Integrating over a given volume s Z $ d¼0

5.4 Physical Interpretation of Bound Currents

365

Fig. 5.11 Two cylindrical surfaces shown in dotted lines either larger or smaller than the cylinder considered in the problem

Applying Gauss’s divergence theorem I )

B ds ¼ 0:

Let us apply the above equation to our cylinder shown in Fig. 5.11 where we show two cylindrical surfaces (dotted lines) either larger (or) smaller than our original cylinder. Let the length of the dotted cylindrical surfaces be L and H radius be r then the top and bottom faces do not contribute to the integral ) B ds ¼ 0 because the radial component Br is perpendicular to ds on the top and bottom faces. Hence only the curved cylindrical surface contributes to the integral. I ) B ds ¼ 0 I ) jBjjdsj ¼ 0 ) 2prc L Br ¼ 0 ) Br ¼ 0 as rc 6¼ 0; L 6¼ 0 Hence both B/ and Br are both zero inside and outside the cylinder. The only surviving component is Bz. We know that the flux density B must go to zero when we go far away from the cylinder. Now consider Fig. 5.12. Applying Ampere’s circuital law to loop 2 lying outside the cylinder

366

5 Magnetic Fields in Materials

Fig. 5.12 An inﬁnitely long circular cylinder shown along with Amperian loops

H

B dl ¼ ½BðPÞ BðQÞL ¼ lo Ienc ¼ 0 as the loop doesn’t enclose any current. ) BðPÞ ¼ BðQÞ ¼ 0 as L 6¼ 0 ) BðPÞ ¼ BðQÞ

Now we can select the loop 2 such that it lies very far away from the cylinder in which case B(Q) = 0 ) BðPÞ ¼ BðQÞ ¼ 0: Hence for all points outside the cylinder B is zero. Now let us apply Ampere’s circuital law to loop 1 then I B dl ¼ B L ¼ lo Ienc ) B L ¼ lo Ienc ¼ lo Kp L

^ : ) B ¼ lo Kp ¼ lo M is the value of B inside the cylinder As Kp ¼ M n

Example 5.3 Consider a long circular cylinder of radius R. It carries a magnetization of where b is a constant. Calculate the magnetic flux density produced by M both M ¼ j rc / inside and outside the cylinder. Solution First let us calculate the value of Jp and Kp. We have in cylindrical coordinates ^ M ¼ Mr ^rc þ Mz^z þ M/ /

5.4 Physical Interpretation of Bound Currents

367

^ then As M ¼ j rc / Mr ¼ 0; Mz ¼ 0; M/ ¼ krc The curl is cylindrical coordinates then is given by Jp ¼ $ M ¼

1 @ rc M/ ^z rc @rc

1 @ ½rc jrc ^z rc @rc 1 Jp ¼ j 2rc^z ¼ 2j^z rc Jp ¼ 2j^z Jp ¼

b ^rc ^ ¼ jrc / and Kp ¼ M n Kp ¼ jR^z as the bound current exists on the surface. Having calculated Jp and Kp now let as calculate B due to Jp and Kp or equivalenty that of M. Consider Fig. 5.13 to ﬁnd the B inside the cylinder for the Amperian loop 1. Applying Ampere’s law for this loop. B 2prc ¼ lo Ienc From Eq. 4.22 we see that Z Ienc ¼

Jp dsper

Because Jp ¼ 2j^z, meaning Jp is in z-direction.

Fig. 5.13 A long circular cylinder of radius R shown along with Amperian loops

368

5 Magnetic Fields in Materials

The perpendicular area dsper from Fig. 5.13 is dsper ¼ dl1 dl2 From Sect. 1.19 in cylindrical coordinates dl1 ¼ drc ; dl2 ¼ rc d/: Hence dsper ¼ rc drc d/ and Z Ienc ¼ 2j rc drc d/ Z2p

Zrc ¼ 2j

rc drc o

d/ o

r2 ¼ 2j c 2p ¼ 2p jr2c 2 Substituting is Ampere’s law B 2prc ¼ lo Ienc ¼ lo 2p jr2c B ¼ lo j rc To ﬁnd B outside the cylinder consider Amperian loop 2 in Fig. 5.13 which encloses both Jp and Kp. Now let us calculate Ienc in Ampere’s law. Z Z Kp dlper Ienc ¼ Jp dsper þ ZR Z2p Ienc ¼

Z2p 2j rc drc d/ þ

o

o

j RðR d/Þ 0

R Because in jp dlper , dlper is the line element perpendicular to the direction of flow of Kp. We see that Kp ¼ jR^z which means that Kp is in negative z-direction. A line element dlper is equal to dl2 in Fig. 5.13, [see Sect. 1.19—cylindrical coordinates], i.e. a line element dl2 lying on the curved outer surface of the cylinder as Kp flows on the surface. Hence dlper ¼ dl2 ¼ Rd/ Thus Ienc ¼ 2p jR2 jR2 2p: Ienc ¼ 0

5.4 Physical Interpretation of Bound Currents

369

Hence the current enclosed by the Amperian loop 2 in Fig. 5.13 is zero. Applying Ampere’s law to Amperian loop 2 in Fig. 5.13 I B dl ¼ lo Ienc ) B 2prc ¼ 0 as Ienc ¼ 0 )B¼0 Hence B is zero outside. Another interesting fact to be noted is Jp ¼ 2j^z, Kp ¼ j R^z meaning that JP flows in z-direction, i.e. current flows up the cylinder and Kp flows in negative z-direction, i.e. current returns back down the surface. Example 5.4 A cylinder of radius R and length L carries a uniform magnetization as shown in Fig. 5.14. Calculate the magnetic flux density B at point P in Fig. 5.14. The magnetization is M ^z. Solution As in Problem 5.2 Jp ¼ $ M ¼ 0 ^ Kp ¼ M ^rc ¼ M^z ^rc ¼ M/

Fig. 5.14 a, b, c A cylinder of radius R and length L carries a uniform magnetization pointing in z directionx

370

5 Magnetic Fields in Materials

From Eq. 4.157 the magnetic flux density along the axis of the single circular loop is B¼

lo I R2

2 x2 þ R2 3=2

From Fig. 5.14b. R ) x ¼ R cot a x ) dx ¼ Rcosec2 ada

tan a ¼

and R2 þ x2 ¼ r2 ; sin a ¼ R r From Fig. 5.14b I ¼ Kp dx Md and hence dB ¼

lo M R2

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ dx 2 x2 þ R2 x2 þ R2 dB ¼

lo M sin a da 2

Thus B due to the entire cylinder is lo M B¼ 2

Za1 sin a da a2

l M B ¼ o ½cos a1 cos a2 ^z 2

5.5

Magnetic Field and Ampere’s Law in Magnetized Materials

In Sect. 5.3 we saw that bound currents Jp and Kp appear in the magnetic material due to magnetization. Now there are two distinct ﬁelds present in the magentic material. One produced by the external agent and the other produced by bound currents Jp and Kp, i.e. magnetization itself. The external agent or current which is

5.5 Magnetic Field and Ampere’s Law in Magnetized Materials

371

responsible for the ﬁeld in the magnetic material let us call it as “free current” and denote it as Jf. Thus the total current can be written as J ¼ Jp þ Jf

ð5:9Þ

Hence Ampere’s law can be written as $ B ¼ lo J 1 ð$ BÞ ¼ J ¼ Jp þ Jf lo

ð5:10Þ

1 ð$ BÞ ¼ Jf þ ð$ MÞ lo 1 B M ¼ Jf )$ lo

ð5:11Þ

) )

ð5:12Þ

Let us denote H¼

1 BM lo

ð5:13Þ

Hence $ H ¼ Jf

ð5:14Þ

Here H is called the magnetic ﬁeld. In free space M = 0, hence. H¼

1 B lo

Integrating Eq. 5.14 and applying Stokes’ theorem we ﬁnally get I H dl ¼ If ðenclosedÞ

ð5:15Þ

ð5:16Þ

where If ðenclosedÞ is the total free current passing there the Amperian loop. H In Sect. 4.9 we derived a similar expression for B, B dl ¼ lo Ienc using which we calculated B in Sects. 4.16–4.21. Similarly in the presence of magnetic materials H can be calculated using the equation: I H dl ¼ If ðenclosedÞ :

372

5 Magnetic Fields in Materials

which is Ampere’s law in magnetic materials. Here H simply is determined by If ðenclosedÞ , i.e. free current. Free current is stuff which is under our control, while bound currents appear is the material on which we don’t have control. Equation 5.16 enables us to calculate H, by considering free current alone which we control. Finally H vector in magnetostatics play analogous role as D in electrostatics.

5.6

Linear and Nonlinear Media

Whenever a paramagnetic or diamagnetic material is subjected to magnetic ﬁeld the material gets magnetized. As the ﬁeld increases the magnetization in the material also increases. Hence we write MaH

ð5:17Þ

M ¼ vm H

ð5:18Þ

where the constant of proportionality vm is called magnetic susceptibility. Materials that obey Eq. 5.17 are called linear media. Then from Eq. 5.13 B ¼ lo ðH þ MÞ ¼ lo ½1 þ vm H

ð5:19Þ

B ¼ lH

ð5:20Þ

where l ¼ l0 ½1 þ vm is called the permeability of the material. ) lr ¼

l ¼ 1 þ vm lo

ð5:21Þ

where lr is called the relative permeability of the medium. For vacuum vm = 0 because there is no material to magnetize. Hence for vacuum l ¼ lo

ð5:22Þ

Hence lo is called permeability of free space. Example 5.5 Consider a wire of radius R carrying a current I and the current is uniformly distributed. The wire is made of a linear material whose susceptibility is vm. Calculate B at a distance r from axis and the bound currents.

5.6 Linear and Nonlinear Media

373

Fig. 5.15 A wire of radius R carrying a uniform current I

Solution We don’t know what the bound currents are and hence directly calculating B is not possible. But however B can be calculated as follows. Applying Ampere’s law (Fig. 5.15). I I

H dl ¼ If ðenclosedÞ H dl ¼ H 2pr

For a point outside the wire If ðenclosedÞ ¼ I. Hence Ampere’s law gives H 2pr ¼ I )H¼

I 2pr

For a linear material outside the wire. B ¼ lo H ¼

lo I 2pr

For a point inside the wire Free current density in the wire is Jf ¼

I pR2

Hence If ðenclosedÞ for a point r < R is If ðenclosedÞ ¼ I

pr2 r2 ¼I 2 2 pR R

374

5 Magnetic Fields in Materials

Hence Ampere’s law gives H 2pr ¼ I H¼

r2 R2

Ir 2pR2

We have B ¼ lH inside the material l ¼ l0 ð1 þ vm Þ Hence B ¼ lo ð1 þ vm ÞH ¼ lo ð1 þ vm Þ

Ir 2pR2

Now let us estimate the bound currents. We have Jp ¼ $ M ¼ $ ð vm H Þ Jp ¼ vm ð$ HÞ Applying Eq. 5.14 Jp ¼ vm Jf Jp ¼ vm

I pR2

with Jp having the same direction of current. Now ^ ¼ vm H n ^ Kp ¼ M n v l I Kp ¼ m o 2pr with the direction opposite to that of current.

5.7

Boundary Conditions

The boundary conditions which we derived in Sect. 4.23 can be rewritten for the case of boundary between two magnetic materials. Noting down that H1 and H2 point in same direction as B1 and B2

5.7 Boundary Conditions

375

Equation 4.225 is B1n ¼ B2n ) l1 H1n ¼ l2 H2n

ð5:23Þ

where l1, l2 are the permeability of the two magnetic mediums. In Fig. 4.33b if medium 1, 2 are magnetic mediums then applying Eq. (5.16) to the Amperian loop in Fig. 4.33b we obtain. I H dl ¼ If ðenclosedÞ Because the side a is negligible for the Amperian loop (Fig. 4.33b) H1t L H2t L ¼ Kf L

ð5:24Þ

where H1t, H2t are the tangential components of magnetic ﬁelds is medium 1,2. Kf is the free surface current in Fig. 4.33b ) H1t H2t ¼ Kf

ð5:25Þ

Example 5.6 We have a magnetic flux density ^ B ¼ 5e2y k

mwb m2

Where l = 4.7 l0. Find (a) H (b) M Solution (a) H ¼

B 5 103 e2y ¼ l 4p 107 4:5

^ A=m H ¼ 884e2y k (b) lr ¼ ll ¼ 4:7 o

vm ¼ lr 1 ¼ 4:7 1 ¼ 3:7 M ¼ vm H ¼ ð3:7Þð884Þe2y A=m M ¼ 3270:8 A=m Example 5.7 At a point of a boundary between two magnetic materials (conductors) the magnetic ﬁeld makes an angle h1 and h2 with the normal of the media of permeability l1 and l2, respectively, as shown in Fig. 5.16. Show that

376

5 Magnetic Fields in Materials

Fig. 5.16 a Boundary between two magnetic material. b Resolution of magnetic flux density vector into component

tan h1 l1 ¼ tan h2 l2 Solution Resolving into components B1n ¼ B1 cos h1 B1t ¼ B1 sin h1 B2n ¼ B2 cos h2 B2t ¼ B2 sin h2

5.7 Boundary Conditions

377

and H1n ¼ H1 cos h1 H1t ¼ H1 sin h1 H2n ¼ H2 cos h2 H2t ¼ H2 sin h2 Applying the boundary conditions. B1n ¼ B2n ) B1 cos h1 ¼ B2 cos h2

ð5:26Þ

As the materials are conductors the boundary doesn’t contain any surface current hence H1t ¼ H2t H1 sin h1 ¼ H2 sin h2

ð5:27Þ

From Eqs. 5.26 and 5.27 H1 sin h1 H2 sin h2 ¼ B1 cos h1 B2 cos h2 1 1 ) tan h1 ¼ tan h2 l1 l2 tan h1 l1 ) ¼ tan h2 l2

ð5:28Þ

where B1 ¼ l1 H1 , B2 ¼ l2 H2 From the above equation we can derive the following conclusions. (a) If h1 = 0, then h2 is also zero. That is the magnetic ﬁeld lines are normal to the boundary in both the regions. (b) If region 2 has high permeability as compared to region 1 and let h2 is less than 90°, then the above equation shows that h1 is small. That is the magnetic ﬁeld lines are normal.

378

5 Magnetic Fields in Materials

Exercises Problem 5.1 Consider an electron orbiting around the nucleus with radius R. The orbiting electron is placed in a magnetic ﬁeld B. Using the fact that the electron speeds up or slows down depending on the orientation of B show that the change in dipole moment is opposite to the direction of B. Problem 5.2 A circular loop carrying a current I is placed is a magnetic ﬁeld B as shown in Fig. 5.17. Show that the torque acting on the loop is N ¼ m B. Problem 5.3 Show that the hysteresis loss in the magnetic material is equal to the area of the hysteresis loop. Problem 5.4 In a region speciﬁed by 0 z 3m lies an inﬁnite slab of magnetic material. The value of lr is 2.4. The value of B ¼ 2y^i 7x ^j mwb m2 inside the slab. Find (a) JP (b) KP at z = 0. Problem 5.5 The average circumference of a metal ring is 140 cm. Its cross-sectional area is 9 cm2. It is carrying a current of 2.5 A and the corresponding flux is found to be 2 milli-webers. It is wound with 500 turns. Calculate the relative permeability of the material.

Fig. 5.17 A circular loop carrying a current is placed in a magnetic ﬁeld. Illustration for problem 5.2

Exercises

379

Fig. 5.18 Boundary between two magnetic materials. Illustration for problem 5.6

Problem 5.6 Two magnetic materials which are conductors are present as shown in Fig. 5.18. ^ and In region 1 lying at z > 0, H1 ¼ 5^i þ 4^j 2k lr1 ¼ 5. In region 2 lying at z < 0, lr2 ¼ 7 Find h2 and H2.

Chapter 6

Time-Varying Fields and Maxwell’s Equation

6.1

Introduction—Ohm’s law

In metals like copper the outermost valence electrons are not attached to a particular atom and are called as free electrons. These free electrons are capable of moving through the entire crystal lattice. The free electrons acquire thermal energy and wander inside the metal in all possible directions. The thermal velocities of the free electrons are in random direction and average out to zero. As a result there is no net current in an isolated conductor. However when a negative terminal of the battery is connected to one end of the conductor and the positive terminal is connected to the other end of the conductor, free electrons start flowing from the negative terminal of the battery to the positive terminal through the conductor and hence constitutes a current in the conductor. When the battery is connected to the conductor as said above, an electric ﬁeld E is established inside the conductor and the free electrons in the conductor accelerate towards the positive terminal under the action of the electric ﬁeld. However the electrons undergo collision with the ions in the lattice. The collision slows down the electron and changes its direction. The electric ﬁeld once again accelerates the electrons towards the positive terminal. The electron suffers numerous collisions with the ions of the lattice as the electric ﬁeld accelerates them towards the positive terminal of the battery (see Problem 6.2 and Fig. 6.29). The electrons slowly drift towards the positive terminal with an average velocity known as drift velocity.

6.1.1

Ohm’s law

When an electric ﬁeld is maintained across the material, electrons in the material flow in the direction opposite to the electric ﬁeld E because the electrons with charge –e experience a force given by © Springer Nature Singapore Pte Ltd. 2020 S. Balaji, Electromagnetics Made Easy, https://doi.org/10.1007/978-981-15-2658-9_6

381

382

6 Time-Varying Fields and Maxwell’s Equation

F ¼ eE

ð6:1Þ

Motion of these charges produces current. In a given material how effectively charges move to produce current depends upon the nature of the material. For most materials, the current density J is proportional to E JaE

ð6:2Þ

) J ¼ rE

ð6:3Þ

Here r the proportionality constant varies from one material to the other material and is known as conductivity of the material. Equation 6.3 is called Ohm’s law. The reciprocal of r,q is known as resistivity of the material r¼

1 q

ð6:4Þ

Conductors are materials in which r is very high. Resistors are materials in which the resisitivity q is very high. Consider a cylindrical resistor of length L whose ends are maintained at a potential difference of V (Fig. 6.1). The area of cross section of the cylindrical resistor is A. The current that flows in the cylindrical resistor is I = J A (see Eq. 4.22) Substituting Eq. 6.3 in the above equation I ¼ rE A )I¼r

V A L

ð6:5Þ ð6:6Þ

where r is the conductivity of the resistor. Equation 6.6 can be written as V¼

L qL I¼ I rA A

) V ¼ RI

ð6:7Þ ð6:8Þ

where R ¼ qAL is called the resistance of the material and Eq. 6.8 is the familiar form of Ohm’s law. R¼

qL A

ð6:9Þ

6.2 Electromotive Force

383

Fig. 6.1 A cylindrical resistor with cross sectional area A and length L

6.2

Electromotive Force

We know that work done by a conservative electrostatic force around a closed path is zero (from Eq. 2.280) I E dl ¼ 0

ð6:10Þ

Thus if there is circuit carrying current, conservative electrostatic forces alone will not be enough to maintain current in the circuit. That is, as the charges flow down the wire they collide with atoms and dissipate energy. The dissipated energy must be supplied to the charges if current is to be maintained in the circuit. The electrostatic forces cannot do any work and hence cannot supply energy to the charges. Thus in order to make up for the dissipated energy and maintain current in the circuit, non-conservative sources like batteries, thermocouples, electric generators, etc., must be present in the circuit. Consider the circuit in Fig. 6.2. The circuit contains a non-conservative source— the battery. The non-conservative source—the battery pushes the charge inside the wire when the key K is closed. As the charges are pushed inside the wire at some point (say point B in Fig. 6.2), charges are accumulating. Hence the electrostatic force of repulsion builds up at point B. Charges flowing towards B are repelled and charges flowing away from B are promoted. This electrostatic force of repulsion evens out the flow of charges. The process takes place at each point of the circuit and hence results in even flow of current I over the entire circuit. Hence as the battery pushes the charge inside the wire, in the remaining part of the circuit the electrostatic force of repulsion maintains the current. Thus there are two ﬁelds responsible for maintaining current in the circuit–one non-conservative ﬁeld and the other conservative ﬁeld, the electrostatic ﬁeld. Let the electric ﬁeld due to the non-conservative source be denoted by Es and the electric ﬁeld due to the conservative electrostatic ﬁeld be denoted by E. The total electric ﬁeld is then

384

6 Time-Varying Fields and Maxwell’s Equation

Fig. 6.2 A circuit consisting of a non-conservative source —a battery and a resistor

ET ¼ E þ Es

ð6:11Þ

Let us deﬁne the line integral of ET around the closed path as I I I ET dl ¼ E dl þ Es dl As

ð6:12Þ

H

E ¼ E dl ¼ 0 I

E¼

Es dl

ð6:13Þ

The term E is called electromotive force or emf and represents the measure of strength of the non-conservative source. In an open-circuited battery, ET is zero inside the battery, as no current flows inside the open-circuited battery. Hence from Eq. 6.11 for a open-circuited battery ET ¼ 0 ¼ E þ Es

ð6:14Þ

) E ¼ Es

ð6:15Þ

In Fig. 6.2 the potential difference between the two terminals a and b of the battery is Zb

Zb E dl ¼

V¼ a

Es dl

ð6:16Þ

a

Es the electric ﬁeld of the source is present within the source alone and Es in the wire in Fig. 6.2 is zero. Hence we can write I

Zb Es dl ¼

V¼

Es dl ¼ E

ð6:17Þ

a

H where the closed path in the integral Es dl represents path aBCRKba in Fig. 6.2. In Eq. 6.17 we have used Eq. 6.13. Thus the potential difference across the battery in open circuit is equal to the emf of the battery.

6.2 Electromotive Force

385

We know that Es is the electric ﬁeld of the non-conservative source. H By deﬁnition, the electric ﬁeld is the force per unit charge. Then the integral Es dl with Es as the force per unit charge represents work done per unit charge. Hence from Eq. 6.13 we can write I dW ¼ Es dl E¼ ð6:18Þ dq

6.3

Motional emf

In this section we will discuss about motional emf’s. Motional emf’s arise when we move a wire in a magnetic ﬁeld. As shown in Fig. 6.3 a magnetic ﬁeld of magnetic induction B is pointing into the page. A wire is connected to the resistor and part of the wire abcd lies in the magnetic ﬁeld with bc = l. Suppose assume that the wire is pulled in the x-direction. The charges in the segment bc experience a magnetic force, thereby a current I flows in the circuit. The charges in the segment ab and bc experience a magnetic force perpendicular to the wire and hence do not contribute to the current. Thus when the charges are pulled along the x-direction by an external agency a magnetic force acting along y-direction which drives the charges along ydirection responsible for a current I in the circuit develops. Therefore the charges acquire two velocity components one in x-direction (due to external agency) and the other along y-direction (due to magnetic force).

Fig. 6.3 A wire abcd lying in a magnetic ﬁeld, connected to a resistor and pulled in the xdirection by an external agency

386

6 Time-Varying Fields and Maxwell’s Equation

Because the charges acquire new velocity component along y-direction responsible for current I another magnetic force component acting along x-direction develops and is given by Fm ¼ Il B Fm acting along negative x-direction opposes the motion of the wire along xdirection. In order to make the wire to move in the x-direction, the external agency must exert the force Fext ¼ Fm ¼ Il B

ð6:19Þ

) Fext ¼ IðlBð^iÞÞ

ð6:20Þ

) Fext ¼ IBl^i

ð6:21Þ

The work done by the external force is dW ¼ IBldx ¼ IBludt

ð6:22Þ

where dx is the distance moved by the wire ab in a time dt when it is pulled with a velocity u along x-direction. But I dt = dq the charge transferred in time dt. Hence dW ¼ Budql The emf

ð6:23Þ

E is given by E¼

dW ¼ Bul dq

ð6:24Þ

Let us derive an expression relating E , the emf and the flux /. Consider a wire of arbitrary shape as shown in Fig. 6.4 being pulled by an external agency in the

Fig. 6.4 A wire of arbitrary shape being pulled in a non-uniform magnetic ﬁeld by an external agency

6.3 Motional emf

387

presence of an non-uniform magnetic ﬁeld with flux density B. Divide the wire into small elements dlW as shown in the ﬁgure. The area PP0 QQ0 is ds ¼ ðdl dlw Þ

ð6:25Þ

) ds ¼ ðu dlw Þdt

ð6:26Þ

where u is the velocity with which wire is pulled. Here we need to take ðdl dlw Þ and not ðdlw dlÞ because by convention the direction of area vector is outside normal. From Eq. 6.19 the force exerted by the external agency is I Fext ¼

Iðdlw BÞ

ð6:27Þ

C

Here C is the path of integration in the direction of current in the wire. With PP0 ¼ dl, the work done is dW ¼ Fext dl

ð6:28Þ

Substituting Eq. 6.27 in 6.28 I dlw B

dW ¼ Idl

dl dq With I ¼ dt and u ¼ Eq. 6.29 can be written as dt I dW ¼ u dlw BðdqÞ

ð6:29Þ

ð6:30Þ

C

The motional emf is then

E

dW ¼ u ¼ dq

I ðdlw BÞ

ð6:31Þ

u ðdlw BÞ

ð6:32Þ

C

(or)

E

dW ¼ ¼ dq

I C

388

6 Time-Varying Fields and Maxwell’s Equation

(or)

E¼

dW ¼ dq

I B ðu dlw Þ

ð6:33Þ

C

The change in flux associated with the area ds, is given by ¼ B ds ðfrom Eq. 4.246)

ð6:34Þ

Substituting Eq. 6.26 in 6.34 ¼ B ðu dlw Þdt The total change in flux over the entire wire is I d/ ¼ B ðu dlw Þdt

ð6:35Þ

ð6:36Þ

C

)

d/ ¼ dt

I B ðu dlw Þ

ð6:37Þ

d/ dt

ð6:38Þ

C

Comparing Eqs. 6.33 and 6.37

E¼

6.4

Faraday’s Law

Michael Faraday in 1831 demonstrated that apart from current flowing in the circuit, when the wire is pulled in the magnetic ﬁeld in Fig. 6.3, a current also flows in a stationary circuit, if the magnetic flux associated with the wire varies with respect to time. In Fig. 6.3 we have seen that when a wire is pulled in the presence of the magnetic ﬁeld in the x-direction a current flows in the circuit. Instead let the wire be held stationary and let the magnetic ﬁeld be time varying so that a time-varying magnetic flux is associated with the wire loop shown in Fig. 6.3. In that case once again current flows in the circuit and clearly it is not the magnetic force responsible for the current flow. The wire is held stationary and hence magnetic force cannot be responsible for the flow of current ½v ¼ 0 in F ¼ qðv BÞ: Faraday stated that “An induced electric ﬁeld is responsible for the flow of current in the circuit

6.4 Faraday’s Law

389

whenever the wire is held stationary”. The induced emf due to the changing magnetic ﬁeld is once again given by I

E¼ We know that / ¼ Hence

R S

d/ dt

E dl ¼

ð6:39Þ

B ds (From Eq. 4.246). I E dl ¼

d dt

I )

Z B ds

ð6:40Þ

@B ds @t

ð6:41Þ

S

Z E dl ¼ S

Equation 6.41 is Faraday’s law in integral form. Applying Stoke’s theorem Z

Z ð$ EÞ ds ¼

S

S

@B ds @t

ð6:42Þ

Because the above equation is valid for an small area ds, for the area ds, we need not integrate ð$ EÞ ds ¼ )$E¼

@B ds @t

ð6:43Þ

@B @t

ð6:44Þ

Equation 6.44 is Faraday’s law in differential form. Let us compare Eq. 6.44 with the electrostatic counterpart $E¼0 That is the induced electric ﬁeld given by Eq. 6.44 curl’s around (or) E associated with changing magnetic flux forms closed loops. However the lines of E associated with static charges do not form closed loops, but instead they originate on a positive charge and end on a negative charge or go of the inﬁnity.

390

6.5

6 Time-Varying Fields and Maxwell’s Equation

Lenz’s Law

Faraday’s law relates the induced electric ﬁeld to the rate of change of magnetic flux density. Now we will discuss about the direction of the induced emf and current which is given by Lenz’s law. It state’s that the direction of the induced current or emf in a closed circuit is such that it opposes the very cause that produces it. In other words the law states that current induced is in the direction, such that it produces a magnetic flux tending to oppose the original change of flux. Suppose assume that the applied flux density B is decreasing in magnitude, then the current induced in the loop will be in a direction such that it will produce a ﬁeld to increase B. If on the other hand when B is increasing the current induced in the loop will be in a direction such that the ﬁeld produced by the induced current will tend to decrease B. Suppose assume as shown in Fig. 6.5 a magnet approaches a loop. When the magnet approaches the loop the induced current Ii set up in the loop has to produce a ﬁeld so that the induced ﬁeld produced by the incoming magnet in the loop opposes the magnetic ﬁeld of the incoming magnet. Hence the induced current flows as shown in Fig. 6.5 so that ﬁeld produced by the current acts leftward to oppose the increase in ﬁeld produced by the incoming magnet. On the other hand, if the magnet is withdrawn then the ﬁeld in the loop decreases and the induced current flows in the opposite direction as shown in Fig. 6.6 so that the ﬁeld produced by the induced current is rightward in order to oppose the decrease in magnetic ﬁeld due to withdrawal of magnet. Example 6.1 Consider a coaxial cable of length L with the intermediate material whose conductivity is r. The inner radius of the cylinder is rc1 and the outer radius is rc2 as shown in Fig. 6.7. A current I flows from the inner conductor to the outer conductor. Find the resistance R between the inner and outer conductors? Solution The current density at an arbitrary point rc between the inner and outer conductors is I ¼ I . J¼A 2prc L Fig. 6.5 Magnet moving towards the loop. The direction of the induced current in the loop is shown

Fig. 6.6 Magnet withdrawn from the loop. The direction of the induced current in the loop is shown

6.5 Lenz’s Law

391

Fig. 6.7 A coaxial cable of length L, ﬁlled with an intermediate material and carrying a current I

The electric ﬁeld E is then J ¼ rE J r I ^rc E¼ 2prrc L )E¼

The potential difference between the conductors is Zrcl V¼

E dl rc2 Zrc1

¼

Edl rc2 Zrcl

¼ rc2

I I rc2 ¼ ln 2prrc L 2prc L rcl

Then the resistance is V 1 rc2 R¼ ¼ ln I 2prL rc1

392

6 Time-Varying Fields and Maxwell’s Equation

Fig. 6.8 Cross-sectional view of Fig. 6.7

Alternate Method Consider Fig. 6.8. Let rc and rc þ drc be the radii of the concentric layer of the intermediate material. The surface area of the concentric layer is 2prc L: From Eq. 6.9 R¼

qL A

1 is the resistivity or speciﬁc resistance of the material. r Equation 6.9 can be written for the concentric layer of intermediate material as

where q ¼

dR ¼

qdrc 2prc L

For the entire material Zrc2 R¼q rcl

q R¼ 2pL

drc 2prc L Zrc2

drc rc rc1 1 rc2 ln R¼ 2prL rc1 Example 6.2 In Example 6.1 rc1 = 0.9 cm and rc2 = 1.4 cm. The intermediate material is having a resistance per unit length of 700 MX/km. What is the resistivity of the intermediate material. If the resistance is supposed to be increased to 1600 MX/km by increasing the thickness of the same intermediate material, how much should be the increase in thickness?

6.5 Lenz’s Law

393

Solution We know that R¼

q rc2 ln 2pL rcl

rc2 1:4 ¼ rc1 0:9 L = 1 km = 105 cm. R = 700 MX. Hence Now

700 ¼

q 1:4 ln 5 2p10 0:9

4396 105 ¼ qð0:44183Þ ) q ¼ 9:9 108 MX-cm: Total insulation has to be 1600 MX/km. Hence the remaining insulation that needs to be increased is 1600–700 = 900 MX/km. Now 900 ¼

9:9 108 rc2 þ x ln rc2 2 3:14 105

Here rc2 is the outer radius of the existing intermediate material (or) inner radius of the additional intermediate material added. rc2 + x is the outer radii of the additional intermediate material added. rc2 = 1.4 cm Hence 9:9 108 1:4 þ x 900 ¼ ln 1:4 2 3:14 105 1:4 þ x 570 103 ¼ ln 1:4 1:4 þ x 1:769 ¼ 1:4 x ¼ 1:078 cm Example 6.3 A metal disc revolves about 15 revolutions per second about the horizontal axis passing through its centre. The disc placed vertically has a diameter of 30 cm. A uniform magnetic flux density of strength 150 Gauss acts perpendicular to the plane of the disc? Calculate the induced emf.

394

6 Time-Varying Fields and Maxwell’s Equation

Solution We know that d/ d ¼ ðBAÞ dt dt dA ðNumerically) ¼B dt

E¼ E

Given B = 150 Gauss = 150 10−4 Wb/m2 r = 15 cm = 0.15 m dA ¼ Area swept by the disc in unit time: dt ¼ pr 2 No: of revolutions per second ¼ ð3:14Þð0:15Þ2 15 ¼ 1:05975 Hence

E ¼ B dA dt E ¼ 150 104 1:05975 E ¼ 15:896 mV:

Example 6.4 A circular path of initial radius 50 cms is centreed at the origin of the x–y plane. The circular radius gradually increases at a rate of 70 m/s. A uniform magnetic flux density of 100 Gauss is directed along the z-axis. Determine the induced emf is the circular path as a function of time. Solution The induced emf is d/ d/ ¼ ðNumericallyÞ dt dt / ¼ Bðpr2 Þ dr E ¼ Bpr dt B ¼ 100 Gauss ¼ 102 Wb=m2

E¼

dr ¼ 70 m=s dt ¼ 102 ð3:14Þð0:5Þð70Þ

r ¼ 0:5 m

E E ¼ 1:099 V

6.5 Lenz’s Law

395

Example 6.5 A metal wire of length 1 m lies perpendicular to the x–y plane. It is moving ^ m/s. The magnetic flux density in the region is with a velocity u ¼ ð4^i þ 3^j þ 2kÞ 2 ^ ^ B ¼ 3i þ 4j Wb/m . Determine the potential difference the ends of the wire. Solution From Eq. 6.33 for the entire length of the wire L

E ¼ ðu BÞ L

^i ^j k ^ u B ¼ 4 3 2 3 4 0 ^ ¼ ^ið0 8Þ ^jð0 6Þ þ kð16 9Þ ^ u B ¼ 8^i þ 6^j þ 7k E ¼ ðu BÞ L ^ k ^ ¼ ð8^i þ 6^j þ 7kÞ ) E ¼ 7V Example 6.6 Consider Fig. 6.9 A metal rod AB moves on ﬁxed conductors CA and DB with a velocity of 5 m/s along the x-direction. A uniform magnetic flux density B of strength 0.7 Wb/m2 acts along the negative z-direction. (a) Calculate the magnitude and direction of the induced emf. (b) Calculate the current in the circuit if the resistance of ABCD is 0.25 X. (c) Calculate the force required so that the motion of metal rod AB is maintained.

Fig. 6.9 A metal rod Ab moves on a ﬁxed conductors CA and DB in a uniform magnetic ﬁeld

396

6 Time-Varying Fields and Maxwell’s Equation

Solution

E ¼ ðu BÞ L ½Numerically ^ L ¼ 0:6^j u ¼ 5^i B ¼ 0:7k ^i ^j ^ k u B ¼ 5 0 0 (a) 0 0 0:7 ^ 0Þ u B ¼ ^ið0 0Þ ^jð3:5 0Þ þ kð0 u B ¼ 3:5^j ) E ¼ ðu BÞ L ) E ¼ 2:1 V The direction of induced emf is from B to A in Fig. 6.9 as per Lenz’s law. 2:1 ¼ 8:4 A 0:25 (c) The force required is F ¼ ILB ¼ ð8:4Þð0:6Þð0:7Þ F ¼ 3:528 N:

(b) I ¼ V R¼

Example 6.7 A very lengthy solenoid has 150 closely and evenly wound turns per cm of its length. The diameter of the solenoid is 7 cm. Exactly at the centre of the solenoid a small coil of 60 turns and 5 cm diameter is placed such that the magnetic ﬁeld produced by the solenoid is parallel to the axis of the coil. The current in the solenoid is changed from 2.5 A to zero and then raised to 2.5 A in the opposite direction at a steady rate over a time period of 0.1 s. Calculate the induced emf in the coil during this change. Solution For a solenoid we know that B ¼ lo

N I L

Here lo = 4p 10−7 Wb/A m N ¼ 150 turns=cm ¼ 150 102 turns=m L I ¼ 2:5 A B ¼ 4p 107 150 102 ð2:5Þ B ¼ 4710 105 Wb/m B ¼ 4:71 102 Wb=m2

2

6.5 Lenz’s Law

397

Area of central coil ¼ ðNo: of turnsÞ area of each turn 2 ¼ ð60ÞðpÞ 2:5 102 ¼ 0:11775: The initial flux is / = BA / ¼ 4:71 102 0:11775 / ¼ 0:5546 102 Wb: When the current is changed from 2.5 to –2.5 A the flux will change from 0.5546 10−2 to –0.5546 10−2 Wb. The net change in flux is then d/ ¼ ½0:5546 ð0:5546Þ 102 d/ ¼ 1:1092 102 Wb: The magnitude of induced emf is then d/ ½Numerically dt 1:1092 102 ¼ 0:1 ¼ 11:092 102 V ¼ 1:1092 m V:

E¼ E E E

6.6

Magnetic Circuits

We come across number of instruments in which magnetism ﬁnds application. Few instruments to mention are transformers, magnetic recording devices, motors, generators, etc. Similar to electric circuit problems where we ﬁnd voltages and currents in various branches of the circuit, we are required to ﬁnd magnetic fluxes, magnetic ﬁeld intensities is various parts of the so-called magnetic circuits. Magnetic circuit problems appear is above-mentioned devices. As an example of a magnetic circuit consider the arrangement shown in Fig. 6.10. A toroidal coil of N turns is wound on a ring of iron of uniform permeability l. The toroidal coil carries a current of I amperes and the cross-sectional area of iron ring is A. The magnetic flux / is conﬁned to the iron ring and has the same value all around the ring, which is similar to current I in the electric circuit. That is the current I is conﬁned to the electric circuit and has the same value all along the circuit. Hence it can be considered that magnetic flux / is the magnetic circuit is analogous to current I is the electric circuit. The magnetic counterpart to the electromotive force (e.m.f) is the magnetomotive force (m.m.f).

398

6 Time-Varying Fields and Maxwell’s Equation

Fig. 6.10 A Toroid carrying a current I

From Eq. 6.39 we have I

E¼

E dl

ð6:45Þ

where E is the e.m.f By analogy we deﬁne m.m.f as I m.m.f ¼

H dl

ð6:46Þ

Applying Ampere’s circuital law to the dotted path shown in Fig. 6.10 we get I H dl ¼ NI ð6:47Þ Comparing Eqs. 6.46 and 6.47 I m:m:f ¼

H dl ¼ NI

ð6:48Þ

Along the dotted path H and dl are parallel and also we have / = BA, hence B ¼ lH )H¼

ð6:49Þ

B / ¼ l lA

ð6:50Þ

From Eqs. 6.46 and 6.50 I m.m.f ¼

I H dl ¼

Hdl

ð6:51Þ

6.6 Magnetic Circuits

399

I m.m.f ¼

/ dl lA

As we expect / to be constant all over circuit I dl m.m.f ¼ / lA

ð6:52Þ

ð6:53Þ

m.m.f )/¼ H dl lA

ð6:54Þ

m.m.f NI ¼H )/¼ H dl dl lA lA

ð6:55Þ

From Eq. 6.48 we can write

In a pure electric circuit we can write Ohm’s law as Current ¼ emf/Resistance

ð6:56Þ

H dl in magnetic circuit plays lA equivalent role of resistance is electric circuit. Then From 6.55 and 6.56 we see that the quantity I R¼

dl lA

ð6:57Þ

is called reluctance of magnetic circuit. Suppose assume that the area of cross section is constant then H R¼

dl L ¼ lA lA

ð6:58Þ

where L is the total length. Reluctance behaves similar to resistance. The total reluctance in series combination is R¼

X Li X ¼ Ri li Ai i i

The total reluctance is parallel is

ð6:59Þ

400

6 Time-Varying Fields and Maxwell’s Equation

Fig. 6.11 A toroidal coil wound on a iron ring with a small air gap d carries a current I

1 X1 ¼ R Ri i

ð6:60Þ

As mentioned previously the magnetic circuit like transformers, relays ﬁnds importance in practical applications. The magnetic flux stored in the particular part of the machine can be calculated using the relation /¼

NI L lA

ð6:61Þ

Now as an example let us investigate the effect of air gap left in the magnetic circuit. Consider Fig. 6.11. A iron ring of cross-sectional area a with a small air gap d wound with a toroidal coil with N turns carries current I as shown in the ﬁgure. d lo a

ð6:62Þ

ld la

ð6:63Þ

The reluctance of air gap ¼ The reluctance of iron ¼ where l is the total length of air gap and iron. The total reluctance is Total reluctance ¼

d ld ðEq. 6:59Þ þ lo a la

ð6:64Þ

NI d þ ld lo a la

ð6:65Þ

The magnetic flux is then given by /¼

The above expression applies to a very narrow gap. In the case of wide gap, the lines of force spread out and the area of cross section varies considerably. Suppose assume that a magnetic circuit consists of two or more sections of magnetic materials as shown in Fig. 6.12a.

6.6 Magnetic Circuits

401

Fig. 6.12 a A magnetic circuit consisting of many sections of magnetic material. b Equivalent circuit for Fig. 6.12a.

The circuit shown in Fig. 6.12a can be represented in terms of reluctance is equivalent form as shown in Fig. 6.12b. The total reluctance of the circuit in Fig. 6.12b can be obtained by series and parallel combination of reluctances. Example 6.8 Consider a magnetic circuit as shown in Fig. 6.13. The entire magnetic circuit is made of cast iron with lr = 1400. The length and area of the different sections is (see Fig. 6.13)

Fig. 6.13 A magnetic circuit made of cast iron. Illustration for Example 6.8

402

6 Time-Varying Fields and Maxwell’s Equation

L1 = L3 = 16 cm, L2 = L4 = 14 cm and corresponding area’s A1 (section AB) = A3 (section CD) = 7 cm2 A2 (section BC) = A4 (section DA) = 5 cm2 Number of turns N1 = 300 N2 = 200 and / = 150 10−6 Wb. Determine I1 if I2 = 0.5 A. Solution The four magnetic paths form series connection Reluctance R = R1 + R2 + R3 + R4 L1 L3 ¼ lo lr A1 lo lr A3 1:6 102 1 R1 ¼ R3 ¼ 7 4p 10 1400 7 104 16 109 ¼ 4p 1400 R1 ¼ R3 ¼

R1 ¼ R3 ¼ 9:1 105 L2 L4 R2 ¼ R4 ¼ ¼ lo lr A2 lo lr A4 14 102 R2 ¼ R4 ¼ 7 ð4p 10 Þð1400Þð5 104 Þ 14 109 R2 ¼ R4 ¼ ð4pÞð1400Þð5Þ R2 ¼ R4 ¼ 1:59 105 Total reluctance R ¼ 9:1 105 þ 1:59 105 þ 9:1 105 þ 1:59 105 R ¼ 21:38 105 Total mmf is m.m.f ¼ R/ ¼ 21:38 105 150 106 ¼ 21:38 15 m.m.f ¼ 320:7

6.6 Magnetic Circuits

403

The currents I1 and I2 produce m.m.f of opposite signs, hence m.m.f ¼ N1 I1 N2 I2 320:7 ¼ 300I1 200ð0:5Þ ) I1 ¼ 1:4 A: Example 6.9 Calculate the magnetic flux in an electromagnet. Solution An electromagnet consists of four parts. (1) (2) (3) (4)

A yoke Two limbs Two pole pieces An air gap as shown in Fig. 6.14.

Let L1 be the effective length, l1 permeability and A1 is the cross-sectional area of the yoke. The reluctance of the yoke is R1 ¼

L1 l1 A1

If L2, l2, A2 are the length, permeability and area of cross section of the limbs then reluctance of the limbs is

Fig. 6.14 An electromagnet

404

6 Time-Varying Fields and Maxwell’s Equation

R2 ¼

2L2 l2 A2

If L3, l3, A3 are the length, permeability and area of cross section of the pole pieces R3 ¼

2L3 l3 A3

R4 ¼

L4 lo A4

Similarly for air gap

Total reluctance R ¼ R1 þ R2 þ R3 þ R4 L1 2L2 2L3 L4 R¼ þ þ þ l1 A1 l2 A2 l3 A3 lo A4 Hence magnetic flux

6.7

/¼

NI R

/¼

NI L1 þ 2L2 þ 2L3 þ L4 l1 A1 l2 A2 l3 A3 lo A4

Induction—Self Induction

So far we have seen how an induced emf can be produced by the action of an externally applied magnetic ﬁeld. In this section, we will see that the induced emf is produced in a loop as a result of varying current in the loop itself, the phenomenon known as self-induction. We know that Z /¼

B ds

ð6:66Þ

S

and l I B¼ o 4p

I

dl ^r

r2

ð6:67Þ

6.7 Induction—Self Induction

405

by Biot–Savart law. Suppose assume a loop carries current I producing a magnetic flux density B. If I changes then B changes, and hence / linked to the same loop changes. As / linked to the loop changes, due to the change in the current I carried by the same loop, then according to the equation

E¼

d/ dt

ð6:68Þ

an induced emf is produced in the loop which is responsible for establishing B or /. The induction is called self-induction because the emf is in the same loop which is responsible for the establishing flux /. From Eq. 6.67 B is proportional to I and hence from 6.66, because / is proportional to B we have //I

ð6:69Þ

/ ¼ LI

ð6:70Þ

(or)

From Eq. 6.68 then

E ¼ L

dI dt

ð6:71Þ

Here the constant of proportionality L is known as coefﬁcient of self-inductance. It depends upon the geometry of the circuit and the permeability of the medium is which the loop is immersed.

6.8

Induction—Mutual Induction

In Fig. 6.15 we show two loops in which loop 1 is carrying current I1, and hence a magnetic flux density B1 is produced at the site of loop 2. The total flux through the loop 2 is then Z /21 ¼

B1 ds2

ð6:72Þ

S2

where S2 is the area of loop 2. We have B1 ¼

lo I1 4p

I

dl1 ^r

r2

ð6:73Þ

406

6 Time-Varying Fields and Maxwell’s Equation

Fig. 6.15 Mutual Induction: Two loops loop 1, loop 2 situated as shown in the ﬁgure and loop 1 carrying current I1

by Biot–Savart law. Thus as I1 changes B1 changes and /21 the flux linked to the loop 2 due to flux density B1 changes. Thus /21 produced by the loop 1 is proportional to I1 /21 / I1

ð6:74Þ

/21 ¼ M21 I1

ð6:75Þ

(or)

The constant of proportionality M21 is known as coefﬁcient of mutual inductance. An emf is induced in the loop 2 given by

E 21 ¼

d/21 dt

ð6:76Þ

From Eq. 6.75

E 21 ¼ M21

dI1 dt

ð6:77Þ

ð$ A1 Þ ds2

ð6:78Þ

We have Z /21 ¼

Z B1 ds2 ¼

S2

S2

Applying Stokes theorem I /21 ¼

A1 dl2

ð6:79Þ

6.8 Induction—Mutual Induction

407

From Eq. 4.112 A1 ¼

lo I1 4p

I

dl1

ð6:80Þ

r

Substituting 6.80 in 6.79 /21

l I1 ¼ o 4p

I I

dl1

r

dl2

ð6:81Þ

Comparing 6.81 to 6.75 M21 ¼

lo 4p

I I

dl1 dl2

r

ð6:82Þ

If we pass the same current in loop 2 instead of loop 1 we can show that M12 ¼

lo 2p

I I

dl1 dl2

r

ð6:83Þ

or in other words M12 ¼ M21 ¼ M

ð6:84Þ

That is irrespective of whatever loop carries the same current, same flux is linked with the other loop. Example 6.10 Consider a solenoid of length L with air core. The total number of turns is N and area of each turn is A. Calculate the self-inductance of the solenoid if the current I flows in the solenoid. Solution We know that the magnetic flux density of a solenoid is given by B ¼ lo

N I L

Then the flux through each turn is BA. l NI ¼ o A [By substituting the value of B] L Then the total flux through the entire solenoid is lo NIA N L l N2 IA /¼ o L

/¼

408

6 Time-Varying Fields and Maxwell’s Equation

Now

E

d/ d lo N2 IA ¼ ¼ dt dt l

E¼

lo N2 A dI l dt

Comparing the above equation with equation L¼

E ¼ L dI dt we get

lo N2 A l

Example 6.11 Consider a circuit consisting of two parallel wires P and Q separated by a distance d as shown in Fig. 6.16. The current I goes through P and returns via Q. The radius of each wire is r. Calculate the self-inductance. Solution Consider a small element dy at a distance y from P and d – y from Q. Noting down l I the B due to a thin wire at a distance R carrying current I is o . 2pR B due to the wires P and Q at dy is B¼

l0 I l0 I þ 2py 2pðd yÞ

d/ through dy is d/ ¼ Bdy

Fig. 6.16 Two parallel wires P and Q carrying current I and separated by distance d

6.8 Induction—Mutual Induction

409

and then the total flux is Zd /¼ 0

Zdr 1 1 lo I þ dy d/ ¼ y d y 2p r

1 1 þ dy y dy r lo I dr ln /¼ p r

/¼

lo I 2p

Zdr

d/ lo d r dI ¼ ln dt r dt p LdI Comparing with Eq. 6.71 E ¼ dt

Now

E¼

L¼

lo dr ln r 4p

Example 6.12 Consider a toroid of rectangular cross section as shown in Fig. 6.17. A coil of N number of turns is wound on it and a current I flows in it. Let a and b are the internal and external radius, respectively. Calculate the self-inductance of the toroid. Solution Applying Ampere’s law I B dl ¼ lo NI

Fig. 6.17 A toroid of rectangular cross-section

410

6 Time-Varying Fields and Maxwell’s Equation

For the path of radius r Bð2prÞ ¼ lo NI l NI )B¼ o 2pr Considering a rectangular strip of width dr and height h at a distance r from the centre the magnetic flux linked with the strip is d/ ¼ BdA l NI ¼ o hdr 2pr as dA = hdr Thus flux linked with one turn of the toroid is Zb /¼

l NIh d/ ¼ o 2p

a

Zb

dr r

a

l NIh loge b=a /¼ o 2p Total flux with N turns ¼ N/ ¼

lo N2 Ih loge b=a 2p

Also the total flux with N turns is = L I l N2 h loge b=a Hence L ¼ o 2p Example 6.13 Consider two concentric solenoids Q1 and Q2 closely wound upon each other as in Fig. 6.18. Q1 is the primary solenoid with nP turns per unit length and Q2 is the secondary solenoid with NS turns. AP and AS are the area of cross of the Q1 and Q2, respectively. A current I flows in the primary. Assuming no leakage in flux calculate the mutual inductance. Fig. 6.18 Two concentric solenoids wound upon each other. Q1 is the primary solenoid and Q2 is the secondary solenoid

6.8 Induction—Mutual Induction

411

Solution The magnetic flux density due to the current I in the primary is B ¼ lo n p I The magnetic flux is the product of B and area of cross section. Hence lo nP AP I is the flux linked with each turn of the secondary. As there is no leakage of flux the total flux linked with NS turns of the secondary is / ¼ lo nP AP INS The induced emf

E¼

d/ dt

¼ lo nP NS AP

dI dt

We know that

E ¼ M

dI dt

Hence comparing M ¼ l0 nP NS AP is the mutual inductance. Example 6.14 Suppose assume that there are two windings in a toroid with one being primary having N1 turns and the other being secondary with N2 turns. Then show that mutual inductance between both the windings is M¼

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ L1 L2

where L1 and L2 are the self-inductance of primary and secondary, respectively. The radius of the toroid is R. Assume no flux leakage.

Fig. 6.19 A toroid with two windings—A primary winding and a secondary winding

412

6 Time-Varying Fields and Maxwell’s Equation

Solution Suppose assume that in Fig. 6.19 CC is the primary and DD is the secondary and a current I1 flows in primary. The magnetic flux density due to the ﬁrst winding is B1 ¼

lo N1 I1 2pR

Let /11 be the flux through the primary winding then /11 ¼ B1 N1 A ¼

lo N21 AI1 2pR

Also /11 ¼ L1 I1 Comparing the above two equations L1 ¼

lo N21 A 2pR

The flux through the secondary winding is /21 ¼ B1 N2 A ¼

lo N1 N2 AI1 2pR

Also /21 ¼ M21 I1 Comparing the above two equations M21 ¼

lo N1 N2 A 2pR

Suppose a current I2 flows is the secondary instead of I1 in primary then following the same procedure lo N22 A 2pR l N1 N2 A ¼ o 2pR

L2 ¼ M12 We see M12 ¼ M21 ¼ M and L1 L2 ¼ Hence M ¼

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ L1 L2 .

l2o N21 N22 A2 ¼ M2 4p2 R2

6.9 Energy Stored in the Circuit in Terms of Self-Inductance

6.9

413

Energy Stored in the Circuit in Terms of Self-Inductance

We have seen in Sect. 6.7 that a loop possesses self-inductance. According to Lenz’s law this self-inductance will oppose any change in current in the loop. Hence this self-inductance is called back emf. Suppose assume a circuit connected to a battery and a key and the key is just closed. The current builds up in the circuit. The current is charging in the circuit and as per Lenz’s law a back emf is induced which oppose the growth in current. In order to build up the current up to a steady value I, work has to be done against this back emf. This work done remains hidden in the circuit. When the key in the circuit is opened current is decreasing and once again a back emf is produced which tries to maintain the current in the circuit. Thus the work done by the current during the growth is recovered. The work done against the back emf on a unit charge is −E on a single round around the circuit. The negative sign is included to note that work is done by the external agency against the back emf. Work done per unit time is

From Eq. 6.71

dW ¼ E I dt

ð6:85Þ

dW dI ¼ LI dt dt

ð6:86Þ

E ¼ L dI dt

The total work done in increasing the current from zero to maximum value of current I is ZI W¼

LI

dI dt dt

ð6:87Þ

0

1 W ¼ LI2 2

6.10

ð6:88Þ

Energy Stored in Magnetic Fields

The work done expressed in terms of self-inductance in Eq. 6.88 can be expressed in terms of magnetic ﬁeld. We know that

414

6 Time-Varying Fields and Maxwell’s Equation

Z

Z

/¼

B ds ¼ S

ð$ AÞ ds

ð6:89Þ

S

Applying Stokes theorem I /¼

A dl

ð6:90Þ

A dl

ð6:91Þ

As / ¼ LI I LI ¼ Substituting 6.91 in 6.88 1 W¼ I 2

I A dl

As I, dl always point in same direction I 1 ðA IÞdl W¼ 2 The above equation is for line current. For volume current Z 1 W¼ ðA JÞds 2

ð6:92Þ

ð6:93Þ

ð6:94Þ

s

With Ampere’s law $ B ¼ lo J W¼

1 2lo

Z A ð$ BÞds

ð6:95Þ

s

From Sect. 1.20 identity 8 is $ ðA BÞ ¼ B ð$ AÞ A ð$ BÞ ) A ð$ BÞ ¼ B ð$ AÞ $ ðA BÞ As B ¼ $ A ) A ð$ BÞ ¼ B B $ ðA BÞ ) A ð$ BÞ ¼ B2 $ ðA BÞ

ð6:96Þ

6.10

Energy Stored in Magnetic Fields

415

Substituting Eq. 6.96 in Eq. 6.95 2 3 Z Z 1 4 B2 ds $ ðA BÞds5 W¼ 2lo

ð6:97Þ

s

Applying Gauss’s divergence theorem 2 3 Z I 1 4 W¼ B2 ds ðA BÞ ds5 2lo s

ð6:98Þ

S

In Eq. 6.94 the integration is to be taken over volume s which encloses current J. But larger volumes are also acceptable because J is zero over there. Hence a larger volume can be selected. When we select a large volume, however, surface integral in Eq. 6.98 vanishes. Because |A| falls of as 1r ; |B| falls off as r12 , and hence magnitudes of A B decreases as r13 while the surface S increases as r2 . The entire H term S ðA BÞ ds falls off as 1r hence when we select entire volume then the surface integral goes to zero leaving W¼

1 2lo

Z B2 ds s

ð6:99Þ

Example 6.15 Consider two coils of self-inductance L1 and L2, respectively, as shown in Fig. 6.20. A current I1 and I2 flow in the each coil. Let M be the mutual inductance between the coils. Calculate the energy stored in the system. Consider both cases when the fluxes due to I1 and I2 are in either in same or opposite direction. Solution Let E 1 and

E 2 be the induced emf is coil 1 and coil 2, respectively. Then

Fig. 6.20 Two coils of self inductance L1 and L2 respectively placed close to each other and carrying currents I1 and I2 respectively

416

6 Time-Varying Fields and Maxwell’s Equation

dI1 dI2 M ðNumerically) dt dt dI2 dI1 M ðNumericallyÞ ¼ L2 dt dt

E 1 ¼ L1 E2

where +M stands for when the fluxes due to I1 and I2 are in same direction in each coil and –M when the fluxes due to I1 and I2 are in opposite direction. We know that from Eq. 6.85 dW ¼ E 1 I1 þ E 2 I2 ðNumerically) dt Substituting for

E1; E2

dW ¼ dt

dI1 dI2 dI2 dI1 L1 M M I1 þ L 2 I2 dt dt dt dt

) dW ¼ ðL1 dI1 MdI2 ÞI1 þ ðL2 dI2 MdI1 ÞI2 dW ¼ L1 I1 dI1 þ L2 I2 dI2 dðMI1 I2 Þ Integrating we get the energy stored in the system 1 1 ¼ L1 I21 þ L2 I22 MI1 I2 2 2 The ﬁrst two terms corresponds to the self-inductance of each coil and the third term corresponds to mutual inductance between two coils. Example 6.16 In Sect. 4.2 we saw that magnetic forces don’t do any work. We will discuss this point in detail here. Suppose assume that a wire ABCD is pulled in the presence of magnetic ﬁeld, the ﬁeld pointing into the plane of the paper. The wire is pulled with a velocity vp as shown in Fig. 6.21. The segments BC and AD experience magnetic forces in opposite direction which cancel out. The charges in the segment AB experience a magnetic force from A to B and hence a current I flows in the circuit. The entire picture is shown in Fig. 6.22. The work done per unit charge by the magnetic force fm at the instant wire is pulled is I ð6:100Þ Wq ¼ f m dl ¼ vp BL where fm is force per unit charge and AB = L. But we know that magnetic forces don’t do any work. Let us analyse the complete situation. As the wire is pulled with a velocity vp, the charges in the wire are pulled along with the velocity vp as shown in Fig. 6.22, and hence a magnetic force per unit charge fm acts on the charges.

6.10

Energy Stored in Magnetic Fields

417

Fig. 6.21 A wire ABCD is pulled in the presence of the magnetic ﬁeld

Fig. 6.22 The magnetic force acting on the charges in the segment AB of the wire

The charges in the wire are pulled horizontally by the force fp where fp is the pulling force per unit charge. But, however, the magnetic force pushes the charges from A to B vertically. Hence the resultant displacement sinLh of the charges will be as shown in Fig. 6.23 as the wire moves. The resultant velocity of the charges vq will be along resultant displacement sinLh of the charges as shown in Fig. 6.24 and makes an angle h with fp. Because the magnetic force fm is perpendicular to the velocity it no longer points in the vertical direction as shown in Fig. 6.22 but tilts by the same angle h as shown in Fig. 6.24. From Fig. 6.24 we see that fm has a component fm sinh which opposes fp. Hence in order to make the wire to move in its direction fp has to exert a minimum force of f p ¼ f m sin h

ð6:101Þ

vp ¼ vq cos h

ð6:102Þ

From Fig. 6.24 we see that

418

6 Time-Varying Fields and Maxwell’s Equation

Fig. 6.23 Resultant displacement of the charge in the segment AB due to different forces acting on the charge

Fig. 6.24 Resolution of magnetic force and the resultant velocity of the charge into components

The work done per unit charge by the pulling force fp is I Wp ¼ f p dl

ð6:103Þ

As f p ¼ f m sin h

L ) Wp ¼ f m sin h cos h sin h

ð6:104Þ

) Wp ¼ f m L cos h

ð6:105Þ

Wp ¼ vq B ðL cos hÞ

ð6:106Þ

Wp ¼ vq cos h ðBLÞ

ð6:107Þ

Substituting for fm

6.10

Energy Stored in Magnetic Fields

419

Wp ¼ vp BL

ð6:108Þ

(In equation for Wq we used vp for fm as in that instant the charges are moving with velocity vp. But in the equation for Wp we are using velocity vq in fm because now charges are moving with velocity vq). Comparing Eqs. 6.100 and 6.108 Wp ¼ Wq Thus the actual work is done by the pulling agent and magnetic force takes credit for it. Then what is the role of magnetic force. When the wire is pulled by fp the charges move with a horizontal velocity vp. Magnetic force simply adds a vertical component vq sin h to the horizontal component vq cos h ¼ vp , thereby producing a current I in the wire.

6.11

Maxwell’s Equation

We have four important laws ðaÞ

$E¼

q eo

ð6:109Þ

ðbÞ

$B¼0

ð6:110Þ

ðcÞ ðdÞ

@B @t

ð6:111Þ

$ B ¼ lo J

ð6:112Þ

$E¼

The equation a, the Gauss’s law and equation b state the divergence of electric and magnetic ﬁeld’s, respectively, while equation c the Faraday’s law and equation d the Ampere’s law state their curls. Taking divergence of equation c @B @ ð$ BÞ $ ð$ EÞ ¼ $ ¼ @t @t

ð6:113Þ

The left land side of the above equation is zero because the divergence curl of any vector is zero (Example 1.8). The right-hand side of the equation is zero by virtue of equation b

420

6 Time-Varying Fields and Maxwell’s Equation

Taking divergence of equation d $ ð$ BÞ ¼ lo $ J

ð6:114Þ

The left-hand side of the above equation is zero because divergence of curl of the vector is zero. However the right-hand side of the equation is zero only for steady currents. Because from equation of continuity (Eq. 4.89) $ Jþ

@q ¼0 @t

ð6:115Þ

@q @t

ð6:116Þ

)$J¼

@q ¼ 0; the steady currents, i.e. only in a [email protected] netostatic condition. In order to generalize for non-steady currents Maxwell modiﬁed the equation as follows We have The above equation is zero if

$J¼

@q @t

ð6:117Þ

Using Gauss’s law @ ðeo $ E Þ @t @E ) $ J ¼ $ eo @t $J¼

ð6:118Þ ð6:119Þ

Thus if we add the above term in Eq. 6.112 to J then Eq. 6.112 will be valid for non-steady currents. Hence Eq. 6.112 now becomes $ B ¼ l o J þ l o eo

@E @t

ð6:120Þ

Taking divergence of above equation $ ð$ BÞ ¼ lo $ J þ lo eo Using Gauss’s law

@ ð$ EÞ @t

ð6:121Þ

6.11

Maxwell’s Equation

421

@q $ ð$ BÞ ¼ lo $ J þ @t

ð6:122Þ

The left-hand side is zero as usual and the right-hand side is zero by equation of continuity. Apart from satisfying for non-steady currents Maxwell equation gives an interesting information. “Faraday’s law says that changing magnetic ﬁeld induces an electric ﬁeld, while Maxwell’s equation states that changing electric ﬁeld induces an magnetic ﬁeld”. Maxwell named the extra term in Eq. 6.121 as displacement current Jd ¼ e o

6.12

@E @t

ð6:123Þ

The Displacement Current

The concept of displacement current was a great theoretical discovery. It helped to retain the notion that flow of current in a circuit is continuous. As an example consider the plates of a parallel-plate capacitor connected to a voltage source as shown in Fig. 6.25. Conduction current flows in the leads of the capacitor charging the plates of the capacitor as shown in Fig. 6.25. But the conduction current I is not continuous across the gap between the plates as there is no transfer of charge between the plates. But from Maxwell’s equation we see that the changing electric ﬁeld between the plates serves the purpose of the conduction current inside the gap. Within the gap the changing electric ﬁeld produces a displacement current which is exactly equal to conduction current as shown below. If q is the charge given to one of the plates of the capacitor at any given instant of time and if A is the area of the plates then the electric ﬁeld between the plates of the capacitor from Eq. 3.331 is given by E¼ ) Fig. 6.25 A capacitor connected to a voltage source

V q ¼ d eo A

@E 1 @q I ¼ ¼ @t eo A @t eo A

ð6:124Þ ð6:125Þ

422

6 Time-Varying Fields and Maxwell’s Equation

) eo )

@E I ¼ @t A

ð6:126Þ

@ I ðeo EÞ ¼ @t A

ð6:127Þ

As D ¼ eo E in free space (P = 0 in D ¼ eo E þ P) )

@ I ðDÞ ¼ @t A

ð6:128Þ

As Jd ¼ @D @t I A

ð6:129Þ

) Jd A ¼ I

ð6:130Þ

) Id ¼ I

ð6:131Þ

) Jd ¼

As Jd A ¼ Id

Thus the displacement current in the gap between the capacitor plates is found to be equal to the conduction current flowing in the leads of the capacitor. The following are the four Maxwell’s equation ðaÞ

$E¼

q eo

ð6:132Þ

ðbÞ

$B¼0

ð6:133Þ

ðcÞ ðdÞ

$E¼

@B @t

$ B ¼ lo J þ lo eo

ð6:134Þ @E @t

ð6:135Þ

In free space with D ¼ eo E and B ¼ lo H we can write the above equations as $D¼q

ð6:136Þ

$H¼0

ð6:137Þ

6.12

The Displacement Current

423

$E¼

@B @t

$ H ¼ Jþ

6.13

@D @t

ð6:138Þ ð6:139Þ

Maxwell’s Equation in Matter

In the presence of materials Maxwell’s equation can be derived as follows. In the materials in the presence of electric and magnetic ﬁelds the material gets polarized and magnetized. With polarization we have seen that in Sects. 3.15–3.18, in addition to free charge density qf a bound charge density appears. Hence q ¼ qf þ q p

ð6:140Þ

In the case of magnetization we have seen that in Sect. 5.3 in addition to free current Jf a bound current Jp appears. Hence J ¼ Jf þ Jp

ð6:141Þ

Equations 6.140 and 6.141 correspond to electrostatic and magnetostatic conditions. However in a non-static case the bound charge density noted down in Eq. 6.140 may change with time giving rise to currents. Hence under non-static conditions there will be additional currents those arise from the change of polarization with time given by J1 ¼

@P @t

ð6:142Þ

In order to arrive at Eq. 6.142 consider a dielectric of arbitrary volume s bounded by surface subjected to a time-varying electric ﬁeld. The dielectric is polarized giving rise to bound charges. Under the action of time-varying electric ﬁeld the bound charges flow out of volume s giving rise to polarization current density J1. The rate at which bound charge flows out through S must be equal to the rate of decrease of bound charge within s: Hence Z S

R

@ J1 ds ¼ @t

Z qp ds

ð6:143Þ

s

The right-hand side of the above equation signiﬁes that the bound charge density R @ q s p ds changes with time as per @ t s qp ds and the negative sign indicates charge is

424

6 Time-Varying Fields and Maxwell’s Equation

decreasing R within volume s. The flow of bound charges outside s gives rise to current S J1 ds. Applying Gauss’s divergence theorem and from Eq. 3.262 qP ¼ $ P Z @ $ J1 ds ¼ $ Pds @t s s Z Z @P ds ) $ J1 ds ¼ $ @t Z

s

ð6:144Þ ð6:145Þ

s

The above equation must be valid even for small volume ds and in that case we need not integrate $ J1 ds ¼ $

@P ds @t

ð6:146Þ

@P @t

ð6:147Þ

) $ J1 ¼ $ ) J1 ¼

@P @t

ð6:148Þ

which is Eq. 6.142. Hence under non-static conditions Eq. 6.141 can be written as J ¼ Jf þ Jp þ J1

ð6:149Þ

Thus variation of bound charge density with time gives rise to current J1. However variation of magnetization with time doesn’t give rise to similar currents. As M varies with time, bound current given by Eq. 6.151 varies with time accordingly. From Eqs. 3.262 and 5.7, we see that qp ¼ $ P

ð6:150Þ

Jp ¼ $ M

ð6:151Þ

Hence Eqs. 6.140 and 6.149 can be written as q ¼ qf $ P J ¼ Jf þ $ M þ

ð6:152Þ @P @t

ð6:153Þ

6.13

Maxwell’s Equation in Matter

425

Maxwell’s ﬁrst equation is q eo

ð6:154Þ

1 ðq $ PÞ eo f

ð6:155Þ

$E¼ Using Eq. 6.152 $E¼

with D ¼ eo E þ P in the presence of materials Eq. 6.155 is ð6:156Þ

$ D ¼ qf Maxwell’s second equation is as usual $B¼0

ð6:157Þ

Maxwell’s third equation is $E¼

@B @t

ð6:158Þ

Maxwell’s fourth equation is $ B ¼ lo J þ eo

@E @t

ð6:159Þ

From Eq. 6.153

@P @E $ B ¼ lo Jf þ $ M þ þ eo @t @t

ð6:160Þ

We know B M lo

ð6:161Þ

D ¼ eo E þ P

ð6:162Þ

H¼ and

426

6 Time-Varying Fields and Maxwell’s Equation

Using the above equations in Eq. 6.160 $ H ¼ Jf þ

@D @t

ð6:163Þ

Equations (6.132–6.135), (6.136–6.139) are the Maxwell’s equation in free space. Equations 6.156, 6.157, 6.158 and 6.161 written below are Maxwell’s equation in the presence of materials $ D ¼ qf

ð6:164Þ

$B¼0

ð6:165Þ

$E¼

@B @t

$ H ¼ Jf þ

6.14

ð6:166Þ

@D @t

ð6:167Þ

Maxwell’s Equation in Integral Form and Boundary Conditions

Maxwell’s Eqs. 6.164–6.167 can be written in integral form as I ðiÞ

D ds ¼ qfðenclosedÞ

ð6:168Þ

S

I B ds ¼ 0

ðiiÞ

ð6:169Þ

S

I ðiiiÞ l

d E dl ¼ dt

Z B ds

I H dl ¼ IfðenclosedÞ þ

ðivÞ

ð6:170Þ

S

d dt

Z D ds

ð6:171Þ

S

l

Consider Fig. 6.26. Using Eq. 6.168 and following procedures mentioned in Sects. 2.25 and 3.22 D1n D2n ¼ rf

ð6:172Þ

6.14

Maxwell’s Equation in Integral Form and Boundary Conditions

427

Fig. 6.26 Boundary between two mediums shown along with a pill box

Fig. 6.27 Boundary between two mediums shown along with a Amperian loop

Consider Fig. 6.27. Using Eq. 6.170 Z I d E dl ¼ B ds dt l

ð6:173Þ

S

In the limit a in Fig. 6.27 goes to zero the surface integral vanishes. Hence E1t l E2t l ¼ 0

ð6:174Þ

428

6 Time-Varying Fields and Maxwell’s Equation

E1t ¼ E2t

ð6:175Þ

Using Eq. 6.169 and following procedures mentioned in Sects. 5.6 and Sect. 4.23 B1n ¼ B2n

ð6:176Þ

Applying Eq. 6.171 to Amperian loop in Fig. 6.27 I H dl ¼ IfðenclosedÞ þ l

d dt

Z D ds

ð6:177Þ

S

In a limit a in Fig. 6.27 goes to zero the surface integral vanishes. Hence Hlt l H2t l ¼ Kf l

ð6:178Þ

H1t H2t ¼ Kf

ð6:179Þ

Thus we have the following boundary conditions

ðiÞ

D1n D2n ¼ rf e1 E1n e2 E2n ¼ rf

ð6:180Þ

ðiiÞ

E1t ¼ E2t

ð6:181Þ

ðiiiÞ

B1n ¼ B2n

ð6:182Þ

ðivÞ

H1t H2t ¼ Kf 1 1 B1t B2t ¼ Kf l1 l2

ð6:183Þ

6.15

Potential Functions

For electrostatic ﬁeld we know that 1 V¼ 4peo

Z s

qds

r

ð6:184Þ

6.15

Potential Functions

429

and for magnetostatic ﬁeld A¼

lo 4p

Z s

Jds

r

ð6:185Þ

Now what will happen to these potentials when the ﬁelds are varying with time. Our fundamental aim is to ﬁnd the value of electric and magnetic ﬁelds. In electrostatic case electric ﬁeld E(r) is given by Coulomb’s law. In magnetostatic case the magnetic flux density B(r) is given by Biot–Savart law. But for time-varying ﬁelds E (r, t), B (r, t) are buried in Eqs. (6.132)–(6.135). As we have seen in Sect. 1.1 it is easy to work with scalars instead of vectors. As E(r, t) and B(r, t) are vectors we will now try to simplify the problem by working with potentials. In magnetostatics $ B ¼ 0 invited us to write B as $ A similarly from Eq. 6.133 we can write B¼$A

ð6:186Þ

Substituting the above Eq. (6.186) in Eq. (6.134) @ ð$ AÞ @t @A ) $ Eþ ¼0 @t $E¼

ð6:187Þ ð6:188Þ

Because curl gradient of any scalar function is zero, we can write from the above equation Eþ

@A ¼ $V @t

) E ¼ $V

@A @t

ð6:189Þ ð6:190Þ

Now substituting the above Eq. 6.190 in Eq. 6.132 @A q $ $V ¼ @t eo ) $2 V þ

@ q ð$ AÞ ¼ @t eo

ð6:191Þ ð6:192Þ

430

6 Time-Varying Fields and Maxwell’s Equation

With the help of Eqs. 6.186, 6.191 and 6.135 can be written as @V @2A $ ð$ AÞ ¼ lo J lo eo $ lo e o 2 @t @t

ð6:193Þ

From Sect. 1.20 identity 7 gives $ ð$ AÞ ¼ $ð$ AÞ $2 A

ð6:194Þ

Hence @V @2A $ ð$ AÞ $2 A ¼ lo J lo eo $ lo e o 2 @t @t )

$2 A lo eo

@2A @V e ¼ lo J $ $ A þ l o o @t2 @t

ð6:195Þ ð6:196Þ

We have to extract A, V from Eqs. 6.192, 6.196 and then substituting them in Eqs. 6.186, 6.190 we can obtain E(r, t), B(r, t). It seems that solving Eqs. 6.192, 6.196 to get A, V is a tedious task. However as we will see in next section gauge transformations simplify this problem.

6.16

Gauge Transformations

Equation 6.186 is B¼$A We know curl gradient of any scalar function is zero. Suppose a is any scalar function then B ¼ $ A þ $ $a

ð6:197Þ

B ¼ $ ½A þ $a

ð6:198Þ

Let the new vector function be A1 ¼ A þ $a

ð6:199Þ

B ¼ $ A1

ð6:200Þ

Hence

6.16

Gauge Transformations

431

“Thus changing the vector potential by adding gradient of any scalar function (Eq. 6.197) doesn’t alter the value of B”. If we change vector potential then what happens to E. Equation 6.190 is E ¼ $V

@A @t

Substituting for A from Eq. 6.199 E ¼ $V

@ ðA1 $aÞ @t

@A1 @a þ$ @t @t @a @A1 ) E ¼ $ V @t @t ) E ¼ $V

ð6:201Þ ð6:202Þ ð6:203Þ

With V1 ¼ V

@a @t

) E ¼ $V1

@A1 @t

ð6:204Þ ð6:205Þ

From Eqs. 6.205, 6.197 and 6.200 we observe that changing vector potential and scalar doesn’t alter the electric and magnetic ﬁelds. We have A1 ¼ A þ $a

ð6:206Þ

@a @t

ð6:207Þ

V1 ¼ V

That is if we add $a to A and simultaneously substract @a @t from V, then the values of E, B doesn’t get altered. We know that the actual physical quantities of interest are E, B and not the potentials A, V. We can change A, V for our advantage without changing E, B. That is we are free to select any form of vector potential and scalar potential without affecting E, B. Because we can change vector potential and scalar potential without affecting E, B we select the A such that $ A ¼ lo eo

@V @t

ð6:208Þ

432

6 Time-Varying Fields and Maxwell’s Equation

What is the advantage of Eq. 6.208. To check, substitute 6.208 in 6.196 then $ 2 A l o eo

@2A ¼ lo J @t2

ð6:209Þ

@ 2 V q ¼ @t2 eo

ð6:210Þ

and in 6.192 $ 2 V lo e o

Thus from Eqs. 6.192, 6.196 we have uncoupled A and V in Eqs. 6.209, 6.210 using the Eq. 6.208. Equations 6.206 and 6.207 are called gauge transformations. Equations 6.209 and 6.210 are inhomogeneous wave equations for A and V, and Eq. 6.208 is called Lorentz condition or Lorentz gauge for potentials.

6.17

Retarded Potentials

Solving Eqs. 6.209 and 6.210 we can get A and V for time-varying conditions. However there is one important point to be considered. First we shall write the potentials for static case, i.e. from electrostatics and magnetostatics VðrÞ ¼

1 4peo

l AðrÞ ¼ o 4p

Z

qðr0 Þ 0 ds

r

s0

Z s0

JðrÞ

r

ds0

ð6:211Þ

ð6:212Þ

Consider Fig. 6.28 where we show either qðr0 Þ or Jðr0 Þ In the static case, qðr0 Þ or Jðr0 Þ doesn’t change with time. But when qðr0 Þ or Jðr0 Þ changes with time then the situation is slightly different. As we will see the electromagnetic waves travel with velocity c the speed of light. At time tr a change produced in qðr0 ; tr Þ or Jðr0 ; tr Þ travels with the velocity c and takes time t to reach point P in Fig. 6.28. Suppose assume that we are interested in knowing the time-varying potentials and hence time-varying ﬁelds at point P at time t. But the potentials and ﬁelds at point P at time t depends on qðr0 Þ or Jðr0 Þ at time tr. The situation is similar to astronomy because stars are few hundreds of light years away from us. Whatever condition of stars we see today is the information which happened in stars few hundred light years before.

6.17

Retarded Potentials

433

Fig. 6.28 Volume charge density or volume current density is shown in the ﬁgure. Changes produced in volume charge density or volume current density at a given time tr travels a long distance and reaches point P at a later time t

Thus at point P whatever we measure at time t is that corresponds to that of qðr0 Þ or Jðr0 Þ at time tr. Thus tr ¼ t

r

ð6:213Þ

c

where cr is the time taken for the change produced in either qðr0 ; tr Þ or Jðr0 ; tr Þ to reach point P from the source. tr is called retarded time. Equations 6.211 and 6.212 can then be generalized to non-static case as Vðr; tÞ ¼

1 4peo

l Aðr; tÞ ¼ o 4p

Z

qðr0 ; tr Þ

r

s0

Z s0

Jð r 0 ; t r Þ

r

ds0

ds0

ð6:214Þ

ð6:215Þ

Once V(r, t) and A (r, t) has been calculated from Eqs. 6.214 and 6.215 E(r, t), B(r, t) can be calculated using equations Eðr; tÞ ¼ $V

@A @t

Bðr; tÞ ¼ $ A

ð6:216Þ ð6:217Þ

434

6.18

6 Time-Varying Fields and Maxwell’s Equation

Time-Harmonic Fields

The electric and magnetic ﬁelds in Maxwell’s equations have arbitrary time dependence. We will now restrict ourselves to time-harmonic ﬁelds. Harmonic ﬁelds are those which varies sinusoidally or periodically with time. Expressing ﬁelds in terms of sinusoids have an advantage that most wave forms can be expressed in terms of sinusoids with the Fourier’s techniques. Calculations with time-harmonic ﬁelds can be easily done with exponential functions. Suppose assume that there are two time-harmonic functions given by f 1 ¼ C1 cosðxt þ d1 Þ and f 2 ¼ C2 cosðxt þ d2 Þ We are interested in adding them f3 ¼ f1 þ f2 f 3 ¼ C1 cosðxt þ d1 Þ þ C2 cosðxt þ d2 Þ However instead of adding f1 and f2 directly, if we carry out the addition of f1 and f2 as stated below calculations become easier. f3 ¼ f1 þ f2 ¼ C1 cosðxt þ d1 Þ þ C2 cosðxt þ d2 Þ h i h i f 3 ¼ Re C1 eiðxt þ d1 Þ þ Re C2 eiðxt þ d2 Þ Similarly if we want to add sine functions C1 sinðxt þ d1 Þ; C2 sinðxt þ d2 Þ then calculations become easier if we carry out the addition as follows ¼ C1 sinðxt þ d1 Þ þ C2 sinðxt þ d2 Þ h i h i ¼ Im C1 eiðxt þ d1 Þ þ Im C2 eiðxt þ d2 Þ The point that, working with exponential functions is much easier than working with time-harmonic functions, will be more clear to the reader when the reader completes Sect. 7.6 in Chap. 7. Having seen that it is much easier to work with exponential functions than with harmonic functions, we will now discuss phasors. Let Aðx; y; z; tÞ be a time-harmonic ﬁeld. The phasor form of A is given by As ðx; y; zÞ which depends on space coordinates x, y, z alone. Then we can write

Aðx; y; z; tÞ ¼ Re As ðx; y; zÞeixt

ð6:218Þ

6.18

Time Harmonic Fields

435

or

A ¼ Re As eixt

ð6:219Þ

As ¼ As ðx; y; zÞ

ð6:220Þ

Here

is the phasor form of Aðx; y; z; tÞ As the phasor, is independent of time and depends on x, y, z only. As an example consider the electric ﬁeld given by Eðz; tÞ ¼ Eo cosðxt bzÞ

ð6:221Þ

where the meaning of b will be clear to the reader in Chap. 7. We can write Eq. 6.221 as h i Eðz; tÞ ¼ Re Eo eiðxtbzÞ

ð6:222Þ

) Eðz; tÞ ¼ Re Eo eibz eixt

ð6:223Þ

Es ¼ Eo eibz

ð6:224Þ

We deﬁne

Es in Eq. 6.224 is the phasor form of E (z, t) in Eq. 6.223. Thus Eq. 6.223 can be written as

Eðz; tÞ ¼ Re Es eixt

ð6:225Þ

In Eq. 6.224 we have thus separated the spatial part in the form of phasors. Phasors of the time-harmonic ﬁelds are a complex quantities which are function of space only. As seen in Eq. 6.225 the time dependency is present in eixt . Thus we have separated spatial part in the form of phasors and time part in eixt . In summary in this section we have seen that we use time-harmonic ﬁelds because all other wave forms can be expressed in the form of time-harmonic ﬁelds. Then we have expressed time-harmonic ﬁelds in exponential form as it is easy to work with exponentials. Finally we separated the exponential form into space part (phasors) and time part eixt . Reader must take care to differentiate between i in eixt which denotes the complex number and ^i where ^i is the unit vector along x-direction. Example 6.17 ~ ¼E ~ o eixt calculate the ratio between the conduction If the electric ﬁeld is given by E current Jc and displacement current Jd.

436

6 Time-Varying Fields and Maxwell’s Equation

Solution We know that the conduction current is given by Jc ¼ rE The displacement current is given by @E @ ~ ixt ¼e Eo e dt dt ) Jd ¼ ixeE r jJc j ¼ ) jJd j xe

Jd ¼ e

Example 6.18 Using the results of Example 6.17 obtain the condition under which the material medium can acts as an conductor or an insulator when an alternating ﬁeld is present. Solution In a conductor, the conduction current density is very high as compared to displacement current density. Jc Jd Then r xe )x

r e

In a insulator the conduction current density is lessor than displacement current density Jc Jd ) r xe r )x e Example 6.19 Check whether at 1 MHz aluminium is a good conductor and Teflon is a good insulator. Given (a) For aluminium r = 35.3 Mega-Mho/m e = eo

6.18

Time Harmonic Fields

437

(b) For Teflon r = 30 nano-Mho/m e = 2.1 eo Note Mho/m is the unit of r. Solution (a) For aluminium

r 35:3 106 ¼ xe 2ð3:14Þ106 ð8:854 102 Þ r ¼ 6:4 1011 xe which is very high than unity or in other words x

r e

Hence aluminium is a conductor. Even at very high frequencies say 100 GHz the following condition is satisﬁed in the conductors x

r or Jc Jd e

Thus conduction current dominates displacement current by a very large value. Hence in general, in conductors, displacement current can be neglected. For Teflon r 30 109 ¼ xe 2p 106 2:1 8:854 1012 ¼ 2:57 104 which is lessor than unity. (or) Jc Jd Hence at 1 MHz Teflon is an insulator. Conduction current density is negligible as compared to displacement current density. However at lower frequencies the r ratio xe may approach unity and the conduction current may become comparable to displacement current. Example 6.20 Calculate the frequency at which Jc ¼ Jd in distilled water for which e = 81eo and r = 2 10−4 Mho/m.

438

6 Time-Varying Fields and Maxwell’s Equation

Solution For Jc ¼ Jd r ¼ xe Frequency x r ¼ 2p 2pe 2 104 f¼ 2pð8:854 1012 Þð81Þ

f¼

) f ¼ 44:4 kHz Example 6.21 The electric ﬁeld E in free space is given by E ¼ Eo cosðxt kzÞ^i Find B. Solution We know ^i @B @ ¼ $E¼ @x @t E cosðxt kzÞ o @B ) ¼ kEo sinðxt kzÞ^j @t kEo cosðxt kzÞ^j )B¼ x Example 6.22 In a source free region if ^ A ¼ x4^i þ z2 t2 k Find Eðr; tÞ; Bðr; tÞ Solution In source free region as qðr0 ; tr Þ ¼ 0 and hence.

^j @ @y

0

^ k @ @z 0

6.18

Time Harmonic Fields

439

Z 1 qðr0 ; tr Þ 0 Vðr; tÞ ¼ ds 4peo s0 r Vðr; tÞ ¼ 0 @A @A ¼ E ¼ $V @t @t ^ ) E ¼ 2z2 tk and B¼$A¼0

6.19

Maxwell’s Equation in Phasor Form

From Eq. 6.218 an electric ﬁeld E (x, y, z, t) can be written as

Eðx; y; z; tÞ ¼ Re Es ðx; y; zÞeixt

ð6:226Þ

@E ¼ Re ixEs eixt @t

ð6:227Þ

@E ¼ Re½ixE @t

ð6:228Þ

From Eq. 6.226

Hence from Eq. 6.228 we can write @ ! ix @t

ð6:229Þ

Similar to Eq. 6.226 writing expressions for D (x, y, z, t), H (x, y, z, t), B (x, y, z, t), J (x, y, z, t) and q (x, y, z, t) we can express Maxwell’s equations 6.136–6.139, 6.164– 6.167 in phasor form. eixt being present in both sides of the equations cancel out. $ Ds ¼ qs

ð6:230Þ

$ Bs ¼ 0

ð6:231Þ

$ Es ¼ ixBs

ð6:232Þ

$ Hs ¼ Js þ ixDs

ð6:233Þ

440

6 Time-Varying Fields and Maxwell’s Equation

The above four Maxwell’s equation in phasor form depends on space part (x, y, z) only and is independent of time. The integral form of Maxwell’s equations 6.168–6.171 can also be derived in phasor form. See Problem 6.20.

Exercises Problem 6.1 Suppose assume that in Problem 3.8 the space between parallel-plate capacitors is ﬁlled with a medium which is homogeneous with ﬁnite conductivity r. Find the resistance of the region between the plates. Problem 6.2 In a material when electric ﬁeld is applied the electrons get accelerated and flow towards the positive terminal. As they move towards the positive terminal they collide with the atoms. Because of these collisions the electron undergoes random motion in all directions as shown in Fig. 6.29. The net result is that electrons acquire a slow drift velocity vd towards the positive terminal. Show that the drift velocity is related to current density as J ¼ nevd where n is the number of electrons per unit volume. Problem 6.3 Repeat Example 6.1 with r ¼ Cq1 þ C2 where C1 and C2 are constants. Problem 6.4 A silver wire of 0.9 mm diameter carries a charge of 75 C in 1 h. Calculate the drift velocity of the electrons in the wire using the results of Problem 6.2 if n for silver is 5.8 1022 electrons/cm3. Problem 6.5 A Nickel wire of diameter 0.5 cm is welded to a copper wire of diameter 0.4 cm. Both the wires as such carry a current of 9 A. What is the current density in each wire. Assuming one free electron per atom in copper calculate drift speed of the

Fig. 6.29 Drift of electrons towards the positive terminal in material subjected to an electric ﬁeld. Illustration for Problem 6.2

Exercises

441

electrons in copper wire. Given atomic weight of copper 64 and density of copper is 9 gm/cc. Hint: Use the results of Problem 6.2. Problem 6.6 Calculate the induced emf of a metal aeroplane vertically diving at a speed of 500 km/s in an earth’s magnetic ﬁeld where the horizontal component is 0.5 10−4 Wb/m2. The wingspan of the aeroplane is 40 m. Problem 6.7 Repeat example 6.4 but this time the magnetic ﬁeld is pointing at an angle of 45° to the z-axis. Problem 6.8 A metal rod is rotating with a constant angular velocity x in a uniform magnetic flux density B. The length of the rod is L. The magnetic flux density B is perpendicular to the length of the rod. What is the emf induced between the ends of the rod? Problem 6.9 A train passes on a railway track whose rails are assumed to be insulated from each other. The velocity of the train is 150 km/h. The distance between the rails in the railway track is 1.6 m. If the vertical component of the earth’s magnetic ﬁeld at the points of the rail is 0.373 Wb/m2 calculate the emf that will exist between the rails? Problem 6.10 A steel core of rectangular cross section 15 mm 10 mm and a mean length of 160 mm is shown in Fig. 6.30. The air gap length is 1 mm. The flux in the air gap is 156 l Wb. (a) Determine the mmf H if the value of H is 200 A/m for B = 1.04 T. Using the relation mmf ¼ H dl: (b) Draw the equivalent circuit for Fig. 6.30, calculate reluctance and hence compute mmf.

Fig. 6.30 A steel core of rectangular cross-section. Illustration for Problem 6.10

442

6 Time-Varying Fields and Maxwell’s Equation

Fig. 6.31 Two coaxial cylinders. Illustration for Problem 6.11

Problem 6.11 Consider two coaxial cylinders as shown in Fig. 6.31. The inner cylinder is a solid cylinder carrying a current I distributed uniformly over its cross section. Calculate the self-inductance per unit length. Problem 6.12 In Example 6.10 calculate the energy stored in the solenoid. Problem 6.13 In Example 6.12 calculate the energy stored in the toroid. Problem 6.14 Consider a long coaxial cable as shown in Fig. 6.32. The current I flows down the surface of the inner radius and comes back along the outer cylinder. Find the self-inductance and magnetic energy stored in the section of length L. Problems 6.15 Consider a material with er = 1 and conductivity r = 4.3 Mho/m. The value of E is E ¼ 200 sin 109 t ^iV=m: Calculate the conduction and displacement current densities. Also ﬁnd the frequency at which they have equal magnitude. Problem 6.16 Consider a coaxial capacitor. The inner and outer radius of the coaxial capacitor is 5 mm and 7 mm, respectively. The length of the coaxial capacitor is 700 mm. The

Fig. 6.32 A long coaxial cable. Illustration for Problem 6.14

Exercises

443

space between capacitor is ﬁlled with dielectric with er = 5. The capacitor is subjected to a ﬁeld. E¼

2 102 V sin 400t ^r m r

Calculate the displacement current and conduction current. Problem 6.17 Consider a parallel-plate capacitor ﬁlled with a dielectric with er = 6. The area of the plates separated by a distance of 4 mm is 20 cm2. The capacitor is connected to a voltage source of 400 V whose frequency is 2 MHz. Calculate the displacement current. Problem 6.18 In free space show that the following ﬁeld vectors satisfy Maxwell’s equation E ¼ Eo cosðxt kzÞ^i Eo H ¼ cosðxt kzÞ^j g Calculate g: Problem 6.19 Consider a spherical capacitor consisting of two concentric spheres of inner and outer radius a, b, respectively. The region between two spheres is ﬁlled with air. Find the displacement current. Problem 6.20 For time-harmonic ﬁelds show that Z [email protected] !

E ix

and hence obtain integral form of Maxwell’s equation 6.168–6.171 in phasor form I I I

Z Ds ds ¼

qs ds

Bs ds ¼ 0

Z Es dl ¼ ix Bs ds I Z Hs dl ¼ ðJs þ ixDs Þ ds

The reader should note that subscript s denotes phasor form while s in ds denotes surface.

Chapter 7

Plane Electromagnetic Waves

7.1

Introduction

In this chapter we will discuss about plane electromagnetic waves. What is a wave? A wave is a disturbance passing through a medium. The wave carries energy and momentum with it. As an example of a wave assume there are some bricks arranged in a line as shown in Fig. 7.1a. Suppose we push one of the bricks at one of the end, then the bricks fall in sequence, one by one, as shown in Fig. 7.1b. The hand shown in Fig. 7.1b produces a disturbance by pushing one of the brick at the end. The disturbance then travels through the medium—the bricks. There are quite a few number of classiﬁcations of a wave. One of the important classiﬁcations of wave worth mentioning here is, longitudinal wave and transverse wave. In a longitudinal wave the displacement of the medium is parallel to the direction of the propagation of the wave. An example of the longitudinal wave is shown in Fig. 7.1. The medium, that is the bricks, is displaced parallel to the direction of propagation of the wave. In Fig. 7.1b the bricks are displaced in x-direction and wave also propagates in the x-direction. In a transverse wave the displacement of the medium is perpendicular to the direction of propagation of the wave. An example of a transverse wave is shown in Fig. 7.2. A person holds one end of the rope. He moves the rope up and down thereby producing a disturbance. The disturbance travels through the medium—the rope. In this example the medium, the rope, is displaced perpendicular to the direction of wave propagation. In Fig. 7.2 the medium, the rope, is displaced in y-direction, while the wave propagates in the x-direction. Let us see the mathematical description for the wave. As an example consider a wave as shown in Fig. 7.3. The wave travels with a velocity v in the z-direction. u (z, t) is the function representing the wave. For example in Fig. 7.3 u(z, t) represents the displacement of the rope at point z at time t. Then the wave equation is given by

© Springer Nature Singapore Pte Ltd. 2020 S. Balaji, Electromagnetics Made Easy, https://doi.org/10.1007/978-981-15-2658-9_7

445

446

7 Plane Electromagnetic Waves

Fig. 7.1 (a) Bricks arranged in a line. (b) On of the bricks is pushed and one by one the bricks fall thereby the disturbance travels through the bricks

Fig. 7.2 A rope is moved up and down thereby a disturbance passes through the rope

Fig. 7.3 A wave is propagating along the z direction and u (z,t) is the function representing the wave.

7.1 Introduction

447

@2u 1 @2u ¼ @z2 v2 @t2

ð7:1Þ

The solution of the above equation is uðz; tÞ ¼ uo eiðxtbzÞ

ð7:2Þ

where uo is the amplitude of the wave and x is the angular frequency and b is given by b¼

2p k

The reader is referred to any standard book on waves to learn complete theory of waves. We will now discuss about propagation of electromagnetic waves in different media and use phasors to our advantage. That is we will discuss initially about the propagation of electromagnetic waves in spatial coordinates-(x, y, z), with the help of phasors. Then we will insert the time part eixt to explain the propagation of electromagnetic waves with respect to time.

7.2

The Wave Equation

Let us now derive the wave equation from Maxwell’s equation. Consider a uniform linear medium with permittivity e, permeability l and conductivity r. Let the medium be source free. That is the medium doesn’t contain any charge and currents other than that determined by Ohm’s law. Then we have D ¼ eE; Jf ¼ rE; B ¼ lH and qf ¼ 0

ð7:3Þ

Hence from Eqs. 6.164–6.167 9 rD¼0 > > > > rB¼0 > = @B rE¼ @t > > > @D > > ; r H ¼ Jf þ @t

ð7:4Þ

Substituting Eq. 7.3 in 7.4. ðaÞ r E ¼ 0 ðbÞ r H ¼ 0 ðcÞ

r E ¼ l

@H @t

9 > > > > > =

> > > @E > > ; ðdÞ r H ¼ rE þ e @t

ð7:5Þ

448

7 Plane Electromagnetic Waves

Taking curl of Eq. 7.5c we have @ ðr HÞ @t

ð7:6Þ

@ @E rE þ e @t @t

ð7:7Þ

r r E ¼ l Substituting Eq. 7.5d in Eq. 7.6 r r E ¼ l

Now using vector identify 7 from Sect. 1.20 @ @E rE þ e rðr EÞ r E ¼ l @t @t

ð7:8Þ

2

Substituting Eq. 7.5a in Eq. 7.8 @ @E rE þ e r E¼l @t @t

ð7:9Þ

2

) r2 E rl

@E @2E el 2 ¼ 0 @t @t

ð7:10Þ

Taking curl of Eq. 7.5d

@ r ðr HÞ ¼ rðr EÞ þ e ðr EÞ @t

ð7:11Þ

Substituting Eq. 7.5c in Eq. 7.11 and using vector identity 7 from Sect. 1.2 and using Eq. 7.5b we have r2 H rl

@H @2H el 2 ¼ 0 @t @t

ð7:12Þ

Equations 7.10, 7.12 are called general wave equations.

7.3

Plane Electromagnetic Wave in Free Space

Let us assume a wave travelling in z-direction without having x or y dependance. The ﬁelds are uniform over the entire plane perpendicular to the z-direction, i.e. the direction of wave propagation. Such a wave is called as uniform plane wave. In order to create a uniform ﬁeld over the entire plane a source of inﬁnite extent will be required. However, such sources are practically impossible because the sources are

7.3 Plane Electromagnetic Wave in Free Space

449

of ﬁnite extent. On the other hand if we are very far away from the source then the wave front which is surface of constant phase is almost spherical and a small portion of a sphere is a plane. For free space r ¼ 0; e ¼ eo ; l ¼ lo

ð7:13Þ

Substituting Eq. 7.13 in Eqs. 7.10, 7.12 r2 E eo lo

@2E ¼0 @t2

ð7:14Þ

r2 H eo lo

@2H ¼0 @t2

ð7:15Þ

The above two equations can be written as

@2 @2 @2 @2E þ þ l ¼0 E e o o @t2 @x2 @y2 @z2

ð7:16aÞ

@2 @2 @2 @2H þ þ l ¼0 H e o o @t2 @x2 @y2 @z2

ð7:16bÞ

For uniform plane waves travelling in the z-direction Eqs. 7.16a, 7.16b can be written as @2E @2E e l ¼0 o o @z2 @t2

ð7:16cÞ

@2H @2H eo l o 2 ¼ 0 2 @z @t

ð7:16dÞ

Comparing Eqs. 7.16c, 7.16d to Eq. 7.1 1 v ¼ pﬃﬃﬃﬃﬃﬃﬃﬃﬃ eo lo

ð7:17Þ

Substituting the values of eo ; lo we get v ¼ c ¼ 3 108 m=sec Hence electromagnetic waves travel with the velocity of light. Let us now assume that E and H have harmonic time dependency. Hence we write the electric ﬁeld as

450

7 Plane Electromagnetic Waves

E ¼ Es eixt

ð7:18aÞ

where Es is the usual phasor which depends on space coordinates. Substituting Eqs. 7.17, 7.18a in Eq. 7.16c @2E @2E ¼ v2 2 2 @t @z

ð7:18bÞ

@ 2 ixt @2 ¼ v2 2 Es eixt Es e 2 @t @z 2 @ 2 ixt 2 ixt @ Es ) Es 2 e ¼v e @t @z2 2 2 v @ Es ) Es ¼ 2 2 x @z @ 2 Es x2 ) ¼ Es @z2 v2 )

We deﬁne b¼

x v

Hence @ 2 Es ¼ b2 Es @z2 @ 2 Es ) þ b2 E s ¼ 0 @z2

ð7:19Þ

H ¼ Hs eixt

ð7:20aÞ

Similarly with

Equation 7.16d can be written as @ 2 Hs þ b2 H s ¼ 0 @z2

ð7:20bÞ

The solution of the differential Eq. 7.19 is Es ¼ Eo eibz þ E0o eibz

ð7:21aÞ

7.3 Plane Electromagnetic Wave in Free Space

451

where the term Eo eibz represents a wave travelling in +z-direction. The term E0o eibz represents a wave travelling in–z-direction. Because we consider a plane wave travelling in the +z-direction E0o ¼ 0: Hence Es ¼ Eo eibz

ð7:21bÞ

Substituting Eq. 7.21b in Eq. 7.18a Eðz; tÞ ¼ Eo eiðxtbzÞ

ð7:22Þ

Similarly we get with Hs ¼ Ho eibz Hðz; tÞ ¼ Ho eiðxtbzÞ

ð7:23Þ

The actual physical ﬁelds are given by real parts of E and H. Eo, Ho are the amplitudes of the ﬁelds and are in general complex numbers. The propagation vector b is given by ^¼ b ¼ bk

2p ^ 2pv ^ x ^ k¼ k¼ k k v v

ð7:24Þ

^ is the unit vector in the direction of wave propagation, the z-direction. Here k Let Eo in Eq. 7.21b be written as ^ Eo ¼ ðEo Þxs^i þ ðEo Þys^j þ ðEo Þzs k

ð7:25Þ

In Eq. 7.16c we have assumed that the wave propagates in the z-direction. Let the wave be uniform plane wave characterized by uniform electric ﬁeld over plane surfaces perpendicular to z with no electric ﬁeld component in z-direction (see Fig. 7.4a).

(a)

Fig. 7.4 Uniform plane wave with uniform electric ﬁeld over a plane surface perpendicular z

452

7 Plane Electromagnetic Waves

(b)

Fig. 7.4 An electromagnetic wave propagating in free space in the z-direction

Hence ðEo Þzs¼0

ð7:26Þ

Eo ¼ ðEo Þxs^i þ ðEo Þys^j

ð7:27Þ

Ho ¼ ðHo Þxs^i þ ðHo Þys^j

ð7:28Þ

r Es ¼ ix lo Hs

ð6:232Þ

Thus Eq. 7.25 is

Similarly we can write

From Eq. 6.232

We have from Eq. 7.21b 9 h i Es ¼ Eo eibz ¼ ðEo Þxs^i þ ðEo Þys^j eibz > > = and similarly h i > ; Hs ¼ Ho eibz ¼ ðHo Þxs^i þ ðHo Þys^j eibz >

ð7:29Þ

Hence from Eq. 6.232 ðE

^i @ @x o Þxs e

^j @ @y ibz

ðEo Þys eibz

^ k h i @ ¼ lo ðixÞ ðHo Þxs^i þ ðHo Þys^j eibz @z 0

) bðEo Þys bðEo Þxs

¼ lo xðHo Þxs ¼ lo xðHo Þys

ð7:30Þ

ð7:31Þ

7.3 Plane Electromagnetic Wave in Free Space

453

Equation 7.31 can be written as lo H o ¼

b ^ k Eo x

ð7:32Þ

Multiplying Eq. 7.32 by eibz on both sides and inserting the time factor eixt then using Eqs. 7.22, 7.23 we can write lo H ¼

b ^ ðk EÞ x

ð7:33Þ

So E, H are mutually perpendicular. As the wave is travelling in z-direction, if E is pointing in x-direction then H must point in y-direction. (Hence electromagnetic waves are transverse waves) (Further see Example 7.1). In such situation the magnitude of the real part in Eq. 7.32 ðH o Þy ¼

b ðE o Þx lo x

ð7:34Þ

From the above equation the ratio of E, H is g¼

ðEo Þx xlo ¼ ¼ ðHo Þy b

rﬃﬃﬃﬃﬃ lo eo

ð7:35Þ

where we have used Eqs. 7.17, 7.24. The unit of η is ohms, as E, is in volt/m and H is in amp-turn/m. Thus jEj ðEo Þx ¼ is in volt/amp or ohms: jHj ðHo Þy Hence η is called intrinsic impedance of the medium. Substituting the values of lo and eo in Eq. 7.35 g ¼ 376:6 ohms

ð7:36Þ

for free space. From Eq. 7.35 we see that the ratio jEj ðEo Þx ¼ is a constant. This means that the ﬁeld vectors E and H are in phase. jHj ðHo Þy Because E, H are pointing in x-direction and y-direction Eqs. 7.22, 7.23 can be written as Eðz; tÞ ¼ ðEo Þx eiðxtbzÞ^i ð7:37Þ Hðz; tÞ ¼ ðHo Þy eiðxtbzÞ^j

ð7:38Þ

454

7 Plane Electromagnetic Waves

The actual electric and magnetic ﬁelds are given by real parts of equations Eðz; tÞ ¼ ðEo Þx cosðxt bzÞ^i

ð7:39Þ

Hðz; tÞ ¼ ðHo Þy cosðxt bzÞ^j

ð7:40Þ

The results we have obtained in this section are pictorially depicted in Fig. 7.4b. In the ﬁgure E, H are in phase. E, H, b are perpendicular to each other. In Fig. 7.1 the medium is brick through which the wave or disturbance propagates. In Fig. 7.2 the medium is a rope through which the wave or disturbance propagates. In Fig. 7.4b there is no physical medium. It is a free space. The variations of electric ﬁeld in Fig. 7.4 are in x-direction. The variations of magnetic ﬁeld in Fig. 7.5 are in y-direction. The propagation of the variations of electric and magnetic ﬁelds is in z-direction. It is this propagation we call as a wave. Hence the electromagnetic waves travel without a need for an external and physical medium. Note: (1) From Eqs. 7.21b and 7.27 we can write Es ¼ Eo eibz h i ¼ ðEo Þxs^i þ ðEo Þys^j eibz Es ¼ Exs^i þ Eys^j where Exs ¼ ðEo Þxs eibz Eys ¼ ðEo Þys eibz Suppose the electric ﬁeld is along x-axis then ðEo Þys ¼ Eys ¼ 0: Hence Es ¼ Exs^i ¼ ðEo Þxs eibz^i (2) In the presence of a medium with r = 0 and permeability l, permittivity e Eqs. 7.10, 7.12, 7.17, 7.35 can be written as @2E ¼ 0; @t2 @2H r2 H el 2 ¼ 0; @t rﬃﬃﬃ 1 l v ¼ pﬃﬃﬃﬃﬃ ; g ¼ el e

r2 E el

7.3 Plane Electromagnetic Wave in Free Space

455

Example 7.1 If an electromagnetic wave is travelling in z-direction in free space with no charge density with an electric and magnetic ﬁelds given by E ¼ Eo eiðxtbzÞ H ¼ Ho eiðxtbzÞ Then prove that electromagnetic waves are transverse waves. Solution ^ Let E ¼ ðEo Þx^i þ ðEo Þy^j þ ðEo Þz k Then h i @ ^@ ^ @ ðEo Þ ^i þ ðEo Þ ^j þ ðEo Þ k ^ þj þk r E ¼ ^i x y z @x @y @z ) r E ¼ ðibÞeiðxtbzÞ ðEo Þz ¼ 0 by using Maxwell’s equation 6.132 for charge free space. ) ðEo Þz ¼ 0 Similarly using Eq. 6.133 we can show that ðHo Þz ¼ 0: Thus as there are no components of electric and magnetic ﬁelds in the direction of propagation, the z-direction, electromagnetic waves are transverse waves.

7.4

Poynting Theorem

Energy is transported through the electromagnetic wave whenever the wave travels from the transmitter to the receiver. In this section we will derive a relation regarding the energy transport in the electromagnetic wave. From Eq. 6.163 r H ¼ Jf þ

@D @t

) Jf ¼ r H

@D @t

ð7:41Þ

456

7 Plane Electromagnetic Waves

Taking the dot product of E on both sides E Jf ¼ E ðr HÞ E

@D @t

ð7:42Þ

Using vector identity 8 from Sect. 1.20 E ðr HÞ ¼ H ðr EÞ r ðE HÞ

ð7:43Þ

Substituting Eq. 7.43 in 7.42 E Jf ¼ H ðr EÞ r ðE HÞ E

@D @t

ð7:44Þ

Substituting Eq. 6.166 in Eq. 7.44 E Jf ¼ H

@B @D r ðE HÞ E @t @t

ð7:45Þ

For a linear, homogeneous and isotropic medium B ¼ lH and D ¼ eE: Hence @B @H l @ðH HÞ ¼ lH ¼ @t @t 2 @t @B l @H2 ¼ )H @t 2 @t

H

ð7:46Þ

Similarly E

@D @E e @E2 ¼ eE ¼ @t @t 2 @t

ð7:47Þ

Substituting Eqs. 7.46, 7.47 in 7.45 E Jf ¼

1 @ l 2 e 2 H þ E r ðE HÞ 2 @t 2 2

ð7:48Þ

Integrating over a volume s Z E Jf ds ¼ s

@ @t

Z Z l 2 e 2 H þ E ds r ðE HÞds 2 2 s

s

ð7:49Þ

7.4 Poynting Theorem

457

Applying Gauss’s divergence theorem to the second term in the right-hand side of Eq. 7.49 we get Z E Jf ds ¼ s

@ @t

Z I l 2 e 2 H þ E ds ðE HÞ ds 2 2 s

ð7:50Þ

S

where S is the surface bounded by volume s. Now, we will interpret Eq. 7.50 term by term. Z @ l 2 e 2 H þ E ds (i) The term @t 2 2 s

For the electrostatic case the energy stored is given by (Eq. 3.32) Z s

1 2 eE ds 2

For the magnetostatic case the energy stored is given by (Eq. 6.99) Z s

1 2 lH ds 2

Assuming that the above equations are valid for time-varying case the total electromagnetic energy stored per unit volume is Z l 2 e 2 H þ E ds 2 2 s

Then the equation @ @t

Z l 2 e 2 H þ E ds 2 s 2

along with the negative sign represents the rate at which the stored energy in volume s is decreasing. R (ii) The term s E Jf ds Let E and B be the time-varying ﬁelds. The force F experienced by a charged particle q due to this time-varying ﬁeld is F ¼ q½E þ ðv BÞ Due to the force the charge q moves a distance dl in time dt.

458

7 Plane Electromagnetic Waves

Then the work done is dW ¼ F dl ¼ q½E þ ðv BÞ dl

ð7:51Þ

The power supplied by the ﬁelds is then P¼

dW dl ¼ q½E þ ðv BÞ dt dt

ð7:52Þ

) P ¼ q½E þ ðv BÞ v

ð7:53Þ

) P ¼ qE v

ð7:54Þ

But q = q ds and Jf = qv where we have assumed charge q is distributed over a volume s with volume charge density q. The entire charge moves with an average velocity v producing current Jf. ) P ¼ qdsE v

ð7:55Þ

) P ¼ E qvds

ð7:56Þ

) P ¼ E Jf ds

ð7:57Þ

Hence the power dissipated per unit volume is ¼ E Jf

ð7:58Þ

Thus the term Z E Jf ds s

represents power dissipated over the total volume s. H (iii) The term S ðE HÞ:ds R The term s E Jf ds represents the rate of energy dissipated in volume s. The Z @ l 2 e 2 H þ E ds represents the rate at which stored energy in the term @t s 2 2 volume s is decreasing. Thus as per law of conservation of energy considering a volume s bounded by surface S. “The rate of electromagnetic energy dissipation in a volume s must equal to the sum of the rate at which the electromagnetic stored energy is decreasing in the volume s and the rate at which energy is entering volume s through the surface S”.

7.4 Poynting Theorem

459

H Hence S ðE HÞ ds must represent the rate of flow of energy inward of H volume s. Thus the expression S ðE HÞ ds represents the rate of outward flow of energy through the surface S which bounds a volume s. Thus (E H) is energy flow per unit time per unit area and is known as Poynting vector, denoted as P:

P¼EH

ð7:59Þ

Equation 7.50 is known as Poynting theorem.

7.5

Average Poynting Vector

We know the electric and magnetic ﬁelds are given by E ¼ Eo eiðxtbzÞ ¼ ðE1 þ iE2 ÞeiðxtbzÞ

ð7:60Þ

as Eo is a complex number. Similarly H ¼ Ho eiðxtbzÞ ¼ ðH1 þ iH2 ÞeiðxtbzÞ

ð7:61Þ

where E1, E2, H1 and H2 are real. Then E ¼ ðE1 þ iE2 Þ½cosðxt bzÞ i sinðxt bzÞ E ¼ ½E1 cosðxt bzÞ þ E2 sinðxt bzÞ i½E1 sinðxt bzÞ E2 cosðxt bzÞ

ð7:62Þ ð7:63Þ

From Eq. 7.63 RealðEÞ ¼ E1 cosðxt bzÞ þ E2 sinðxt bzÞ

ð7:64Þ

RealðHÞ ¼ H1 cosðxt bzÞ þ H2 sinðxt bzÞ

ð7:65Þ

Similarly

460

7 Plane Electromagnetic Waves

Now hP i ¼ hRe E Re Hi

ð7:66Þ

Re E Re H ¼ ½E1 cosðxt bzÞ þ E2 sinðxt bzÞ ½H1 cosðxt bzÞ þ H2 cosðxt bzÞ ¼ ðE1 H1 Þ cos2 ðxt bzÞ

ð7:67Þ

þ ðE2 H2 Þ sin ðxt bzÞ þ ½ðE1 H2 Þ þ ðE2 H1 Þ sinðxt bzÞ cosðxt bzÞ 2

¼ ðE1 H1 Þ cos2 ðxt bzÞ þ ðE2 H2 Þ sin2 ðxt bzÞ þ ½ðE1 H2 Þ þ ðE2 H1 Þ sinðxt bzÞ cosðxt bzÞ

ð7:68Þ

As with x ¼ 2p=T (see Eq. 7.79a) 1 T

ZT

1 sin ðxt bzÞdt ¼ T

ZT cos2 ðxt bzÞdt ¼

2

0

1 2

0

and 1 T

ZT sinðxt bzÞ cosðxt bzÞdt ¼ 0 0

We have 1 hReE ReHi ¼ ½ðE1 H1 Þ þ ðE2 H2 Þ 2

ð7:69Þ

The complex conjugate of H is H ¼ ðH1 iH2 ÞeiðxtbzÞ

ð7:70Þ

h i E H ¼ ðE1 þ iE2 ÞeiðxtbzÞ ðH1 iH2 ÞeiðxtbzÞ

ð7:71Þ

E H ¼ ½ðE1 H1 Þ þ ðE2 H2 Þ þ i½ðE2 H1 Þ ðE1 H2 Þ ReðE H Þ ¼ ðE1 H1 Þ þ ðE2 H2 Þ

ð7:72Þ ð7:73Þ

7.5 Average Poynting Vector

461

Substituting Eq. 7.73 in Eq. 7.69, 1 hRe E Re Hi ¼ Re ðE H Þ 2

ð7:74Þ

Substituting Eq. 7.74 in Eq. 7.66 1 hP i ¼ PðE H Þ 2

ð7:75Þ

The above equation can be used to calculate the average value of Poynting vector directly from E, H.

7.6

Poynting Vector for Wave Propagation in Free Space

We know that the Poynting vector is given by

P¼EH

ð7:76Þ

For plane waves travelling in z-direction with electric and magnetic ﬁelds in x-direction and y-direction, respectively, from Eqs. 7.39, 7.40 We have Eðz; tÞ ¼ Eo cosðxt bzÞ^i

ð7:77Þ

Eo cosðxt bzÞ^j g

ð7:78Þ

Hðz; tÞ ¼

Here we have used ðEo Þx ¼ Eo ; ðHo Þy ¼ Ho and Eq. 7.35. Substituting Eqs. 7.75, 7.76 in 7.59

P¼

ðEo Þ2 cos2 ðxt bzÞð^i ^jÞ g

ð7:79Þ

E2o ^ cos2 ðxt bzÞk g

ð7:80Þ

P¼

For the case of time-varying ﬁelds P ¼ E H gives the instantaneous value of Poynting vector. For time-varying electric and magnetic ﬁelds whenever we make a measurement it will be over large number of cycles if the frequency is very high. In such instances average value will be of interest. We denote the average value by < > brackets,

462

7 Plane Electromagnetic Waves

Thus hP i ¼

E2o ^ hcos2 ðxt bzÞik g

1 hcos ðxt bzÞi ¼ T

ð7:80aÞ

ZT cos2 ðxt bzÞdt

2

0

where T is the time period of the wave. Now, 1 hcos ðxt bzÞi ¼ T 2

ZT cos 2ðxt bzÞ þ 1 dt 2 0

1 ¼ 2T

ZT

1 cos2ðxt bzÞdt þ 2T

0

¼

1 2T

ZT dt 0

2p=x Z

cos2ðxt bzÞdt þ

1 ½T 2T

0

1 sin 2ðxt bzÞ 2p=x 1 ¼ þ 2T x 2 0 1 sinð4p 2bzÞ sinð2bzÞ 1 ¼ þ 2T x x 2 1 sin 4p cos 2bz cos 4p sin 2bz sin 2bz 1 þ ¼ þ 2T x x 2 1 sin 2bz sin 2bz 1 þ ¼ þ 2T x x 2 Hence hcos2 ðxt bzÞi ¼

1 2

ð7:80bÞ

Hence Eq. 7.80a is then 1 E2o ^ k hP i ¼ 2 g

ð7:81aÞ

Let us illustrate the advantage of working with exponentials than with time-harmonic ﬁelds as discussed in Sect. 6.18.

7.6 Poynting Vector for Wave Propagation in Free Space

463

From Eqs. 7.37, 7.38 with ðEo Þx ¼ Eo ; ðHo Þy ¼ Ho and Eq. 7.35 we can write Eðz; tÞ ¼ Eo eiðxtbzÞ^i E H ðz; tÞ ¼ o eiðxtbzÞ^j g Substituting the above equations in (7.74) 1 Eo iðxtbzÞ iðxtbzÞ ^ ^ hP i ¼ Eo e e ði jÞ 2 g ) hP i ¼

1 jEo j2 ^ k 2 g

ð7:81bÞ

Compare the above equation with Eq. 7.80a. Directly working with time-harmonic ﬁelds in Eqs. 7.75, 7.76 so many lengthy and laborious steps were involved in arriving at Eqs. 7.80a. Using the exponential form the same result was arrived at in just few steps in Eq. 7.80b. It is encouraged that the reader can solve Problem 7.4 to verify further the advantage of working with exponential functions than with time-harmonic ﬁelds.

7.7

Plane Electromagnetic Waves in Lossy Dielectrics

In the previous sections we discussed the plane wave propagation in free space. In this section we will discuss the plane wave propagation in lossy dielectrics. For a lossy dielectric r 6¼ 0; e ¼ er eo l ¼ lr lo ð7:82Þ A lossy dielectric is a partially conducting dielectric in which the electromagnetic wave losses energy as it propagates in the medium. A lossy dielectric can be an imperfect dielectric or imperfect conductor. Let the medium be free of charge qf ¼ 0: Using Eq. 7.82 we can write Eqs. 7.10, 7.12 as r2 E ¼ rl

@E @2E þ el 2 @t @t

ð7:83aÞ

r2 H ¼ rl

@H @2H þ el 2 @t @t

ð7:83bÞ

464

7 Plane Electromagnetic Waves

Now using the phasor form of the electric ﬁeld vector (Eq. 6.219) E ¼ Es eixt Eq. 7.83a becomes @ ixt @2 Es e r2 Es eixt ¼ rl þ el 2 Es eixt @t @t ) r2 Es ¼ ixlðr þ ixeÞEs ixt Similarly with H ¼ Hs e Eq. 7.82b becomes

ð7:84aÞ

r2 Hs ¼ ixlðr þ ixeÞHs

ð7:84bÞ

c2 ¼ ðixlÞðr þ ixeÞ

ð7:85Þ

Let

c is known as propagation constant. Then Eqs. 7.83a, 7.83b can be written as r2 Es c2 Es ¼ 0

ð7:86aÞ

r2 Hs c2 Hs ¼ 0

ð7:86bÞ

Let us assume the wave propagates in the z-direction with the electric ﬁeld along x-axis and the magnetic ﬁeld along y-axis. Then Eqs. 7.86a, 7.86b can be written as @ 2 Es c2 Es ¼ 0 @z2

ð7:87aÞ

@ 2 Hs c2 H s ¼ 0 @z2

ð7:87bÞ

Because c is complex we can write c ¼ a þ ib

ð7:88Þ

a is called attenuation constant, b the phase constant. From Eq. 7.88 c2 ¼ ða þ ibÞ2 ¼ a2 b2 þ 2aib

ð7:89Þ

7.7 Plane Electromagnetic Waves in Lossy Dielectrics

465

Equating the real and imaginary parts of Eqs. 7.88, 7.85 b2 a2 ¼ x2 le

ð7:90aÞ

2ab ¼ xlr

ð7:90bÞ

2 qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 2 c ¼ a2 b2 þ ð2abÞ2 ) c2 ¼ a2 þ b2

ð7:91Þ

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 2 c ¼ xl r2 þ x2 e2

ð7:92Þ

From Eq. 7.88

From Eq. 7.85

Comparing Eqs. 7.91, 7.92 a2 þ b2 ¼ xl

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ r2 þ x2 e2

ð7:93Þ

Solving Eqs. 7.90a, 7.93, 7.90b vﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ # u "rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ

r 2 ule t a¼x 1 1þ 2 xe vﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ #ﬃ u "rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ

r 2 ule b ¼ xt þ1 1þ 2 xe

ð7:94aÞ

ð7:94bÞ

Because we assume the wave propagates in z-direction and the electric ﬁeld is pointing in x-direction Es ¼ Exs ðzÞ^i

ð7:95Þ

Substituting Eq. 7.95 in Eq. 7.87a

d2 2 c Exs ðzÞ ¼ 0 dz2

ð7:96Þ

466

7 Plane Electromagnetic Waves

The solution of the above equation is Exs ðzÞ ¼ Eo ecz þ E00 ecz

ð7:97Þ

The term ecz denotes a wave travelling in the –z-direction while we assume a wave travelling in the +z-diection. Hence E0o ¼ 0: Thus Exs ðzÞ ¼ Eo ecz

ð7:98aÞ

Substituting for Exs ðzÞ in Eq. 7.95 Es ¼ Eo ecz^i

ð7:98bÞ

Now inserting the time factor eixt and using Eq. 7.87 and then taking the real part h i Eðz; tÞ ¼ Re Es eixt ¼ Re Eo ecz eixt^i h i ¼ Re Eo eaz eiðxtbzÞ^i

ð7:99Þ

Eðz; tÞ ¼ Eo eaz cosðxt bzÞ^i Now using the Maxwell’s Eq. 6.232 (phasor form) Hs ¼

1 r Es ixl

Because the electric ﬁeld has only x-component ^j ^ ^i k @ @ @ r Es ¼ @x @y @z Exs 0 0 @ExS ^ ¼ j @z 1 @Exs ^ Hence Hs ¼ j ixl @z with Esx ¼ Eo ecz Hs ¼ with H ¼ HS eixt

c Eo ecz^j ixl

ð7:100Þ

7.7 Plane Electromagnetic Waves in Lossy Dielectrics

H¼

467

c Eo ecz eixt^j ixl

Substituting for c from Eq. 7.85 and simplifying sﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ r þ ixe az iðxtbzÞ^ Eo e e Hðz; tÞ ¼ j ixl Let rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ixl ixl Eo ¼ and Ho ¼ g¼ r þ ixe c g

ð7:101aÞ

Hðz; tÞ ¼ Ho eaz eiðxtbzÞ^j

ð7:101bÞ

then

Here η, the intrinsic impedance of the medium, is complex and hence we can write g ¼ jgjeih g

ð7:102Þ

With Eqs. 7.101a, 7.102 we can write Eq. 7.101b as Eo az iðxtbzhg Þ^ e e j jgj

ð7:103Þ

Eo az e cos xt bz hg ^j jgj

ð7:104Þ

H¼ Now taking the real part H¼

From Eqs. 7.99, 7.103 we see that E, H are out of phase by angle hη, because intrinsic impedance is complex. The factor eaz in Eqs. 7.99, 7.104 is called attenuation factor. It shows that the amplitude of E, H decreases as z increases. The SI unit of the attenuation constant is Neper per metre (Np/m). Thus E, H attenuate as the electromagnetic wave propagates inside the medium as shown in Fig. 7.5a. From Eqs. 7.102, 7.103 we observe that E leads H (or) H lags E by an angle of hη which is shown pictorially in Fig. 7.5a. In Problem 7.4 we calculate the average Poynting vector for the case electromagnetic wave propagation in lossy dielectrics. From Eq. 7.84

468

7 Plane Electromagnetic Waves

(a)

Fig. 7.5 (a) Propagation of an electromagneic wave in a lossy dielectric. Electric and magnetic ﬁelds attenuate as they propagate inside the material

h ri ixlðr þ ireÞ ¼ x2 le 1 i ¼ x2 lec xe

ð7:105Þ

h ri ec ¼ e 1 i re

ð7:106Þ

) ec ¼ e0 ie00

ð7:107Þ

where

where r ð7:108Þ x ec is called the complex permittivity of the medium. b in Eq. 7.99, 7.104 is the usual wave number given by Eq. 7.24. We know that the conduction current density and displacement current density are given by e0 ¼ e; e00 ¼

Jcs ¼ rEs

ð7:109Þ

@E @t

ð7:110aÞ

Jd ¼ e

) Jds ¼ ixeEs

ð7:110bÞ

where we have used equations E ¼ Es eixt and Jd ¼ Jds eixt in Eqs. 7.110a, 7.110b. Now taking the ratio )

Jcs r ¼ Jds ixe

ð7:111Þ

7.7 Plane Electromagnetic Waves in Lossy Dielectrics

469

Hence the two vectors point in the same direction in space, but they are 90° out of phase in time. This is shown pictorially in Fig. 7.5b. In the ﬁgure we have used Es as reference. The total current density is (adding vectorically). Jts ¼ Jcs þ Jds

ð7:112Þ

Jts ; Jcs ; Jds is plotted in Fig. 7.5c. Jds makes on angle d with Jts. From the ﬁgure it is clear that r jJcs j jrEs j ð7:113Þ tan d ¼ ¼ ¼ jJds j jjxeEs j xe

(b)

(c)

Fig. 7.5 (b) Relationship between conduction current density and displacement current density in a lossy dielectric. (c) Relationship between displacement current density, conduction current density and total current density in a lossy dielectric

470

7 Plane Electromagnetic Waves

Here tan d is known as loss tangent. d is the loss angle of the medium. A perfect dielectric is one in which tan d is very small, i.e. r xe. For a good conductor tan d is very large, i.e. r xe. It should be noted that tan d depends on frequency of the wave. Hence if a medium is a good conductor at low frequencies it shall be a good dielectric at high frequencies. The calculation of Poynting vector for the case of lossy dielectric is given as an exercise (see Problem 7.4). Special Case: For a perfect lossless dielectric r ’ 0; e ¼ eo er ; l ¼ lo lr

ð7:114Þ

pﬃﬃﬃﬃﬃ a ¼ 0; b ¼ r le

ð7:115aÞ

Hence

Also r xe. Hence 7.100 takes the form. rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ pﬃﬃﬃﬃﬃﬃﬃﬃ rﬃﬃﬃ ixl l ixl ¼ g¼ ¼ r þ ixe e ixe

ð7:115bÞ

η is real and hence E, H are in phase.

7.8

Plane Electromagnetic Waves in Good Conductors

As mentioned in the previous section a good conductor is a medium in which r xe. Thus from Eqs. 7.94a and 7.94b rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ xlr a¼b¼ 2

ð7:116Þ

rﬃﬃﬃﬃﬃﬃ xl g¼ \45 r

ð7:117Þ

From Eq. 7.101a

Hence E leads H by 45 . The E and H are given by E ¼ Eo eaz cosðxt bzÞ^i

ð7:118Þ

7.8 Plane Electromagnetic Waves in Good Conductors

H¼

Ho az e cosðxt bz 45 Þ^j jgj

471

ð7:119Þ

Thus as the electromagnetic wave propagates through the conducting medium the amplitude of E and H gets attenuated by a factor of eaz : We deﬁne the penetration depth or skin depth of the electromagnetic wave in the conducting medium as, the distance travelled by the wave in the conducting medium such that the wave amplitude decreases by a factor of e−1 (about 37%). We denote the skin depth as d. Hence Eo ead ¼ Eo e1 )d¼

1 a

ð7:120Þ ð7:121Þ

Substituting Eq. 7.116 in 7.121 1 d ¼ qﬃﬃﬃﬃﬃﬃﬃ ¼ xlr 2

sﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 2 xlr

ð7:122Þ

With x ¼ 2pv where m is the frequency of the em wave we have sﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 1 d¼ pvlr

ð7:123Þ

Let us calculate the skin depth for silver in microwave frequencies. For silver r’107 mho=m: At a typical microwave frequency of 108 Hz the skin depth is 10−4 cm. Thus the depth of penetration of the em wave in silver at microwave frequencies is very small. Depending upon the frequency and conducting material the em wave attenuates within a small distance of the conductor. Thus electromagnetic energy doesn’t penetrate deeper inside the conductor. It travels in the region surrounding the conductor. Hence the conductor merely guides the waves. Suppose we wish to make a waveguide made of silver at microwave frequencies. As we know that the skin depth of em wave at microwaves frequency in silver is 104 cm we can coat silver on brass or glass with the silver coating thickness of 104 cm. Thus the silver coating on brass or glass will serve the purpose. No need to make the entire waveguide of silver. Hence lot of silver is saved; thereby reducing the cost of waveguide.

472

7 Plane Electromagnetic Waves

From 7.116, 7.121, 7.123 and with b¼

2p k

The wavelength can be written as k ¼ 2pd

ð7:124Þ

We know the velocity of the wave is x b ) v ¼ xd

v¼

ð7:125Þ

Example 7.2 A plane wave with frequency of 9 GHz is travelling in free space with an electric ﬁeld of amplitude 1 V/m. (a) Find the velocity and the propagation constant of the em wave. (b) Calculate the characteristic impedance of the medium. (c) Find the amplitude of the magnetic ﬁeld intensity. Solution (a) For free space 1 v ¼ pﬃﬃﬃﬃﬃﬃﬃﬃﬃ lo eo v ¼ 3 108 m=s: k¼

v 3 108 1 ¼ ¼ 101 m: v 9 109 3 ¼ 3:3cm:

Now b ¼

2p k b¼

2p ¼ 1:903 rad=cm 3:3

7.8 Plane Electromagnetic Waves in Good Conductors

473

(b) The characteristic impedance of the medium is rﬃﬃﬃﬃﬃ lo g¼ ¼ 377X eo (c) The amplitude of the magnetic ﬁeld intensity is given by g¼

ðEo Þx ðHo Þy

) ðHo Þy ¼ ðH o Þy ¼

7.9

ðEo Þx g

1 ¼ 2:65 103 A=m: 377

Plane Electromagnetic Waves in Lossy Dielectrics Using Maxwell’s Equation in Phasor Form

The results which we derived in Sect. 7.7 for the case of lossy dielectrics can be derived directly from Maxwell’s equation in phasor form as follows: From Eqs. 6.230–6.233 we can write ðaÞ

r Ds ¼ qs ¼ 0 as qs ¼ 0 With Ds ¼ eEs we have then r Es ¼ 0

ð7:126Þ ð7:127Þ

ðbÞ

r Bs ¼ 0 With Bs ¼ lHs

r Hs ¼ 0

ð7:128Þ

ðcÞ

r Es ¼ ix Bs ) r Es ¼ ixlHs

ð7:129Þ

ðdÞ

r Hs ¼ Js þ ixDs

ð7:130Þ

) r Hs ¼ rEs þ ixeEs

ð7:131Þ

) r Hs ¼ ðr þ ixeÞEs

ð7:132Þ

Taking curl on both sides of Eq. 7.129

474

7 Plane Electromagnetic Waves

r r Es ¼ ixlr Hs

ð7:133Þ

Using vector identity 7 from Sect. 1.20 and using Eqs. 7.132, 7.127 r2 Es ¼ ixlðr þ ixeÞEs

ð7:134Þ

c2 ¼ ixlðr þ ixeÞ

ð7:135Þ

) r2 Es ¼ c2 Es

ð7:136Þ

Let

Here c is the propagation constant of the medium. Similarly using Eqs. 7.132, 7.128 it can be shown that r2 Hs ¼ c2 Hs

ð7:137Þ

Equations 7.136, 7.137 are nothing but Eqs. 7.86a, 7.86b. Hence all the derivations done in Sect. 7.7 from Eqs. 7.87a, 7.87b onwards can be derived from Eqs. 7.136, 7.137 that is from Maxwell’s equation in phasor form.

7.10

Wave Polarization

In Fig. 7.4 we saw that when the wave is propagating along z-axis, the electric ﬁeld E is pointing in x-direction and the magnetic ﬁeld H is pointing in y-direction. “The wave polarization is deﬁned as its electric ﬁeld vector orientation as a function of time, at a ﬁxed position in space” Because the direction of H is related to E by Maxwell’s equation specifying the direction of E alone is sufﬁcient.

Fig. 7.6 A linearly polarized electromagnetic wave

7.10

Wave Polarization

475

Fig. 7.7 An electric ﬁeld vector of an polarized electromagnetic wave

In Fig. 7.4 E is in ﬁxed straight orientation in all times and positions. Such a wave is said to be linearly polarized. In Fig. 7.4 E always lies in the x-z plane. But E can be oriented in any ﬁxed direction in the x-y plane and be linearly polarized. This is shown in Fig. 7.6. H is not shown in Fig. 7.6. If the tip of the electric ﬁeld traces a circle the wave is said to be circularly polarized. If the tip of the electric ﬁeld traces an ellipse the wave is said to be elliptically polarized. Consider Fig. 7.7 in which we show an electric ﬁeld vector E. The electric ﬁeld vector may correspond to a linearly or circularly or elliptically polarized wave. The conditions under which the electric ﬁeld vector corresponds to linearly or circularly or elliptically polarized wave can be derived as follows: In general the electric ﬁeld vector E shown in Fig. 7.7 will be a superposition of two linearly polarized waves-one polarized wave in x-direction and the other polarized wave in the y-direction and lagging in phase by angle / at any value of z. Then we can write E ¼ Exo cosðxt bzÞ^i þ Eyo cosðxt bz þ /Þ^j

ð7:138Þ

Let us examine the direction change of E at z = 0 as t changes. Hence Eq. 7.138 becomes [with z = 0]. E ¼ Exo cos xt^i þ Eyo cosðxt þ /Þ^j

ð7:139Þ

The above equation can be written as E ¼ Ex^i þ Ey^j

ð7:140Þ

Ex ¼ Exo cos xt

ð7:141Þ

with

476

7 Plane Electromagnetic Waves

Ey ¼ Eyo cosðxt þ /Þ

ð7:142Þ

From Eqs. 7.140 and 7.141 cos xt ¼

sin xt ¼

Ex Exo

ð7:143Þ

E cos xt cos / E y

yo

sin /

Ex cos / Ey E Eyo sin xt ¼ xo sin /

ð7:144Þ

Now sin2 xt þ cos2 xt ¼ 1 ¼

Ex Exo

2

2 E EE ðEx =Exo Þ2 cos2 / þ E y 2 E x Ey cos / yo xo yo þ sin2 /

)

Ex Exo

2

Ey þ Eyo

2

2Ex Ey cos / ¼ sin2 / Exo Eyo

ð7:145Þ

ð7:146Þ

(a) Linear Polarization If Ex and Ey are in phase so that / = 0 Eq. 7.145 reduces to

Ex Ey Exo Eyo )

2 ¼0

Ey Eyo ¼ Ex Exo

ð7:147Þ ð7:148Þ

Because Eyo =Exo the ratio between amplitudes is constant Eq. 7.148 represents a straight line as shown in Fig. 7.8. Ex, Ey are given by Eqs. 7.141, 7.142 with / = 0 Ex ¼ Exo cos xt

ð7:149Þ

Ey ¼ Eyo cos xt

ð7:150Þ

The above discussion is at z = 0. If we allow z to vary then the linearly polarized wave will be as shown in Fig. 7.6. In Fig. 7.8

7.10

Wave Polarization

477

Fig. 7.8 Relationship between Ex and Ey for a linearly polarized electromagnetic wave

tan h ¼

Eyo Exo

ð7:151Þ

(b) Circularly Polarized wave If / ¼ p=2 and Exo ¼ Eyo ¼ Eo then Eq. 7.146 reduces to E2x þ E2y ¼ E2o

ð7:152Þ

which is the equation of the circle. The electric ﬁeld E rotates about as shown in Fig. 7.9. Figure 7.9 is a circle. The tip of the E rotates about the circle thereby the wave is a circularly polarized wave. The above discussion is at z = 0 as the time varies. On the other hand if z also varies in addition to time then the circularly polarised wave will be as shown in Fig. 7.10.

478

7 Plane Electromagnetic Waves

Fig. 7.9 Relationship between Ex and Ey for a circularly polarized electromagnetic wave. The electric ﬁeld vector rotates about the circle

Fig. 7.10 Circularly polarized electromagnetic wave

Figure 7.9 shows that E rotates at a uniform rate with an angular velocity x in a counter clockwise direction. By right-hand rule when ﬁngers of the right hand point in the direction of rotation of E then the thumb points in the direction of the wave. This is right hand circularly polarized wave. In Fig. 7.9 thus the wave propagates in the direction of outward drawn normal to the plane of the paper at point O and is a right circularly polarized wave. Hence when Ey lags Ex by p/2 in time phase we get right circularly polarized wave.

7.10

Wave Polarization

479

Fig. 7.11 Relationship between Ex and Ey for a elliptically polarized electromagnetic wave. The electric ﬁeld vector rotates about the ellipse.

(c) Elliptically Polarized wave If / ¼ p=2 and Eox 6¼ Eoy then Eq. 7.146 reduces to

Ex Eox

2

Ey þ Eoy

2 ¼1

ð7:153Þ

which is the equation of ellipse. The tip of the E then rotates about in the form of the ellipse as shown in ﬁgure 7.11. Similar to circularly polarized wave we can deﬁne right and left elliptically polarized wave. An elliptically polarized wave is shown in ﬁgure 7.12 Example 7.3 A plane wave is travelling in a perfect dielectric medium. Its electric ﬁeld intensity is given by Ex ðz; tÞ ¼ 5 cos 2p 107 t 0:2pz V=m (a) Calculate the velocity of propagation. (b) Find the magnetic ﬁeld intensity if l = lo. Solution We know that Ex ðz; tÞ ¼ Eo cosðxt bzÞ Hence x ¼ 2pv ¼ 2p 107 ) v ¼ 107 Hz:

480

7 Plane Electromagnetic Waves

Fig. 7.12 Elliptically polarized electromagnetic wave

2p ¼ 0:2p: k Hence k ¼ 10m: The velocity of propagation is then

Also b ¼

v ¼ vk ¼ 107 10 ¼ 108 m=sec: Velocity in free space = 3 108 m/sec. Hence

pﬃﬃﬃﬃﬃﬃﬃﬃ 3 10 ¼ 3: er lr ¼ 2 10

as lr ¼ 1; er ¼ 3 ¼ 9: rﬃﬃﬃ l 120p ¼ ¼ 40p Now g ¼ e 3 Ex As g ¼ Hy 5 cos 2p 107 t 0:2pz A=m: Hy ¼ 40p Here Ex and Hy are the components of electric and magnetic ﬁelds, respectively.

7.11

Reflection and Transmission of the Plane Wave at Normal Incidence

In the previous section we discussed the propagation of uniform plane waves in unbounded, homogeneous medium. We now consider a monochromatic uniform plane wave travelling from one medium and entering another medium of inﬁnite

7.11

Reﬂection and Transmission of the Plane Wave at Normal Incidence

481

Fig. 7.13 The ﬁgure shows medium 1 and medium 2. An electromagnetic wave is incident normal to the boundary between two mediums. Part of the incident wave is reflected and remaining part is transmitted

extent. Whenever an electromagnetic wave from one medium enters a different medium it is partly reflected and partly transmitted. The parameters e, l; r determine the proportion of the wave reflected and transmitted. In this section we will restrict our discussion for normal incidence of the plane wave. In later sections we shall discuss about oblique incidence of the plane wave. Consider a linearly polarized uniform plane wave travelling in the z-direction. The wave is incident at the interface between two media, the interface being situated at z = 0. Medium 1 present at z < 0 is characterized by r1, e1, l1 and medium 2 present at z > 0 is characterized by r2, e2, l2 as shown in Fig. 7.13. As shown in the same ﬁgure incident wave (Ei, Hi) is travelling in the z-direction in medium 1. The transmitted wave (Et, Ht) is travelling in the z-direction in medium 2. The reflected wave (Er, Hr) is travelling in the –z-direction. We deﬁne the reflection coefﬁcient C as the ratio between the amplitude of the reflected wave Ero and amplitude of the incident wave Eio at the interface z = 0 C¼

Ero Eio

ð7:154Þ

The transmission coefﬁcient T is deﬁned as the ratio between the amplitude of the transmitted wave Eto and amplitude of the incident wave Eio at the interface z = 0. T¼

Eto Eio

ð7:155Þ

482

7 Plane Electromagnetic Waves

The electric and magnetic ﬁelds of the incident wave in phasor form are Eis ðzÞ ¼ Eio ec1 z ^i

ð7:156Þ

His ðzÞ ¼ Hio ec1 z ^j Eio c1 z ^ ¼ e j g1

ð7:157Þ

and

noting down that the incident wave is in medium 1. The reflected wave is travelling in the –z-direction in medium 1. Hence

and

Ers ðzÞ ¼ Ero ec1 z ^i

ð7:158Þ

Hrs ðzÞ ¼ Hro ec1 z ð^jÞ Ero c1 z ^ ¼ e j g1

ð7:159Þ

because in Fig. 7.13 Hr is pointing in the –y-direction. The transmitted wave is travelling in the +z-direction in medium 2. Thus

and

Ets ðzÞ ¼ Eto ec 2z ^i

ð7:160Þ

Hts ðzÞ ¼ Hto ec 2z ^j Eto c2 z ^ ¼ e j g2

ð7:161Þ

In medium 1 we have both incident and reflected ﬁelds, whereas in medium 2 only transmitted ﬁeld exists. Denoting the ﬁelds in medium 1 and medium 2 as E1s, H1s and E2s, H2s, respectively, we have E1s ¼ Eis þ Ers

ð7:162Þ

H1s ¼ His þ Hrs

ð7:163Þ

E2s ¼ Ets

ð7:164Þ

H2s ¼ Hts

ð7:165Þ

7.11

Reﬂection and Transmission of the Plane Wave at Normal Incidence

483

Now the boundary conditions require that tangential components to be continuous at the interface [assuming no charges and currents at the interface (Eqs. 6.181, 6.183) Thus E1tan ¼ E2tan

ð7:166Þ

H1tan ¼ H2tan

ð7:167Þ

at the interface at z = 0. As the electromagnetic waves are transverse both the electric and magnetic ﬁelds are tangential at the interface without any normal components. Hence substituting Eqs. 7.162–7.165 in Eqs. 7.166, 7.167 E1s ð0Þ ¼ E2s ð0Þ

ð7:168Þ

H1s ð0Þ ¼ H2s ð0Þ

ð7:169Þ

Eis ð0Þ þ Ers ð0Þ ¼ Ets ð0Þ

ð7:170Þ

From Eq. 7.168

Using Eqs. 7.156, 7.158, 7.160 then Eio þ Ero ¼ Eto

ð7:171Þ

His ð0Þ þ Hrs ð0Þ ¼ Hts ð0Þ

ð7:172Þ

From Eq. 7.169

Using Eqs. 7.157, 7.159, 7.161 then 1 Eto ðEio Ero Þ ¼ g1 g2

ð7:173Þ

where we have already noted that Eio, Ero and Eto are the amplitudes of the incident, reflected and transmitted waves at the interface z = 0. Dividing Eq. 7.171 throughout by Eio and using Eqs. 7.154, 7.155 1þC ¼ T

ð7:174Þ

Dividing Eq. 7.173 throughout by Eio and using Eqs. 7.154, 7.155 1C¼

g1 T g2

ð7:175Þ

484

7 Plane Electromagnetic Waves

From Eqs. 7.174, 7.175 C¼

g2 g1 g2 þ g1

ð7:176Þ

T¼

2g2 g1 þ g2

ð7:177Þ

and

In general C and T are complex. Special Cases (a) Perfect Dielectric–Conductor Boundary Let us consider that medium 1 is a perfect dielectric and medium 2 is a perfect conductor. Then for medium 1 r1 ¼ 0 and for medium 2 r2 ¼ 1: From Eq. 7.117 g2 ¼

rﬃﬃﬃﬃﬃﬃﬃﬃ xl2 ¼0 r2

ð7:178Þ

as r2 is inﬁnity. Hence from Eqs. 7.176, 7.177 C ¼ 1 T¼0

ð7:179Þ

Thus the wave is totally reflected back into medium 1 without any transmission into medium 2. On physical grounds the same result should be expected. We know that the electric ﬁeld inside the conductor is zero. Hence whatever energy is brought by the incident wave is simply reflected back. The amplitudes of the incident and reflected wave are thus equal and together form a standing wave pattern. The standing wave consists of two travelling waves (incident and reflected) which are travelling in opposite direction. The electric ﬁeld of the standing wave can be obtained from Eq. 7.162 E1s ¼ Eis þ Ers

ð7:180Þ

Substituting Eqs. 7.156, 7.158 in Eq. 7.180 E1s ¼ ½Eio ec1 z þ Ero ec1 z ^i

ð7:181Þ

7.11

Reﬂection and Transmission of the Plane Wave at Normal Incidence

485

Substituting Eqs. 7.154, 7.179 in Eq. 7.181 E1s ¼ Eio ½ec1 z þ Cec1 z ^i

ð7:182Þ

E1s ¼ Eio ½ec1 z ec1 z ^i

ð7:183Þ

From Eqs. 7.88, 7.115a [a1 = 0] c1 ¼ a1 þ ib1 ¼ ib1

ð7:184Þ

Substituting Eq. 7.184 in 7.183 E1s ¼ Eio eib1 z eib1 z ^i

ð7:185Þ

) E1s ¼ 2iEio sin b1 z ^i

ð7:186Þ

Now inserting the time part E1 ¼ Re E1s eixt

ð7:187Þ

E1 ¼ 2 Eio sin b1 z sin xt^i

ð7:188Þ

Hence

By following similar procedure H1 ¼

2Eio cos b1 cos xt^j g1

ð7:189Þ

In Fig. 7.14 we sketch the electric and magnetic ﬁelds given by Eqs. 7.188, 7.189. In Fig. 7.14a we plot the electric ﬁeld along x-axis. In Fig. 7.14b we plot the magnetic ﬁeld along y-axis. x-axis and y-axis are orthogonal to each other in Fig. 7.14. For the sake of explanation they are drawn in the same direction. Figure 7.14 shows that the ﬁelds oscillate with time and they do not travel. Thus E and H are mutually orthogonal to each other and represent standing waves. (b) Perfect Dielectric–Perfect Dielectric Boundary Let the medium 1 and medium 2 be perfect lossless dielectric. In that case r1 ¼ r2 ¼ 0: Hence from Eqs. 7.87, 7.114a a1 ¼ 0; a2 ¼ 0; c1 ¼ ib1 ; c2 ¼ ib2

ð7:190Þ

Hence from Eq. 7.115b g1 and g2 are real and hence from Eqs. 7.176, 7.177 C and T are real.

486

7 Plane Electromagnetic Waves

Fig. 7.14 Standing waves pattern for (a) electric ﬁeld and (b) magnetic ﬁelds for an electromagnetic wave reflected at the perfect dielectric - conductor boundary

Because both the mediums are perfect dielectrics part of the incident wave is transmitted into medium 2 and part of the wave is reflected back into medium 1. Hence whatever energy is brought by the incident wave is divided between the reflected wave and transmitted wave. Thus in this case the amplitudes of the incident and reflected waves are unequal. Energy division between the reflected and transmitted waves is determined by the nature of the dielectric mediums. Let us calculate the electric ﬁeld in the medium 1. From Eq. 7.162 E1s ¼ Eis þ Ers

7.11

Reﬂection and Transmission of the Plane Wave at Normal Incidence

487

Using Eqs. 7.156, 7.158 E1s ¼ Eio ec1 z ^i þ Ero ec1 z ^i

ð7:191Þ

E1s ¼ Eio ½ec1 z þ Cec1 z ^i

ð7:192Þ

E1s ¼ Eio eib1 z þ Ceib1 z ^i

ð7:193Þ

Using Eq. 7.154

Using Eqs. 7.190

Using the Euler’s formula eih ¼ cos h þ i sin h in Eq. 7.194 E1s ¼ Eio ½ð1 þ CÞ cos b1 z ið1 CÞ sin b1 z^i

ð7:194Þ

Considering the magnitude jE1s j ¼ Eio

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ð1 þ CÞ2 cos2 b1 z þ ð1 CÞ2 sin2 b1 z

ð7:195Þ

Now let us consider two cases: Case (i): Let g2 [ g1 ; C [ 0 then the maximum values of jE1s j occur at b1 zmax ¼ np

ð7:196Þ

where n = 0, 1, 2, 3… The negative sign indicates the wave is travelling in the z-direction. With the help of Eq. 7.24 we can write Eq. 7.196 as zmax ¼

nk1 2

ð7:197Þ

Minimum values of jE1s j occur at b1 zmin ¼ ð2n þ 1Þp=2

ð7:198Þ

ð2n þ 1Þ k1 4

ð7:199Þ

) zmin ¼ with n = 0, 1, 2, 3…

488

7 Plane Electromagnetic Waves

Hence from Eqs. 7.195 jE1s jmax ¼ Eio ð1 þ CÞ

ð7:200Þ

jE1s jmin ¼ Eio ð1 CÞ

ð7:201Þ

Case (ii): Let g2 \g1 then C\0: Then using same procedure from Eqs. 7.196–7.201 we can show that jE1s jmax ¼ Eio ð1 CÞ

ð7:202Þ

jE1s jmin ¼ Eio ð1 þ CÞ

ð7:203Þ

Hence in this case the locations of jE1s jmax and jE1s jmin are interchanged with respect to case (i) C [ 0: The incident and reflected waves of unequal amplitudes form a standing wave in the medium 1. The ratio jE1s jmax to jE1s jmin is called standing wave ratio S. Hence S¼

jE1s jmax 1 þ jCj ¼ jE1s jmin 1 jCj

) jCj ¼

S1 Sþ1

ð7:204Þ ð7:205Þ

Calculation of H1 and Poynting vector is given as a problem. See Problem 7.6.

7.12

Reflection and Transmission of the Plane Wave at Oblique Incidence

In the previous section we discussed the reflection of the plane wave at normal incidence. In this section we will discuss the reflection of the plane wave at oblique incidence. Oblique incidence refers to the case in which the electromagnetic wave is incident on a plane boundary at some arbitrary angle. Consider two mediums separated by an interface at z = 0 as shown in Fig. 7.15. The plane containing the normal to the interface and direction of propagation of incident wave is called “Plane of incidence”. The plane of incidence is shown in Fig. 7.15. In general the electric ﬁeld vector of the incident wave Ei can make an arbitrary angle with plane of incidence.

7.12

Reﬂection and Transmission of the Plane Wave at Oblique Incidence

489

Fig. 7.15 (a) Oblique incidence of an electromagnetic wave: An electromagnetic wave is on the boundary between two mediums. Part of the incident wave is reflected and remaining part is transmitted. (b) Perpendicular polarization: The electric and magnetic ﬁelds of the incident , reflected and transmitted wave is shown in the ﬁgure

However as discussed in Sect. 1.9a the Ei can be resolved into two components one component perpendicular to the plane of incidence and the other parallel to the plane of incidence. We will consider each case separately. (a) Perpendicular Polarization Consider Fig. 7.15. In this ﬁgure there are two mediums separated by an interface at z = 0. In Fig. 7.15b the electric ﬁeld of the incident wave is perpendicular to the plane of incidence. The case in which Ei is perpendicular to the plane of incidence is called perpendicular polarization or s-polarized wave. Let the two media be linear, isotropic and homogeneous but ﬁnitely conducting media. The electromagnetic

490

7 Plane Electromagnetic Waves

Fig. 7.16 The unit vector in the direction of the incident wave and the radius vector r is shown in the ﬁgure

wave is incident on the boundary between the two medium and partly gets reflected and partly gets transmitted. The electric ﬁeld of the incident wave is given by (in phasor form) Eis ¼ Eio ec1 ðn^i rÞ^i ð7:206Þ ^i the unit vector in the direction of Consider Fig. 7.16 where we show the n incident wave. r in the ﬁgure is the radius vector which points from O to P1. Using Eq. 1.10 we can write ^ þ y^j r ¼ zk

ð7:207Þ

^i . That is G points from P1 Let G be an vector which points in the direction of n to O. Using Eq. 1.10 and Fig. 7.16 ^ G ¼ Gy^j þ Gz k

ð7:208Þ

From Fig. 7.16 and Eq. 1.11 we can write cos hi ¼

Gz Gy ; sin hi ¼ jGj jGj

ð7:209Þ

7.12

Reﬂection and Transmission of the Plane Wave at Oblique Incidence

491

Substituting Eq. 7.209 in 7.208 ^ G ¼ jGj sin hi^j þ jGj cos hi k

ð7:210Þ

^ ) jGj^ ni ¼ jGj sin hi^j þ jGj cos hi k

ð7:211Þ

^ ^i ¼ sin hi^j þ cos hi k )n

ð7:212Þ

Substituting Eqs. 7.207, 7.212 in Eq. 7.206 Eis ¼ Eio ec1 ðz cos hi y sin hi Þ^i

ð7:213Þ

From the Maxwell’s equation r Es ¼ ixl1 Hs We obtain the magnetic ﬁeld as His ¼

i Eio h^ k sin hi þ ^j cos hi ec1 ðz cos hi y sin hi Þ g1

ð7:214Þ

1 where g1 ¼ ixl c1 is the intrinsic impedance of the medium (See Eq. 7.100a). Now consider the reflected wave. Let us assume the electric ﬁeld is still polar^r be the unit vector in the direction of propagation of ized in the x-direction. Let n the reflected wave as shown in Fig. 7.17. The electric ﬁeld of the reflected wave is given by

Ers ¼ Ero ec1 ðn^r :rÞ^i

ð7:215Þ

In Fig. 7.17 r is the radius vector pointing from O to P2. Hence ^ þ y^j r ¼ zk

ð7:216aÞ

^r : That is N points from O to Let N be a vector which points in the direction of n P2. Hence from Fig. 7.17 ^ N ¼ Ny^j þ Nz k

ð7:216bÞ

From Fig. 7.17 and Eq. 1.11 we can write cos hr ¼

Nz Ny ; sin hr ¼ jNj jNj

ð7:217Þ

492

7 Plane Electromagnetic Waves

Fig. 7.17 The unit vector in the direction of the reflected wave and the radius vector r is shown in the ﬁgure

Substituting Eq. 7.217 in Eq. 7.216a we obtain ^ N ¼ jNj sin hr^j jNj cos hr k

ð7:218Þ

^ ) jNj^ nr ¼ jNj sin hr^j jNj cos hr k

ð7:219Þ

^ ^r ¼ sin hr^j cos hr k )n

ð7:220Þ

Substituting Eqs. 7.220, 7.216a in Eq. 7.215 Ers ¼ Ero ec1 ðy sin hr z cos hr Þ^i

ð7:221Þ

From the Maxwell’s equation (Eq. 6.232) $ Es ¼ ixl1 Hs We obtain the magnetic ﬁeld as Hrs ¼

i Ero h ^ cos hr^j ec1 ðy sin hr z cos hr Þ sin hr k g1

ð7:222Þ

7.12

Reﬂection and Transmission of the Plane Wave at Oblique Incidence

493

Fig. 7.18 The unit vector in the direction of the transmitted wave and the radius vector r is shown in the ﬁgure

Let us now consider the transmitted wave. Let us assume that the electric ﬁeld of the transmitted wave is polarized in the x-direction. ^t be the unit vector in the direction of propagation of the transmitted wave Let n as shown in Fig. 7.18. The electric ﬁeld of the transmitted wave is given by Ets ¼ Eto ec2 ðn^t :rÞ^i

ð7:223Þ

In Fig. 7.18 the radius vector r points from O to P3. Hence ^ þ y^j r ¼ zk

ð7:224Þ

^t . That is U points from O to Let U be an vector which points in the direction of n P3 in Fig. 7.18 Hence ^ U ¼ Uy^j þ Uz k

ð7:225Þ

From Fig. 7.18 and Eq. 1.11 we can write cos ht ¼

Uz Uy ; sin ht ¼ jUj jUj

ð7:226Þ

494

7 Plane Electromagnetic Waves

Hence as usual we get ^ U ¼ jUj sin ht^j þ jUj cos ht k

ð7:227Þ

^ ^t ¼ sin ht^j þ cos ht k )n

ð7:228Þ

Substituting Eqs. 7.227, 7.224 in Eq. 7.223 Ets ¼ Eto ec2 ðy sin ht þ z cos ht Þ^i

ð7:229Þ

From the Maxwell’s equation r Es ¼ ixlHs We obtain the magnetic ﬁeld as Hts ¼

i Eto h ^ þ cos ht^j ec2 ðy sin ht þ z cos ht Þ sin ht k g2

ð7:230Þ

2 Here g2 ¼ ixl c2 is the intrinsic impedance of medium 2. At the boundary z = 0 the boundary condition is

E1 tan ¼ E2 tan

ð7:231Þ

where E1 tan ; E2 tan are the tangential components of the electric ﬁelds in medium 1, medium 2, respectively. Noting down that in medium 1 there are incident and reflected waves and in medium 2 only transmitted wave using Eqs. 7.213, 7.215, 7.221 in 7.230 Eio ec1 ðy sin hi Þ þ Ero ec1 ðy sin hr Þ ¼ Eto ec2 ðy sin ht Þ

ð7:232Þ

Using Eqs. 7.154, 7.155 in Eq. 7.239 ec1 ðy sin hi Þ þ Cec1 ðy sin hr Þ ¼ Tec2 ðy sin ht Þ

ð7:233Þ

The above equation must be valid for all values of y that is every where on the interface. At y = 0 Eq. 7.232 becomes 1þC ¼ T

ð7:234Þ

For Eqs. 7.233, 7.231 to hold for all values of y c1 sin hi ¼ c1 sin hr ¼ c2 sin ht

ð7:235Þ

7.12

Reﬂection and Transmission of the Plane Wave at Oblique Incidence

495

In Eq. 7.234 from the ﬁrst equality hi ¼ hr

ð7:236Þ

Hence angle of incidence is equal to angle of reflection. This is the well known Snell’s law of reflection. Also from Eq. 7.234 c1 sin hi ¼ c2 sin ht

ð7:237Þ

which is well known Snell’s law of refraction. Assuming that there are no currents at the boundary at z = 0 the boundary condition for H is H1 tan ¼ H2 tan

ð7:238Þ

Noting down that the y components are the only tangential components for H in Eqs. 7.214, 7.222, 7.230 we can then write Eq. 7.238 as ½cos hi c1 y sin hi C½cos hr c1 y sin hr T½cos ht c2 y sin ht e e ¼ e g1 g1 g2

ð7:239Þ

where we have used Eqs. 7.154, 7.155 in Eq. 7.239. Equation 7.239 can be simpliﬁed with Eq. 7.235 as cos hi C cos hr T cos ht ¼ g2 cos hi g1 g1

ð7:240Þ

With hi ¼ hr we can write the above equation as 1C¼

g1 cos ht T g2 cos hi

ð7:241Þ

From Eqs. 7.234, 7.241 we get C? ¼

g2 cos hi g1 cos ht g2 cos hi þ g1 cos ht

ð7:242Þ

T? ¼

2g2 cos hi g2 cos hi þ g1 cos ht

ð7:243Þ

496

7 Plane Electromagnetic Waves

Fig. 7.19 Parallel polarization: The electric and magnetic ﬁelds of the incident, reflected and transmitted wave is shown in the ﬁgure.

where ? in the above equations denote that the reflection and transmission coefﬁcients corresponding to perpendicular polarizations. (b) Parallel Polarization Let us now consider the case in which the electric ﬁeld is parallel to the plane of incidence as shown in Fig. 7.19. When E is parallel to the plane of incidence the situation is known as parallel polarization or p-polarized. The magnetic ﬁeld His of the incident wave is given by (in phasor form). His ¼ Hio ec1 ðn^i :rÞ^i

ð7:244Þ

Following the similar procedures as from Eqs. 7.206–7.213 we get His ¼ Hio ec1 ðz cos hi y sin hi Þ^i

ð7:245Þ

Now consider the Maxwell’s equation (in current-free region (Eq. 6.233) r Hi ¼ ixec Ei where ec the complex permittivity of the dielectric is given by Eq. 7.105. From the above Maxwell’s equation Eis ¼

h i c1 ^ ec1 ½z cos hi y sin hi ðHio Þ cos hi^j þ sin hi k ixec1

ð7:246Þ

7.12

Reﬂection and Transmission of the Plane Wave at Oblique Incidence

497

From Eqs. 7.85, 7.106 we get pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ixl1 ðr1 þ ixe1 Þ

ðixÞ r1 þixixe1 rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ c1 ixl1 ¼ ¼ g1 ixec1 r1 þ ixe1 c1 ¼ ixecl

ð7:247Þ

from Eq. 7.101a. Substituting Eq. 7.247 in 7.246 h i ^ ec1 ½z cos hi y sin hi Eis ¼ Hio g1 cos hi^j þ sin hi k

ð7:248Þ

In the similar manner we obtain the ﬁelds associated with the reflected and transmitted waves as Hrs ¼ Hro ec1 ðy sin hr z cos hr Þ^i h i ^ ec1 ðy sin hr z cos hr Þ Ers ¼ Hro g1 cos hr^j þ sin hr k

ð7:249Þ

Hts ¼ Hto ec2 ðy sin ht þ z cos ht Þ^i h i ^ ec2 ðy sin ht þ z cos ht Þ Ets ¼ ðHto Þg2 cos ht^j þ sin ht k

ð7:251Þ

ð7:250Þ

ð7:252Þ

Let Eio ¼ Hio g1 ; Ero ¼ Hro g1 Eto ¼ Hto g2

ð7:253Þ

and using Eqs. 7.154, 7.155 we can write Eio c1 ðz cos hi y sin hi Þ^ e i g1 h i ^ ec1 ðz cos hi y sin hi Þ Eis ¼ Eio cos hi^j þ sin hi k His ¼

Eio C c1 ðy sin hr z cos hr Þ^ e i g1 h i ^ ec1 ðy sin hr z cos hr Þ Ers ¼ CEio cos hr^j þ sin hr k Hrs ¼

ð7:254Þ ð7:255Þ ð7:256Þ ð7:257Þ

498

7 Plane Electromagnetic Waves

Eio T c2 ðy sin ht þ z cos ht Þ e g2 h i ^ ec2 ðy sin ht þ z cos ht Þ Ets ¼ Eio T cos ht^j þ sin ht k Hts ¼

ð7:258Þ ð7:259Þ

Now using Eqs. 7.254, 7.256, 7.258 in the boundary condition for H at z = 0, i.e. Eq. 7.238 ec1 ðy sin hi Þ þ Cec1 ðy sin hr Þ ¼

g1 T c2 ðy sin ht Þ e g2

ð7:260Þ

The above equation must be valid for all values of y. Hence for y = 0 1þC ¼

g1 T g2

ð7:261Þ

For Eq. 7.260 to be valid for all values of y c1 sin hi ¼ c2 sin hr ¼ c2 sin ht

ð7:262Þ

which simply reproduces Snell’s law of reflection and refraction. Noting down that the y components are the only tangential components for E in Eqs. 7.255, 7.257, 7.259 applying the boundary condition 7.231 at the boundary at z = 0 and using Eq. 7.262 1C¼T

cos ht cos hi

ð7:263Þ

From Eqs. 7.261, 7.263 we get Ck ¼

g1 cos hi g2 cos ht g1 cos hi þ g2 cos ht

ð7:264Þ

Tk ¼

2g2 cos hi g1 cos hi þ g2 cos ht

ð7:265Þ

where || in the above equations denote that the reflection and transmission coefﬁcients corresponding to parallel polarization.

7.13

Total Reﬂection and Total Transmission

7.13

499

Total Reflection and Total Transmission

E2 Let us consider that medium 2 is a conductor. In that case g2 ¼ H ¼ 0 as there are 2 no ﬁelds inside the conductor. Then from Eqs. 7.242, 7.264 C? ¼ 1; Ck ¼ 1: Hence all the wave is reflected without any transmission. Irrespective of the incident angle or polarization total reflection occurs. We deﬁne the refractive index or index of refraction as

n¼

pﬃﬃﬃﬃﬃﬃﬃﬃ er lr

ð7:266Þ

Consider the case of dielectric–dielectric interface with r1 ¼ 0; r2 ¼ 0 and l1 ’ l2 ’ lo : In that case c1 ib1 ¼ ¼ c2 ib2

rﬃﬃﬃﬃﬃﬃ er1 n1 ¼ er2 n2

ð7:267Þ

where n1, n2 are the refractive index of medium 1, medium 2, respectively. From Eqs. 7.266, 7.237 sin ht ¼

n1 sin hi n2

ð7:268Þ

Now 2

3

1=2 6 sin hi 7 cos ht ¼ 1 sin2 ht ¼ 41 2 5 n2 n1 2

ð7:269Þ

2 If sin hi ¼ n n1 in that case we observe that cos hi ¼ 0: Hence C? ¼ 1; Ck ¼ 0 meaning that the incident wave suffers total reflection. 2 Another case is when sin hi [ n n1 and hence cos ht is imaginary. Consider Eq. 7.242

C? ¼

a2 ija1 j a2 þ ija1 j

ð7:270Þ

where a2 ¼ g2 cos hi

ð7:271Þ

a1 ¼ ig1 cos ht

ð7:272Þ

500

7 Plane Electromagnetic Waves

) C2? ¼ 1: The same is true for Ck as the reader may observe from Eq. 7.264. This means that the wave suffers total power reflection whenever sin hi [ n2 =n1 : From the above discussion we can write the incident wave suffers total reflection when sin hi

n2 n1

ð7:273Þ

We deﬁne the critical angle for total reflection as hi hc

ð7:274Þ

where hc is given by sin hc ¼

n2 n1

ð7:275Þ

For the case of total transmission, reflection coefﬁcient must be zero. That is C ¼ 0: Hence from Eq. 7.264 for Ck to be zero g1 cos hi ¼ g2 cos ht

ð7:276Þ

The angle hi ¼ hBk when the value of Ck goes to zero is called Brewster angle. At this angle the parallel component of the ﬁeld of wave suffers total transmission. Using Eq. 7.267 and with l1 = l2 Eq. 7.276 can be written as sin hBk ¼

1

e1 e2

2

ð7:277Þ

1 sin hBk ¼ rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ

ﬃ 1 þ ee12

ð7:278Þ

2

1

e1 e2

For the case of perpendicular polarization for total reflection to be zero that is C? ¼ 0 from Eq. 7.242 g2 cos hi ¼ g1 cos ht

ð7:279Þ

The corresponding angle hi ¼ hB? is the Brewster angle for no reflection for the case of perpendicular polarization.

7.13

Total Reﬂection and Total Transmission

501

From Eqs. 7.279, 7.276 we get sin2 hB? ¼

1 l1 e2 =l2 e1

2 1 ll1

ð7:280Þ

2

Brewster’s angle given by Eqs. 7.277, 7.280 for the case parallel and perpendicular polarizations are completely different. Hence if an unpolarized wave is incident on a media it is possible to separate parallel and perpendicular polarizations from the unpolarized wave. Suppose assume that the light incident on a medium satisﬁes Eq. 7.277 then the perpendicular polarization component will be reflected. The above facts have applications in reducing glare, controlling polarization of laser light, etc. Thus still now we have discussed perpendicular and parallel polarizations, where E is perpendicular and parallel to the plane of incidence as shown in Fig. 7.15b and 7.19, respectively, thereby deriving their reflection and transmission coefﬁcients. Now consider Fig. 7.20 where E is neither parallel nor perpendicular to the plane of incidence. In that case we can resolve E into two components, E? the perpendicular component and E? the parallel component, the reflection and transmission coefﬁcients are given by Eqs. 7.242, 7.243, respectively, and that of Ek is given by Eqs. 7.264, 7.265. The results for E in Fig. 7.20 can then be obtained by superposing the result of E? and Ek :

Fig. 7.20 Oblique incidence of an electromagnetic wave: The electric ﬁeld is neither parallel nor perpendicular to the plane of incidence. In that case both reflection and transmission coefﬁcients can be obtained by resolution of electric ﬁeld

502

7.14

7 Plane Electromagnetic Waves

Dispersion—Group Velocity

In the previous sections we discussed the propagation of electromagnetic waves in conductors, dielectrics, reflection and transmission of these waves at the boundaries. The reader would have observed that depending upon r, e, l the electromagnetic waves get reflected, transmitted, attenuated, etc., in the medium. However to some extent the parameters r, e, l depend on the frequency of the incident wave. Hence depending upon the frequency of the wave the wave behaviour inside the medium gets modiﬁed. We know that the velocity of the wave is given by v¼

x 1 ¼ pﬃﬃﬃﬃﬃ b el

v is called phase velocity of the wave. Because now e, l depend on x, waves of different frequency travel with different phase velocities in a medium. This phenomenon is known as dispersion. A given carrier wave will have its main frequency and will be associated with a minor spread in frequencies. Therefore the electromagnetic wave has a group of frequencies and forms a wave packet. The envelope of the wave packet travels with a single velocity known as group velocity as shown in Fig. 7.21. Let the electric ﬁeld of the electromagnetic wave be given by Eðz; tÞ ¼ Eo cosðxt bzÞ

ð7:281Þ

Consider the above wave to be wave packet having equal amplitudes but with two frequencies xo þ Dx; xo Dx and the corresponding propagation constants are bo þ Db; bo Db: For frequency xo þ Dx the electric ﬁeld is E1 ¼ Eo cos½ðxo þ DxÞt ðbo þ DbÞz

ð7:282Þ

For frequency xo Dx the electric ﬁeld is E2 ¼ Eo cos½ðxo DxÞt ðb DbÞz

Fig. 7.21 A wave packet

ð7:283Þ

7.14

Dispersion—Group Velocity

503

The net electric ﬁeld is E ¼ E1 þ E2 ¼ 2Eo cosðxo t bzÞ cosðDxt DbzÞ

ð7:284Þ

The phase velocity vp shown in Fig. 7.21 is the carrier velocity. That is velocity of the carrier wave inside the envelope and is given by v¼

dz xo ¼ dt bo

ð7:285Þ

The group velocity is the velocity of envelope of the wave packet as shown in Fig. 7.21 and is given by vg ¼

dz Dx ¼ dt Db

ð7:286Þ

Example 7.4 Consider a electromagnetic wave travelling in the z-direction with peak amplitude of Eio ¼ 700V=m: The wave is incident normally form medium 1 to medium 2 the interface being situated at z = 0. Given e1 ¼ 12 1012 F=m; l1 ¼ 2:5 1006 H=m; r1 ¼ 0: e2 ¼ 7 1012 F=m; l2 ¼ 5 106 H=m; r2 ¼ 0; x ¼ 108 rad=s: Find Eis ; Ers Solution Let us calculate the intrinsic impedance of the medium rﬃﬃﬃﬃﬃ rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ l1 2:5 106 ¼ 456:4 g1 ¼ ¼ e1 12 1012 r ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ﬃ rﬃﬃﬃﬃﬃ l2 5 106 g2 ¼ ¼ 845:2 ¼ e2 7 1012

504

7 Plane Electromagnetic Waves

Now 1 v1 ¼ pﬃﬃﬃﬃﬃﬃﬃﬃﬃ e1 l1 1 ¼ pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 12 1012 2:5 106 v1 ¼ 1:8 108 m=s: Hence x1 108 ¼ v1 1:8 108 b1 ¼ 0:6 b1 ¼

Hence Eis ¼ Eio cosðxt b1 zÞ^i ¼ 700 cos 108 t 0:6z ^iV=m Assuming the electric ﬁeld is in the x-direction Ero g2 g1 ¼ Hence C ¼ Eio g2 þ g1 C¼

845:2 456:4 ¼ 0:3 845:2 þ 456:4

Hence Ero ¼ C Eio ¼ 210V=m: Ers ¼ 210 cos 108 t 0:6z ^iV=m: Example 7.5 Consider a electromagnetic wave travelling in a medium. The electric ﬁeld of the wave is given by E ¼ 0:7ez=2 cos 108 t bz ^i V=m: Determine (a) b (b) the loss tangent. Given er ¼ 9; lr ¼ 2:5:

7.14

Dispersion—Group Velocity

505

Solution Because E ¼ Eo eaz cosðxt bzÞ^i we get 1 a¼ ; 2

x ¼ 108

(a) The velocity of the wave is 1 1 v ¼ pﬃﬃﬃﬃﬃ ¼ pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ¼ 0:63 108 m=s le lo lr eo er x 108 ¼ v 0:63 108 ¼ 1:59rad=m

b¼

( ðbÞ

a2 ¼ x2

le 2

"rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ #)

r 2 þ1 1 xe

( "rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ #) 2

r 2 8 2 1 1 ) ¼ 10 þ1 1 2 xe 2 ð0:63 108 Þ2 r ) ¼ 0:66 is the loss tangent: xe Example 7.6 For a medium er ¼ 52; lr ¼ 1; r ¼ 25f=m: Calculate the propagation constant, attenuation constant, phase constant for the medium at x ¼ 1011 rad=s: The medium is a lossy dielectric. Solution r 25 ¼ xe 1011 52 8:85 1012 ¼ 0:5432:

506

7 Plane Electromagnetic Waves

Now c2 ¼ ixlðr þ ixeÞ rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ﬃ pﬃﬃﬃﬃﬃ r ) c ¼ ix le 1i xe Now pﬃﬃﬃﬃﬃ pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ le ¼ lr er lo eo pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ¼ 1 52 4p 107 8:85 1012 ¼ 24:042 109 : Hence pﬃﬃﬃﬃﬃ x le ¼ 2404:2 Thus pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ c ¼ i 2404:2 1 i 0:5432 Now 1 i 0:5432 ¼ reih ¼ 1:138ei 28:51 ) ð1 i0:5432Þ1=2 ¼ ð1:138Þ1=2 ei

28:51 2

¼ 1:0339 i 0:262 Hence c ¼ i 2404:2ð1:0339 i 0:262Þ ) c ¼ 629:9 þ i 2485:7 ¼ a þ ib is the propagation constant. Thus a ¼ attenuation constant ¼ 626:9 b ¼ phase constant

¼ 2485:7

Exercises

507

Example 7.7 In Example 7.6 calculate the wavelength of the wave and the intrinsic impedance of the medium. Solution 2p 2 3:14 ¼ b 2485:7 k ¼ 2:52mm:

k¼

Now rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ rﬃﬃﬃsﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ixl l 1 ¼ g¼ r þ ixe e 1 iðr=xeÞ rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃrﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 4p 107 1 g¼ 8:85 1012 52 1 i0:5432 ) g ¼ 47:4 12:02i is the intrinsic impedance.

Exercises Problem 7.1 For water er ¼ 78 at 300 MHz. If water is assumed to be a lossless medium and 300 MHz of em wave is propagating through water calculate H. Given Eo ¼ 0:25V=m l1 ¼ 1 Eo is the amplitude of the electric ﬁeld. Problem 7.2 An em wave of frequency 70 Hz is propagating in copper whose conductivity is 5:8 107 s=m: Calculate v, k attenuation constant and b. Problem 7.3 A plane TEM wave travelling in z-direction and E and H in x-direction and y-direction have a power density of 1.4 W/m2 in a medium with lr ¼ 1 and er ¼ 3: Find the amplitude of E and H. Problem 7.4 Calculate the average value of Poynting vector for the case of lossy dielectric using Eqs. 7.59 and 7.74 separately and verify the statement that exponential is easier to work with than time-harmonic functions.

508

7 Plane Electromagnetic Waves

Problem 7.5 From the results obtained in Sect. 7.7 for the case of lossy dielectric obtain the results for the case of free space that we derived in Sect. 7.3 with r ¼ 0; e ¼ eo ; l ¼ lo . Problem 7.6 For the case of two perfect dielectrics discussed in Eqs. 7.189–7.204 calculate H1 and hence Poynting vector. Problem 7.7 A plane wave is propagating in a medium. The electric ﬁeld of the wave is E ¼ 0:7ez=2 cos 109 t bz ^i V=m and er ¼ 7; lr ¼ 2: Calculate the (a) loss tangent (b) wave velocity (c) wave impedance. Problem 7.8 A electromagnetic wave is travelling in free space in z-direction with a frequency of 2 MHz. Its electric ﬁeld is pointing in the x-direction and the peak value of the ﬁeld is given by 1:4pðmv=mÞ: Obtain the expressions for Eðz; tÞ and Hðz; tÞ: Problem 7.9 A 1000 MHz standing wave pattern exists in a non-magnetic dielectric. The distance between maximum and its adjacent minimum is 1.6 cm. Compute the value of er. Problem 7.10 A perpendicularly polarised electromagnetic wave travels from medium 1 to medium 2. Medium 2 is vacuum. For the case of medium 1. er1 ¼ 7:0; lr ¼ 1; r ¼ 0: The angle of incidence is 16°. If Eio ¼ 2lV=m: Calculate Ero, Hro, Hto. Problem 7.11 Calculate the reflection and transmission coefﬁcients when an electromagnetic wave is incident normally on a dielectric from free space. Given l2 ¼ lo ; er2 ¼ 4: Problem 7.12 For a dielectric er ¼ 2:5; l ¼ lo : From free space air is incident on the dielectric with hi ¼ 29 : Calculate the angle of transmission ht.

Chapter 8

Transmission Lines

8.1

Introduction

In the previous chapter we saw plane electromagnetic wave propagation in free space. The wave propagating in free space is an unguided wave and propagates in all possible directions. These type of unguided wave propagation will be helpful when there are multiple receivers located at different points. Good examples of such wave propagation are radio or TV broadcasting systems. On the other hand for a point (source) to point (load) transmission such unguided wave propagation will be inefﬁcient. For such transmission, electromagnetic energy should be transferred in a well-deﬁned path without any spreading. A good example where such transmission is helpful is in a situation like telephone conversation-wave propagation in such transmission is said to be guided. Typical examples in which the wave propagation is guided from source to load in a well-deﬁned path are transmission line and waveguides. Transmission lines are discussed in this chapter and waveguides in the next chapter. A transmission line consists of two or more parallel conductors used to transmit electromagnetic energy. The most common types of transmission lines are shown in Fig. 8.1. Figure 8.1a shows a coaxial cable which consists of a inner conductor and a conducting sheath separated by a dielectric medium. Figure 8.1b shows a parallel-plate transmission line separated by a dielectric slab of uniform thickness. Two-wire transmission line shown in Fig. 8.1c consists of two wires separated by a uniform distance. Figure 8.1d shows a microstrip line.

© Springer Nature Singapore Pte Ltd. 2020 S. Balaji, Electromagnetics Made Easy, https://doi.org/10.1007/978-981-15-2658-9_8

509

510

8 Transmission Lines

Fig. 8.1 Different types of transmission lines a A coaxial cable. b A parallel plate transmission line. c A two-wire transmission line. d Microstrip line

8.2

Description of a Transmission Line

In Sect. 6.2 we saw that when the key K is closed (Fig. 6.2) a current flows in the circuit. It takes a small time duration for the current to flow in the circuit as soon as the key K is closed. However due to small dimensions of the circuit the time delay is negligible, and for all practical purposes the current flow is instantaneous. On the other hand assume the resistance R in Fig. 6.2 lies thousands of miles away and is connected to the battery. When the key K is closed a current will start flowing in the circuit, but now time delay effects become appreciable. From the point of closing the key and for the current to flow in the resistance R a propagation delay is introduced due to large dimensions of the circuit. The time duration required for the flow of current in the circuit thus becomes appreciable. Let us assume that in Fig. 6.2 the resistance lying at a thousands of miles away is absent. Or in other words the circuit is having large dimensions, but the other end is open. Will there be current in the circuit. To answer the above question the following points should be noted. A transmission line used in circuits of large dimensions has capacitance, inductance, resistance and a ﬁnite conductivity.

8.2 Description of a Transmission Line

511

(1) Any pair of two conductors separated by a insulating medium forms a capacitor. (2) The current flowing through the conductors produces a magnetic ﬁeld around it. The flux of the magnetic ﬁeld is linked to the conductor through inductance. (3) Each conductor has a resistance. (4) The dielectric separating the two conductors in the transmission line has ﬁnite conductivity. Due to large dimensions of the transmission line we characterize the above parameters on per unit length basis. (1) Let C be the capacitance per unit length of the transmission line. (2) Let L be the inductance per unit length of the transmission line. (3) Let R be the resistance per unit length of the transmission line. [Not to be confused with R in Fig. 6.2] (4) Let G be the conductance per unit length of the dielectric medium separating the conductors. Note that G 6¼

1 R

Like resistors, capacitors have speciﬁc symbols in electronic circuits, similarly transmission lines have speciﬁc symbols. Irrespective of whatever be the shape of the transmission line its symbolic representation is parallel wire conﬁguration as shown in Fig. 8.2a.

Fig. 8.2 a Symbolic representation of a transmission line. b Circuit model of a transmission line

512

8.3

8 Transmission Lines

Wave Propagation in a Transmission Line

Transmission lines allow transverse electromagnetic wave (TEM) to propagate along their length. The electric ﬁeld, magnetic ﬁeld and direction of propagation in such TEM waves are mutually perpendicular to each other. Consider that in Fig. 8.2a in the parallel wire representation the transmission line is a coaxial cable. Suppose in Fig. 8.2a the source is switched on and it generates a signal. A TEM wave propagates along the transmission line towards the load in Fig. 8.2a. The electric ﬁeld and magnetic ﬁeld of coaxial line is shown in Fig. 8.3. The electric ﬁeld points in the radial direction from the inner conductor to the outer conductor. The magnetic ﬁeld circles the inner conductor. Switching on the source in Fig. 8.2a establishes a voltage and current in the transmission line. The voltage and current established in the transmission line are related to the electric and magnetic ﬁelds through Eqs. 2.65 and 5.16. Z V ¼ E dl I I ¼ H dl Thus the TEM wave propagation in the transmission line can be described in terms of electric and magnetic ﬁelds or in terms of voltage and current (electric circuit theory). As we will see in future sections the lumped parameter model leading to transmission line equations describe the wave propagation in transmission lines in terms of voltage and current waves. In Sect. 8.11 we will describe the wave propagation in parallel-plate transmission line in terms of ﬁelds.

Fig. 8.3 Electric and magnetic ﬁelds for a TEM wave propagation in a coaxial transmission line

8.4 Transmission Line Equations

8.4

513

Transmission Line Equations

As explained, transmission line problems can be solved using electromagnetic theory and electric circuit theory. In this section we will solve the transmission line problem using electric circuit theory. For the circuit shown in Fig. 6.2 and similar circuits the dimensions are so small compared to the wavelength k of the electromagnetic wave they can be considered as lumped parameter circuits. But for the circuit shown in Fig. 8.2a, b their dimensions are very large as compared to the wavelength of the electromagnetic wave and their circuit parameters are distributed over a length. In those distributed parameter circuits we divide the length of the transmission line of small elements Δz such that Dz\\k: and analyse them using usual methods of circuit theory. One such division is shown in Fig. 8.4. Noting down that Kirchhoff’s law cannot be applied to distributed circuits but can be applied to lumped parameter circuits we use Kirchhoff’s voltage law (KVL) and Kirchhoff’s current law (KCL) to analyse Fig. 8.4. Applying KVL to Fig. 8.4 Vðz; tÞ ¼ ðRDzÞIðz; tÞ þ ðLDzÞ

@Iðz; tÞ þ Vðz þ Dz; tÞ @t

ð8:1Þ

where V(z, t), I (z, t) and V (z + Δz, t), I (z + Δz, t) are the instantaneous voltages and currents at z, z + Δz, respectively. The above equation can be written as Vðz þ Dz; tÞ Vðz; tÞ @Iðz; tÞ ¼ R Iðz; tÞ þ L Dz @t

ð8:2Þ

In the limit Δz ! 0

Fig. 8.4 Equivalent circuit for small length of the two conductor transmission line

@Vðz; tÞ @ Iðz; tÞ ¼ R Iðz; tÞ þ L @z @t

ð8:3Þ

514

8 Transmission Lines

Applying KCL to point A in Fig. 8.4 Iðz; tÞ ¼ ðGDzÞVðz þ Dz; tÞ þ ðCDzÞ

@ Vðz þ Dz; tÞ þ Iðz þ Dz; tÞ @t

Iðz þ Dz; tÞ Iðz; tÞ @ Vðz þ Dz; tÞ ¼ G Vðz þ Dz; tÞ þ C Dz @t

ð8:4Þ ð8:5Þ

In the limit Δz ! 0

@Iðz; tÞ @ Vðz; tÞ ¼ G Vðz; tÞ þ C @z @t

ð8:6Þ

Assuming harmonic time dependence and using the phasor notation Vðz; tÞ ¼ Re Vs ðzÞeixt

ð8:7Þ

Iðz; tÞ ¼ Re Is ðzÞeixt

ð8:8Þ

With the aid of Eqs. 8.7 and 8.8, Eqs. 8.3 and 8.6 become

dVs ¼ ðR þ ixLÞIs dz

ð8:9Þ

dIs ¼ ðG þ ixCÞVs dz

ð8:10Þ

In Eqs. 8.9 and 8.10 the quantities Vs and Is are coupled. From Eqs. 8.9 and 8.10

d2 V s dIs ¼ ðR þ ixLÞ d z2 dz

ð8:11Þ

d2 I s dVs ¼ ðG þ ixCÞ 2 dz dz

ð8:12Þ

Substituting Eqs. 8.10 in 8.11 and 8.9 in 8.12 d2 V s ¼ ðR þ ixLÞ ðG þ ixCÞVs d z2

ð8:13Þ

d2 I s ¼ ðR þ ixLÞ ðG þ ixCÞIs d z2

ð8:14Þ

Thus Vs, Is coupled in Eqs. 8.9, 8.10 has been separated in Eqs. 8.13 and 8.14.

8.5 Wave Propagation in the Transmission Line—Circuit Model

8.5

515

Wave Propagation in the Transmission Line—Circuit Model

Equations 8.13 and 8.14 can be written as d2 Vs c2 Vs ¼ 0 dz2

ð8:15Þ

d2 I s c2 I s ¼ 0 dz2

ð8:16Þ

where c ¼ a þ ib ¼

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ðR þ ixLÞ ðG þ ixCÞ

ð8:17Þ

Equations 8.15 and 8.16 are having similar form as that of Eqs. 7.87a, 7.87b. Hence we conclude Eqs. 8.15 and 8.16 are wave equations for voltage and current, respectively. c in Eq. 8.17 is the propagation constant where a is the attenuation constant (Nq/m) and b is the phase constant (rad/m). a and b are given by a ¼ Re½c hpﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃi ¼ Re ðR þ ixLÞ ðG þ ixCÞ

ð8:18Þ

b ¼ Im½c hpﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃi ¼ Im ðR þ ixLÞ ðG þ ixCÞ

ð8:19Þ

The solutions for Eqs. 8.15 and 8.16 are similar to Eq. 7.97 cz Vs ðzÞ ¼ Voþ ecz þ V o e

ð8:20Þ

cz Is ðzÞ ¼ Ioþ ecz þ I o e

ð8:21Þ

and

The e–cz term represents a forward travelling wave in the z-direction, and ecz term represents a backward travelling wave in the −z-direction. V+o , V–o, I+o , I–o are the wave amplitudes. The forward and backward travelling waves are shown in Fig. 8.5. The instantaneous expression for the voltage is then, Vðz; tÞ ¼ Re Vs ðzÞeixt

ð8:22Þ

516

8 Transmission Lines

Fig. 8.5 Forward and backward travelling wave in a transmission line

cz ixt ¼ Re Voþ ecz þ V e oe

ð8:23Þ

Noting down that c = a + ib and in general Voþ and V o are complex Vðz; tÞ ¼ Re

n

o ða þ ibÞz e Voþ eða þ ibÞz þ V eixt o

ð8:24Þ

Characteristic impedance As the source in Fig. 8.2 is switched on, voltage and current waves start propagating in the transmission line. Let us assume that at a given instant of time the disturbance produced travels and reaches point A in Fig. 8.2. At that instant the source “feels” the impedance of the transmission line only and not the impedance of the load. The impedance of the transmission line is called the characteristic impedance. The characteristic impedance Zo of the transmission line is deﬁned as the ratio of voltage to current for the wave propagating in the z-direction. Substituting Eq. 8.20 in Eq. 8.9 d þ cz cz V e þ V ¼ ðR þ ixLÞIs o e dz o

ð8:25Þ

cVoþ cV o ecz ecz ðR þ ixLÞ ðR þ ixLÞ

ð8:26Þ

) Is ¼

Comparing Eq. 8.26 with 8.21 Ioþ ¼

cVoþ R þ ixL

ð8:27Þ

I o ¼

cV o R þ ixL

ð8:28Þ

8.5 Wave Propagation in the Transmission Line—Circuit Model

517

From Eq. 8.27 and 8.28 we see that Zo ¼

Voþ V R þ ixL o þ ¼ ¼ c Io Io

ð8:29Þ

Substituting Eq. 8.17 in Eq. 8.29 rﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ R þ ixL Zo ¼ G þ ixC

ð8:30Þ

The wavelength k and wave velocity v are given by, 2p b

ð8:31Þ

x ¼ tk b

ð8:32Þ

k¼ v¼

We will now consider special cases of wave propagation in transmission lines.

8.6

Lossless Line

In the case of lossless propagation whatever power is launched by the source into the transmission line arrives to load without any loss. There is no dissipation along the transmission line, and the wave propagation in the transmission is said to be lossless. For no dissipation to occur the resistance of the conductor R and the conductivity of the dielectric G should be zero. Thus R = 0, G = 0. The lossless transmission line is shown in ﬁgure 8.6 with R = G = 0. Initially when the source is switched on in Fig. 8.6 no instant voltage appears at the load. Instead current increases in La until Ca is charged, next current in Lb increases charging Cb, and this sequential charging process continues until the voltage appears at the load. Thus as the source is switched on the voltage doesn’t appear instantaneously everywhere on the line but rather travels from the source to the load at a certain velocity. The velocity of this wave can be estimated as follows.

Fig. 8.6 Lumped circuit model of a lossless transmission line

518

8 Transmission Lines

For a lossless line, R¼G¼0

ð8:33Þ

pﬃﬃﬃﬃﬃﬃﬃ a þ ib ¼ ix LC

ð8:34Þ

From Eq. 8.17

Comparing real and imaginary parts on both sides a¼0

ð8:35Þ

pﬃﬃﬃﬃﬃﬃﬃ c ¼ ib ¼ ix LC

ð8:36Þ

Thus the velocity of the wave is v¼

x 1 ¼ pﬃﬃﬃﬃﬃﬃﬃ b LC

ð8:37Þ

The characteristic impedance of the lossless transmission line is Zo ¼

rﬃﬃﬃﬃ L C

ð8:38Þ

We have discussed about the propagation of plane electromagnetic waves in lossy dielectric in Sect. 7.7. From Eq. 7.83a, r2 E r l

@E @2E ¼ el 2 @t @t

ð8:39Þ

For a lossless dielectric the conductivity r is zero. Hence in one dimension the above equation can be written as @2E @2E ¼0 el 2 @z @t

ð8:40Þ

Comparing Eq. 8.40 with Eq. 7.1 1 v ¼ pﬃﬃﬃﬃﬃﬃ el

ð8:41Þ

Comparing Eq. 8.41 with Eq. 8.37 l e ¼ LC

ð8:42Þ

8.7 Low-loss Line

8.7

519

Low-loss Line

In Sect. 8.5 we discussed the propagation of the wave in a lossy transmission line. a, attenuation constant given by Eq. 8.18 and the phase velocity given by Eq. 8.32 are in general dependent on frequency. In the case of lossless propagation (Sect. 8.6) a, and phase velocity given by Eqs. 8.35 and 8.37, respectively, are independent of frequency. There is an another special case in which R xL; G xC

ð8:43Þ

The above case is called low-loss propagation. Under this condition Eq. 8.17 is 1

1

c ¼ ðR þ ixLÞ2 ðG þ ixCÞ2 1 1 1 1 R 2 G 2 ¼ ðixLÞ2 ðixCÞ2 1 þ 1þ ixL ixC 12 12 pﬃﬃﬃﬃﬃﬃﬃ R G c ¼ ix LC 1 þ 1þ ixL ixC

ð8:44Þ

As R xL; G xC applying binomial approximation pﬃﬃﬃﬃﬃﬃﬃ c ’ ix LC 1 þ pﬃﬃﬃﬃﬃﬃﬃ c ’ ix LC 1 þ

Neglecting the

R 2ixL

G 1þ 2ixC

R G RG þ þ 2ixL 2ixC 4x2 LC

ð8:45Þ ð8:46Þ

1 term x2

pﬃﬃﬃﬃﬃﬃﬃ c ix LC 1 þ

R G þ 2ixL 2ixC rﬃﬃﬃﬃ rﬃﬃﬃﬃ! pﬃﬃﬃﬃﬃﬃﬃ 1 C L R c ix LC þ þG 2 L C

Noting down that c ¼ a þ ib 1 a¼ R 2

rﬃﬃﬃﬃ rﬃﬃﬃﬃ! C L þG L C

pﬃﬃﬃﬃﬃﬃﬃ b ¼ x LC

ð8:47Þ

ð8:48Þ

ð8:49Þ ð8:50Þ

520

8 Transmission Lines

The phase velocity is given by v¼

x 1 ¼ pﬃﬃﬃﬃﬃﬃﬃ b LC

ð8:51Þ

From Eq. 8.30 the characteristic impedance is given by 1

Zo ¼

ðR þ ixLÞ2 1

ðG þ ixCÞ2

pﬃﬃﬃﬃﬃﬃﬃﬃ 0 ixL 1 þ Zo ¼ pﬃﬃﬃﬃﬃﬃﬃﬃ @ ixC 1 þ

11 R 2 ixLA G i xC

rﬃﬃﬃﬃ 1 1 L R 2 G 2 1þ Zo ¼ 1þ C ixL ixC

ð8:52Þ

ð8:53Þ

Using binomial approximation rﬃﬃﬃﬃ L R G Zo 1þ 1 C 2ixL 2ixC Neglecting the

1 x2

term rﬃﬃﬃﬃ L 1 R G 1þ Zo C 2ix L C

8.8

ð8:54Þ

ð8:55Þ

Distortionless Line

Transmission lines in general are used to transmit signals from source to receiver. The signal which carries the information usually has a band of frequencies. For the case of lossless propagation a the attenuation constant is zero (Eq. 8.35). So there is no attenuation. The phase velocity given by Eq. 8.37 is independent of frequency. It means that all the waves with different frequencies travel at the same velocity. For the case of low-loss propagation the attenuation constant a given by Eq. 8.49 is approximately independent of frequency. Hence all waves with different frequencies are attenuated to the same extent. The phase velocity v given by Eq. 8.51 is approximately independent of frequency. Thus waves with different frequencies travel with same velocity. However under low-loss condition when the transmission line is lengthy, loss becomes

8.8 Distortionless Line

521

Fig. 8.7 Distortion of the input signal in a long dispersive line

appreciable and a, v depends on frequency. The result is waves of different frequencies get attenuated to a different extent and travel with different velocities. Therefore the input signal carrying information in a band of frequencies get distorted as shown in Fig. 8.7. Even though the line is lossy if the following condition is satisﬁed then the transmission line would become distortionless R G ¼ L C

ð8:56Þ

The above condition is known as Heaviside condition. The propagation constant as given by Eq. 8.17 is c¼

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ðR þ ixLÞðG þ ixCÞ

sﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ R G þ ix þ ix c ¼ LC L C Using Eq. 8.56

ð8:57Þ ð8:58Þ

pﬃﬃﬃﬃﬃﬃﬃ R þ ix c ¼ LC L

ð8:59Þ

rﬃﬃﬃﬃ pﬃﬃﬃﬃﬃﬃﬃ C c¼R þ ix LC L

ð8:60Þ

As c ¼ a þ ib; from Eq. 8.60 we can write rﬃﬃﬃﬃ C a¼R L pﬃﬃﬃﬃﬃﬃﬃ b ¼ x LC

ð8:61Þ ð8:62Þ

The phase velocity is then v¼

x 1 ¼ pﬃﬃﬃﬃﬃﬃﬃ b LC

ð8:63Þ

522

8 Transmission Lines

The attenuation constant a given by Eq. 8.61 and phase velocity v given by Eq. 8.63 are independent of frequency. Thus all frequencies are attenuated to an equal extent and travel with the same velocity. Therefore when the condition given by 8.56 is met then the line is distortionless. For a distortionless line the characteristic impedance given by Eq. 8.30 is sﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ R þ ixL Zo ¼ RC þ ixC L rﬃﬃﬃﬃ L Zo ¼ C

8.9

ð8:64Þ

ð8:65Þ

Relationship Between G and C

For a given transmission line the inductance per unit length L and capacitance per unit length C are related by Eq. 8.42. Assuming that the conductor in the transmission line is having very high conductivity we can set R = 0 in Eq. 8.17. Then c¼

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ixLðG þ ixCÞ

1 pﬃﬃﬃﬃﬃﬃﬃ G 2 c ¼ ix LC 1 þ ixC

ð8:66Þ ð8:67Þ

For a lossy dielectric from Eq. 7.85 c¼

pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ðix lÞðr þ ixeÞ

ð8:68Þ

The above equation can be written as pﬃﬃﬃﬃﬃ r 12 c ¼ ix le 1 þ ixe

ð8:69Þ

Comparing Eqs. 8.67 and 8.69 LC ¼ le

ð8:70Þ

G r ¼ C e

ð8:71Þ

Equation 8.71 gives the relationship between conductance per unit length G and capacitance per unit length C of the transmission line.

8.10

8.10

Determination of Transmission Line Parameters

523

Determination of Transmission Line Parameters

(A) Parallel-plate transmission line The parallel-plate transmission line is shown in Fig. 8.1b. Let d be the separation between the two plates in Fig. 8.1b. Let “a” be the width of the transmission line. Then from the Eq. 3.333 the capacitance per unit length of the transmission line can be written as ea C ¼ F=m ð8:72Þ d In the above equation the fringing ﬁelds of the transmission line have been neglected by assuming d