Preface tthe Fine Eton these chapters was carefully selected to contain only what is essential ‘The last two and one-half chapters may be used in a semester course oF 8 additional reading. Considerable effort was made to give a clear ‘explanation of Landau damping—one that does not depend on a know! ‘edge of contour integration. I am indebted to Tom O'Neil and George Schmide for help in simplifying the physical picture originally given by John Dawson. Some readers will be distressed by the use of egs electrostatic units is, of cours, senseless co argue about units; any experienced physicist can defend his favorite system eloquently and with faultless logic. The system here is explained in Appendix I and was chosen to void unnecessary writing ofc, and éo, as well at to be consistent with the ‘majority of research papers in plasma physics, | would like to thank Miss Lisa Tatar and Mrs. Betty Rae Brown for a highly intuitive job of deciphering my handwriting, Me, ‘Tim Lambert for a similar degree of understanding in the preparation of the drawings, and most of all Ande Chen for putting up with a large number of deserted evenings, Francis F, Chen Los Angeles, 1974 Preface to the Second Edition CONTENTS Preface to the First Edidon x 1. INTRODUCTION 1 Geewrence of Plasmas in Nature 1 «Definition of Plame $ Concept of Tempers «Debye Skalding Se The Plena Portmeter “Tl + Criaria for Pasar Ie Apptiatins of Posma Phos 13 SINGLE-PARTICLE MOTIONS 19 Inpadacin “19+ Cniform E and B Fields 19 + Nomuniform BiFied 25+" Nonunform E Fie 36° + Time-Varying Fiele 30" « "Tine eying B Fidd Sl ©” Summary of Cutting Geter Drie “a9 Aad tavariant 43 |. PLASMAS AS FLUIDS 53 Iuoduction 83» Relation of Plasma Phycier to Ontinary Electomag ties SH + The Fluid Equation of Mion 38 8 Fld Defi Perpendicular B68» eid Dif Peale B75 + The Plasma Appraximation 77 PT4. WAVES IN PLASMAS 0 Repranition of Waste 79» Growp Velucity 81+ Plasma Onis "82's" Elewoe Plame Wawss 87.1 Sound Wows "Ot "+ fon Woon 95. + Valtityof the Plasma Apfresima Hien 98.” « " Gimparzon of fon ond Elton Watts 89. s Bacr atc Elcton Oxilations Perpendicalarto 100. + Electaste fm Wace, Perpendicular to B 100 + The Lower Hid Freeney Vie + Blsronaeati Waves with yd 1d 0 priest Apc ton + "Bectromagntic Waves Pependiclar By 122s Calo and Resmeeness 126+ Elcomagnetic ‘Waves Purell Be 128 «Experimental Conequeness TSI + Hydromagnee Waves 136." » “Magnetvnic Waver 142» Sammaryof Elements Plasma Waces UM ne TheCMA Diagram 116 5, DIFFUSION AND RESISTIVITY 155 Diffasion end Mabiiy in WeohlyLnived Gases 155» Day ofa Plame ‘by Difiwion 180s Steady Sia Salutions 185 + Recon fon “161 «Difwionacros'a Megnatc Fld 169° Calon in Filly Ioised Plumes 176.» The SingleFlui ACHD Equations St 5 Dipuson in Fully Ionized Plasmas 185+ Solution ofthe Difution Equation “188” « "Bohm Difason ond Newclasicl Difusion 180 6, EQUILIBRIUM AND STABILITY 199 Introduction 199+ HydromaguaicEquilibriym 201+ The Cone aptaf e203 6” Difuion of Monee Field ino a Plasma 205+ Chasteason of fetailer 208 + TuoSteam Inatabiety 311 The *Graviatinal” Inability 215 "+ Resistive Drift Waves 218 The Weibel Instability 223 7. KINETIC THEORY 225 The Meaning of 6) 225 «Equant of Kinaic Theory 290 Deaton of he Fld Equctios” 296. + Plasma Oslin end Landay Damping 240.» "The Meaning of Landau Damping 3430+ a ‘Phe! Derivation of Landau Deming 258 = BCK and Van Kampen Modes O61” "Exteimental Verifcation 962 «Ton Landon Damp ing 267" +” Kinet Efe na Mogwai Feld 214 8, NONLINEAR EFFECTS 287 Iran “387+ Sheathe 290» Lom Acoustic Shock Wang 207 + The Pmderomative Force’ 905. Parametric states 309° «Plame ‘Fehr 394° «Nonlinear Landon Damping 828” Equations of Nolinear Plasma Physics "380 APPENDICES cnc Appendix A. Units, Conon and Formulas, Ver Relations 949 Apher B, Theor of Weees in «Cold Uniform Plasma 955 Appendic C. Semple Three-Hour Final Exam $01 [Aphendis D. Anowers @ Some Paslens 369 Index a7 Index to Problems anINTRODUCTION TO PLASMA PHYSICS AND CONTROLLED FUSION SECOND EDITION Volume I: Plasma PhysicsChapter One INTRODUCTION OCCURRENCE OF PLASMAS IN NATURE Tt has often been said that 99% of the matter in the universe isin the plasma state; that is, in the form of an elecirifed gas with the atoms dissociated into positive ions and negative electrons. This estimate may rot be very accurate, but ic is certainly & reasonable one in view of the fact that stellar interiors and atmospheres, gaseous nebulae, and much of the interstellar hydrogen are plasmas. fp our own neighborhood, as soon as one leaves the earth's atmosphere, one encounters the plasma comprising the Van Allen radiation belts and the solar wind, On the other hand, in our everyday lives encounters with plasmas are limited to 2 few examples: the flash of a lightning bolt, the soft glow of the Aurora Borealis, the conducting gas inside a fluorescent tube of neon sign, and the slight amount of ionization in a rocket exhaust, It would seem that we live in the 18 of the universe in which plasmas do not ‘The reason for this can be seen from the Saha equation, which tells ustheamountof ionization tobe expected in agasin thermal equilibrium: 4x to DE user omy Here; and n are, respectively, the density (number perm’) of ionized atoms and of neutral atoms, Tis the gas temperature in °K, K is Boltzmann's constant, and U isthe ionization energy of the gas—that Ma2 (Chaier One jp FIGURE Lt is, the number of ergs required to remove the outermost electron from an atom. (The mks or International System of units will be used in this book.) For ordinary air at room temperature, we may take m= 3x10" m™* (ee Problem 1-1), T=300°K, and Ui» l45eV (for nitrogen), where 1 eV = 1.6% 10"! The fractional ionization m/v + 14) = nutty predicted by Bq, (1-1 is liculousy lowe =10" As the temperature is raised, the degree of ionization remains low until U; is only a few times KT: Then ny/n, rises abruptly, and the gas isin a plasma state. Further increase in temperature makes ny, less than ‘and the plasma eventually becomes fully ionized. This isthe reasan plasmas exit in astronomical bodies with temperatures of millions of degrees, but not on the earth. Life could not easly coexist with a plasma—at least, plasma of the type we are talking about. The natural ‘occurrence of plasmas at high temperatures isthe reason for the designa- tion “the fourth state of mater.” Although we do not intend to emphasize the Saha equation, we should point out its physical meaning. Atoms in a gas have a spread of thermal energies, and an atom is ionized when, by chance, it sullers a a Mhustrating the long range of electrostatic force in a plasma. collision of high enough energy to knock out an electron, In a cold gas, such energetic collisions occur infrequently, since an atom must be accelerated to much higher than the average energy by a series of “favorable” collisions. The exponential factor in Eq, [I-1] expresses the fac thatthe number of fast atoms falls exponentially with U./KT. Once an atom is ionized, it remains charged unl it meets an electron; it then very likely recombines with the electron to become neutral again. The recombination rate clearly depends on the density of electrons, which we can take as equal to ny. The equilibrium ion density therefore, should decrease with nj: and this is the reason for the factor nj! on the right-hand side of Eq (1-1 The plasma inthe interstellar medium owes its existence to the low value of n; (about | per em’), and hence the low recombination rate. DEFINITION OF PLASMA ‘Any fonized gas cannot be called a plasma, of course: there is always some small degree of ionization in any gas. A useful definition is as follows: A plasma is « quasineutral gas of charged and neutral particles which exhibits cllective behavior ‘We must now define “quasineutral” and “collective behavior.” The meaning of quasineutrality will be made cleat in Section 1.4. What is ‘meant by “collective behavior" isa follows Consider the forces acting on a molecule of, say, ordinary air. Since the molecule is neutral, there is no net electromagnetic force on it, and the force of gravity is negligible. The molecule moves undisturbed until it makes a collision with another molecule, and these collisions control the particle's motion. A macroscopic force applied toa neutral gs, such as from a loudspeaker generating sound waves, is tansmicied (0 the individual atoms by collsions. The situation is totaly different in a plasma, which has charged particles. As these charges move around, they can generate local concentrations of positive or negative charge. which sive rise to electric fields. Motion of charges also generates currents, and hence magnetic fields. These fields affect the motion of other charged particles far away. Let us consider the effect on each other of two slightly charged regions of plasma separated by a distance r (Fig. Io). The Coulomb force between A and B diminishes as 1/r". However, for «given solid angle (hat is, r/r » constant), the volume of plasma in B that can affect 12 3 Introduction(Ghaper One 13 A increases as r°, Therefore, elements of plasma exert a force on one another even at large distances, It is this long-ranged Coulomb force that gives the plasma a large repertoire of posible motions and enriches {he field of study known as plasma physics. In fact, the most interesting Fesults concern so-called “collisionless” plasmas, in which the long-range 0) = ma, we Find na exp (/KT,) This equation will be derived with more physical insight in Section 8.5, Substituting for n, and m, in Bq, [1-13], we have fe-ofeo())- In the region where |¢¢/KT,| « 1, we can expand the exponential in a f41 (a6) 4...] open lee a ash ° Intradction10 No simplification is possible for the region near the grid, where |e/ KT ‘may be large. Fortunately, this region does not contribute much to the thickness of the cloud (called a sheath), because the potential falls very rapidly there, Keeping only the linear terms in Eq, [1-18], we have ran} Defining 15, where m stands for nm, we can write the solution of Eq. (1-14]a8 8 = doexp (lel /Aa) a8 ‘The quantity Ap, called the Debye length, is 2 measure ofthe shielding distance or thickness of the sheath, Note that as the density is increated, Ap decreases, as one would ‘expect, since each layer of plasma contains more electrons, Furthermore, Ap increases with increasing KT,. Without thermal agitation, the charige cloud would collapse to an infinitely thin layer. Finally itis the electron ‘emperature which is used in the definition of dy because the electrons, being more mobile than the ions, generally da the shielding by moving so a8 co create a surplus or deficit of negative charge, Only in special silyations i this not true (ee Problem 1-3). ‘The following are useful forms of Eq (1-15); Ap * 6(T/n)""m, Tink Ap= 743K T In)!" mm, KT ine We are now in 2 position t define “quasineutrality” If the dimensions L of a system are much larger than Ap, then whenever local ‘concentrations of charge arise or external potentials are introduced into i system, these are shielded out ina digtance short compared with leaving the bulk of the plasma free of large electric potentials or fields, Outside of the sheath on the wall or on an obstacle, V°4 is very small, and n; is equal to m,. typically, to better than one part in 10% It takes ‘nly a small charge imbalance to give rise to potentials of the order of KT], The plasma is “quasineutral’; that is, neutral enough so that one ‘ean cake mon, =n, where m is a common density called the plasma an density, but not so neutral tha all the interesting electromagnetic forces vanish, A criterion for an ionized gus to be a plasma is that it be dense enough that Ap is much smaller than L, ‘The phenomenon of Debye shielding also occurs—in modified form—in single-species systems, such asthe electron streame in klystrons and magnetrons or the proton beam in a cyclotron. In such cases. any local bunching of particles causes large unshielded electri feld unless the density is extremely low (which it often is). An externally imposed Potential—from a wire probe, for instance—would be shielded out by an adjustment of the density near the electrode. Single-speces systems, for unneutralized plasmas, are not strictly plasmas; but the mathematical tools of plasma physics can be used to study such systems, ‘THE PLASMA PARAMETER ‘The picture of Debye shielding that we have given above is valid only i there are enough particles in the charge cloud. Clearly, if there are only one or two particles in the sheath region, Debye shielding would not be a statistically valid concept. Using Eq (I-17), we can compute the number No of parties in a “Debye sphere” Noe nbeab = 138% 10° ja" (Pinky as In addition to Ay « L, “collective behavior” requires Not 9 CRITERIA FOR PLASMAS We have given two conditions that an ionized gas must satisfy to be called plasma. A third condition has toda with collisions. The weakly ionized tas in a jet exhaust, for example, does not qualify as a plasma because the charged particles collide so frequently with neutral atoms that th motion is controlled by ordinary hydrodynamic forces rather than by I for the gas to behave like a plasma rather than a neutral gas. 1s 16 u Inersction12 Chaper One PROBLEMS ‘The three conditions plasma must satisfy are therefore Lape L. 2 Np 1 Sorel 413. On a log-log plot of vs. KT, with n, from LO" eo 10" mand KT; from 0.4 co 10%ev, draw lines of constant 2y and No, On this graph, place the following points (x ia ms, KT in eV} 1. Tyleal fasion reactor!» = 22 Typical fusion experiments = 10" 1600 pinch. ‘TypicaLionosphere: n = 10", KT = 0.08, ‘Typical glow discharge: n = 10”, KT = 2 ‘Typical fame: n = 10", KT = 0.1 ‘Typical Ce plasma; ='10", Interplanetary space: n= 10%, KT = 0.01 KT ~ 100 corus: = 10%, KT = Convince yourself that thee are plasmas 14, Compute the pressure, in atmospheres and intone, exerted by a ther- ‘momuclear plasma on ts container. Assume KT, = KT, = 20 keV. = 10" m", and p= nKT, where T= T,+ 7, 1.5, Ina stil steady state situation, both the ions and the electrons wl follow the Bolan relation y= necnp (-4s/KT) For the ase ofan infinite, transparent grid charged co «potential & show that the shielding distance then given approximately by g(t) Berta) Show tha determined by he temperate ofthe colder pes. 16, An alerntve derivation of Ao wil ive further ivgh 10 its meaning Consider two infinite, paralel plates x= ='td, set st potential @ = 0. The space between them is uniformly fille by a gas of density n of panicles of charge (a) Using Poisson's equation, show thatthe potential distribution between the plats is Migs fe (Show that for d> Ao, the enerny needed to transport 2 particle from a plate to the midplane is greater than the average Kinetic coerey ofthe particles 1. Compute Ap and Ny fr the following cases (@) A gow discharge, with n = 10m”, KT, = 2e¥. () The earth's ionosphere wth a = 10" Mm", KT, = 0.1e¥, (2 A epinch, with = 10m", AT, = 800ev, APPLICATIONS OF PLASMA PHYSICS Plasmas can be characterized by the two parameters n and KT, Plasma applications cover an extremely wide range of x and KT, varies over 28 orders of magnitude from 10" to 10" m3, and KT can vary over seven orders from 0.1 to 10° eV. Some ofthese applications are discussed very briefly below. The cremendous range of density can be appreciated when one realizes that air and water differ in density by only 10°, while water and white dwarf sars are separated by only a factor of 10%. Even neutron stars are only 10" dimes denser than water, Vet gaseous plasmas in the entire density range of 10% can be described by the same set of ‘equations, since only the classical (non-quantum mechanical) laws of physies are needed, Gas Discharges (Gaseous Electronics) ‘The earliest work with plasmas was that of Langmuir, Tonks, and their collaborators in the 1920's. This research wat inspired by the need to develop vacuum tubes that could carry large currents, and therefore hhad to be filled with fonized gases. The research was done with weakly ionized giow discharges and positive columns typically with KT, = 2eV and 10* and KT, = 40 keV. Modern Astroph Stellar interiors and atmospheres are hot enough to be in the plasma state. The temperature atthe core ofthe sun, for instance, is estimated 0 be 2 keV; thermonuclear reactions occurring at this temperature are ‘esponsible for the sun’ radiation. The solar corona isa tenuous plasiia with temperatures up to 200 eV. The interstellar medium contains ion- ized hydrogen with n = 10° m”*. Various plasma theories have been used to explain the acceleration of cosmic rays, Although the stars ina galaxy are not charged, they behave like particles in a plasma; and plasma Kinetic theory has been used to predict the development of galaxies. Radio astronomy has uncovered numerous sources of radiation that ‘mos likely originate from plasmas. The Crab nebula i rich source of plasma phenomena because i is known to contain a magnetic field, It 2iso contains a visual pulsar. Current theories of pulsars picture them a rapidly rotating neutron stars with plasmas emitting synchrotron radiation from the surface MHD Energy Conversion and Ion Propulsion Geuting back down (0 earth, we come to two practical applications of plasma physics. Magnetohydrodynamic (MHD) energy conversion util jes a dense plasma jet propelled across a magnetic Beld to generate elecricity (Fig, 1-5). The Lorentz force gv x B, where vis the jet velocity, causes the ions to drift upward and the electrons downward, charging the to electrodes to different potentials. Electrical current can then be ‘drawn from the electrodes without the inefficiency of a heat cycle. “The same principle in reverse has been used to develop engines for interplanetary missions. In Fig. 1-6, currentis driven through a plasma by applying a voltage to the two electrodes. The j xB force shoots the plasma out ofthe rocket, and the ensuing reaction force accelerates the rocket. The plasma ejected must always be neutral otherwise, the space sip will charge to high potential Solid State Plasmas "The free electrons and holes in semiconductors constitute a plasma exhibiting the same sort of oscillations and instabilities as a gaseous plasma, Plasmas injected into InSb have been particularly useful in tens j -e8 Principle of the MHD generator Lr 176 FIGURE 1-5 Intraduton16 Chether One FIGURE 1-6 Lat majet engine for spacecraft propulsion. sudies of these phenomena. Because of the latce effects, che effective collision frequency is much less than one would expect in a solid with n= 10m", Furthermore, the holes in a semiconductor can have a very low effective mass—a little as 0.0m,—and therefore have high ‘yclotron frequencies even in moderate magnetic fields. If one were to ‘aleulate No for a solid state plasma, it would be lest than unity because of the low temperature and high density. Quantum mechanical effects {uncertainty principle), however, give the plasma an effective tem- perature high enough to make Np respectably large. Cersain liquids, such as solutions of sodium in ammonia, have been found to behave like plasmas also . Gas Lasers ‘The most common method to “pump” a gas laser—that is, to invert the population in the sacs that give rise to light amplifcation—is to use a 15 discharge. This can be a low-pressure glow discharge for a de laser for a highspressure avalanche discharge in a pulsed laser, The He-Ne lasers commonly used for alignment and surveying and the Ar and Kr lasers used in light shows are examples of de gas lasers. The powerful CO laser is finding commercial application asa cutting cool, Molecular lasers make possible studies of the hitherto inaccessible far infrared region of the electromagnetic spectrurm. These can be directly excited by an electrical discharge, asin the hydrogen cyanide (HCN) laser, of ‘an be optically pumped by a COs laser, a8 with the methyl fluoride (CHF) or methyl alcohol (CHsOH) lasers. Even solid state lasers, such as Nad-glass, depend on a plasma for their operation, since the fash {ubes used for pumping contain gas discharges, 1-8, In laser Fusion, the core of a stall pellet of DT i compressed toa density ‘of 10" m~ ata temperature of $0,000,000°K. Estimate she numberof patles ina Debye sphere in this plasma, 1-9, & distane galaay contains a cloud of protons and satiprotons, ech with Adensiy n = 10°m and temperature 100°. Whats the Debye length? 1.10, A spherical conductor of rads i immersed in a plasma and charged (0.8 potential de. The electrons remain Maswelian and move wo form a Debye shield, but the ions ae stationary during the time fame of the experimen. Assuming #e < KT¥/e, derive an expression for the potential 38a funtion of F interms of , dy, and Ay (Hint: Assume a solation of the form ¢-"/¢) TIL, A felsefece transistor (FET) is basically an electron valve that operates fn a fnte-Debyeslength elect. Conduetion electrons low from the soutee $10 the drain D through a semiconducting material when a potential applied between them. When a negative potentials applied tothe Insulated gate G. no currentcan low through G, but the applied potential esksimathe semiconductor and repels electrons. The channel width i narrowed and the electron Bow impeded in proportion 1 the gate potential I the thicknes ofthe devices oo large, Debye shielding prevents the gate voltage from penetrating far enough Estimate tie maximum thicknes of the conduction layer of an mchenel FET ite has doping level (plasma density) of 10" mi at room temperature. and isan berno more than 10 Debye lensth thick. (See Fig PI“), "7 Inrobucton FIGURE PIAL PROBLEMSChapter Two SINGLE-PARTICLE MOTIONS INTRODUCTION ‘What makes plasmas particularly difficult wo analyze is the fact that the eensities fall in an intermediate range. Fluids like water are so dense that the motions of individual molecules do not have to be considered. Collisions dominate, and che simple equations of ordinary fluid dynamics suffice, At the other extreme in very low-density devices like the alternating-gradient synchrotron, only single-particle trajectories need be considered; collective effects are often unimportant. Plasmas behave sometimes like Buids, and sometimes like a collection of individual particles. The first step in learning how to deal with this schizophrenic personality is to understand how single particles behave in electric and magnetic elds. This chapter differs from succeeding ones in that the E and B fields are assumed to be prescribed and not affected by the charged particles, UNIFORM E AND B FIELDS E=0 In this case, a charged particle has a simple cyclotron gyration. The ‘equation of motion is, tv mt = qvxB Ba ant teal Qa 22 2.21‘Taking & to be the direction of B (B= B2), we have mie aBo, mi, -(2)s 24 =-(2). This describes a simple harmonic oscillator at the elotron frequency, which we define i be alas od? 23) By the convention we have chosen, w is always nonnegative, B is measured in tesla, o webers/m®, a unit equal to 10° gauss, The solution of Eq. [2-2] is then Vey =U exp (ind + 15,5) the # denoting the sign of g. We may choose the phase 8 so that ame ae ea] where v4 is a positive constant denoting the speed in the plane perpen- dicular to B. Then ea Integrating once agin, we have xoxo ie o (25) Me dfn the Lmtd tobe nate re “Taing the veal part of 251, we have m= nnd en GUIDING @sB CENTER, 7 . ©. 1ON ELECTRON Larmor orbits ina magnetic field This describes a circular orbit a guiding center (xy, yq) which is fixed (Fig. 2-1). The direction of the gyration is always such that the magnetic field enerated by the charged particle is opposite to the externally imposed. eld. Plasma particles, therefore, tend to reduce the magnetic field, and plasmas are diamagnetic. In addition to this motion, there is an arbitrary velocity v, along B which is not affected by B. The trajectory of a charged particle in space is, in general, a helix: Finite E If now we allow an electric field to be present, the motion will be found to be the sum of two motions: the usual circular Larmor gyration plus a drift of the guiding center. We may choose E to lie in the x- plane so that E, = 0. As before, the = component of velocity is unrelated to the transverse components and can be treated separately. The equation of dv a E+ VB) es) whose z component is Petree ea) 21 Single-Paricle FIGURE 2-1 a 22.2 “Motions‘This is a straightforward acceleration along B. The transverse com- ponents of Eq, (2-8] are ta fp 200 Wel, or | s=ra(fe aun) arr) so that Eq. [2-11] is reduced to the previous case if we replace v, by ¥, + (Ex/B). Equation [2-4] is therefore replaced by | us ean jo ete Et ‘The Larmor motion is the same as before, but there is superimposed a drift vg of the guiding center in the ~y direction (for E, > 0) (Fig. 2-2), ION ELECTRON | RE22 Particle drifts in crossed electric and magnetic fields. To obtain a general formula for ve, we can solve Eq. [2-8] in vector form. We may omit the mdv/dt term in Eq. [2-8], since this term gives only the circular mot th we already know about. Then Eq. [2-8] becomes 1 at ay EtvxB 3) ‘Taking the cross product with B, we have EX B= BX (vB) = vB*~ Biy-B) pan ‘The transverse components of this equation are vig EX B/B® 2151 We define this to be ve, the electric field drift of the guiding center. In magnitude, this drift is, E(W/m) m Blesla) sec Le I is important co note that ve is independent of g, m, and v,, The reason is obvious from the following physical picture. In the first half- cycle of the ion’s orbit in Fig. 2-2, it gains energy from the electric field and increases in vs and, hence, in r1, In the second half-cyele, it loses energy and decreases in rj. This difference in rom the left and right sides of the orbit causes the drift v-. A negative electron gyrates in the opposite direction but also gains energy in the opposite direction; itends up drifting in the same direction as an ion, For particles of the same velocity but different mass, the lighter one will have smaller rand hence drift less per cycle. However, its gyration frequency is also larger, and the two effects exactly cancel. Two particles of the same mass but different energy would have the same w,. The slower one will have smaller 7, and hence gain less energy from E in a half-cycle. However, for less energetic particles the fractional change in rz for a given change in energy is larger, and these two effects cancel (Problem 2-4). ‘The three-dimensional orbit in space is therefore a slanted helix with changing pitch (Fig. 2-8). Gravitational Field ‘The foregoing result can be applied to other forces by replacing gE in the equation of motion [2-8] by a general force F. The guiding center 23, Single Particle Motions0 25 2 : Chapter Ee ‘Single-Pariicl a 1 ins Cyn? FIGURE2-3 The actual orbit ofa gyrating particle in space, drift caused by Fis then yt EB fe aR ea In particular, if Fis the force of gravity mg, there isa drift ve BEER (2-18 one ts) ‘This is similar to the drift vein that iis perpendicular to both the force and B, but it differs in one important respect. The drift ve changes sign with the particle's charge. Under a gravitational force, ions and electrons rift in opposite directions, so there isa net current density inthe plasma given by xB = (e+ my EE f janie mem 19) ‘The physical reason for ths drift (Fig. 2-4) is again the change in Larmor radius as the particle gains and loses energy in the gravitational field, Now the electrons gyrate in the opposite sense to the ions, but the force ‘on them isin the same direction, so the drifcis in the opposite direction. ‘The magnitude of v, is usually negligible (Problem 2-6), but when the lines of force are curved, here is an effective gravitational Force due to Aenaananancnanazane LQ @s “The drift of «gyrating particle in a gravitational ld, centrifugal force. This force, which is not negligible, is independent of mass; this is why we did not stress the m dependence of Eq. (2-18). Centrifugal force isthe basis of a plasma instability called the "gravita- tional” instability, which has nothing to do with real gravity. 2:1, Compute r forthe following cats if is aegigibe (@) A 10-keV electron inthe earths magnetic field of > 10". (©)-A solar wind proton with steaming velocity 300 km/sec, B= 5% 10°*T, (0A keV He" ion in the solar stmorphere near a sunspot, where B = Bx 10", (@) AS5MeV He th particle in an 8-1 DT fusion reactor 2.2, In the TFTR (Tohamak Fusion Test Reactor at Princeton, the pata wil be heated by injection of 200-keV neutral deuterium atoms, which, ater entering ‘he magnetic el, are converted t0200-LeV D fons (A = 2) by charge exchante “Thete ions are confined only ifr. «a, where @ = 0.6m isthe miner radius of| the toroidal plasma, Compute the enarimum Larmor rave ina 5-1 field to see ita cai, 23. An jon engine (fee Fig. 1-6) has & I-T magnetic field, and a hydrogen plasma isto be shot out at an EB velocey of 1000 km/sec. iow much intreal ‘ected must be present in the pla? 2-4, Show tha i the same fortwo fons of equal mass and charge but different ‘eergics, by using the folowing physical picture (sce Fig. 2-2). Approximate the Fight af of the orbit by 2 vemicrle corresponding to the ion energy ater scceeration by the E fel, and the left hall by a semicitce corresponding to the energy after deceleration. You may assume that E is weak, s0 thatthe fractional change inv, is smal FIGURE 24 PROBLEMS2.5. Suppose electrons obey the Boltzmann relation of Problem 1-inaeylindi cally symmetsic plasma column in which m() varies with a scale length A that is, dnfoe = =n (@) Using E = ~04, find the radial electric fel forgiven 8 (b) For electrons, show chat finite Larmor radi effects ae largeif vei large a8 ey, Specifically show that r= 24 op = (6) 15) aso eve for ions? Hine Do nl se Poisson's equation 2.6. Suppose that a soxalled Qumachine has a uniform field of 0.27 and a crlindrcal plasma with KT, = KT, = 0.2 eV. The density profile ifound experi ‘mentally tobe of Une form xp exp (-r7/2")— 1) Assume the density obeys the electron Boltzmann relation n= nyexp e6/KT®), (0) Calculate the susnimur vg if @ = Lm, (6) Compare this wth », due tothe earths gravitational fle (©) Towhat value can B be lowered before the ions of potassium (A = 99,2 = 1) hhavea Larmor radius equal a? 2:7. An unneviralize electron beam has density n,~ 10" mand radias @ = Jem and flows along a 2T magnetic Feld. If Bis in the +2 direcion and Bi the electrostatic field due to the beam’s charge, calculate the magnitude and direction of the EX B drift at r =a. (See Fig. P27) nr — Be =< NONUNIFORM B FIELD Now that the concept of a guiding center drift is firmly established, we ‘can discuss the motion of particles in inhomogeneous fields and B Fields which vary in space or time. For uniform elds we were able 10 ‘obtain exact expressions for the guiding center drifts. As soon as we introduce inhomogeneity, the problem becomes too complicated to solve he maar y ® cooee | Qa 600.0 2 90,02 ger 1 5 e000 4% ‘The deft ofa eye ng particle in a nonuniform magnetic Bld exactly. To get an approximate answer, itis customary to expand in the small ratio ri/L, where Lis the scale length of the inhomogeneity. This type of theory, called orbit theory, can become extremely involved. We shall examine only the simples cases, where only one inhomogeneity ‘VB 1B: Grad-b Dri Here the lines of force® are straight, but their density increases, say, in the y direction (Fig 25) We can anticipate the resultby using our simple physical picture. The gradient in |B] causes the Larmor radius to be larger at the bottom of the orbit than at the top, and this should lead to a drift, in opposite directions for ions and electrons, perpendicular to both B and VB. The drift velocity should obviously be proportional ton/L and tov, Consider the Lorente force F= qv% B, averaged over a gyration, Clearly, = O,since the particle spends as much time moving upasdown. ‘We wish tocalculate Finan approximate fashion, by using the undisturbed orbit of the particle to find the average. ‘The undisturbed orbit is given by Eqs. (2-4) and [2-7] for a uniform B Feld, Taking the real part of Eq, [2-4], we have where we have made a Taylor expansion of B field about the point 0" 0, 30 Oand have used Eq. (2-7: Bo +(e WB + By = By 3@B,/a9) + ren cf Hines ae often cle “nes of force” They are ot Hines of Fore “The misnomer i prpciaated hereto prepare the sadent forthe (eeacheres oF hs28 Ghaper Teo 282 ‘This expansion of course requires ri/ « 1, where L is the scale length of aB,/2. The first term of Eq. (2-20) averages to zero in a gyration, and the average of cos" aut is, x0 that = Fqours@B/0y) rea} ‘The guiding er drift velocity is then. oR Dat peas where we have used Eq, [2-17]. Since the choice of they axis was arbitrary, this can be generalized to (24) ‘This has all the dependences we expected from the physical picture; only the factor # (arising from the averaging) was not predicted. Note that the * stands for the sign of the charge, and lightface B stands for 1. The quantity ven is called the grad-B drift; itis in opposite directions for ions and electrons and causes a current transverse (0 B. An exact ‘orbit, induding the ‘aleulation of vee would require using the exa drift, in the averaging process, Curved B: Curvature Drift Here we assume the lines of force to be curved with a constant radius fof curvature Ry, and we take [| to be constant (Fig. 2-6). Such a field does not obey Maxwell's equations in a vacuum, soin practice the grad-B drift will always be added to the effec derived here. A guiding center drift arises from the centrifugal force felt by the particles as they move along the feld lines in their thermal motion. If uj denotes the average square of the component of random velocity along B, the average centrifugal force is 28) A curved magneti fel According to Eq, (2-17), this gives rise toa drift (225) The drift vx is called the curvature drift. We must now compute the grad-B drift which accompanies this when the decrease of || with radius is taken ino account, In a vacuum, wwe have V xB = 0. In the cylindrical coordinates of Fig. 2-6, ¥ xB has only a: component, since B has only a 0 component and WB only an r component. We then have (xB), 27 Thus Lo valk [Bix BRE eas Using Bq. [2-24], we have Luiteg y von FS EB eas) 29 Single-Potcle ‘Motions FIGURE 2-6‘Adding this to vp, we have the total drift in a curved vacuum field (230) tis unfortunate that these drifts add. This means tha if one bends a ‘magnetic field into a torus for the purpose of confining a thermonuclear plasma, the particles will drift out ofthe torus na matter how one juggles the temperatures and magnetic fields For a Maxwellian distribution, Eqs. 17] and (1-10) ind ‘Uf and ful are each equal to KT/m, since vs involves two degrees of freedom, Equations {2-3] and [1-6] then allow us to write the average ccurved/field drift as te that Ro Re where j here is the direction of Rx B. This shows that ¥a+eu depends fon the charge ofthe species but not on its mass ong (2.308) ‘VBIB: Magnetic Mirrors Now we consider magnetic Feld which ie pointed primarily in the 2 direction and whose magnitude varies in the x direction. Let the field be axisymmetric, with By = 0 and 3/90 = 0. Since the lines of force converge and diverge, there is necessarily a component B, (Fig. 2-7)-We wish to show that this gives rise to a force which can trap a particle in a magnetic field Daift ofa particle ia a magnetic misror field We can obtain 8, from ¥ B= 0: La aes wan If aBy/a is given at approximately 0 and does not vary much with r, we have es) fe ‘The variation of |B] with r causes a grad-B dbife of guiding centers about the axis of symmetry, but there is no radial grad-B drift, because 8/30 = 0. The components of the Lorentz force are F, = a(veB.~ vB) o F av, +0B,) 33) e' 8 = aloe UB) ‘0 ‘Two terms vanish if By = 0, and terms I and 2 give rise to the usual Larmor gyration. Term 8 vanishes on the axis; when it does not vanish, this azimuthal force causes 2 drift in the radial direction. This drift merely maker the guiding centers follow the lines of force, Tera 4 is the one we are interested in. Using Eq, [2-52], we obtain Sqver(0B./22) es We must now average over one gyration. For simplicity, consider a particle whose guiding center lies on the axis. Then vp is a constant during a gyration; depending on the sign of q, vs is v,. Since =r, the average force is teas BF =F 351 ‘We define the magnetic moment of the gyrating particle to be = bmol/B 2.36)32 Ghaper Tw sothat u(@B./22) est) ‘This isa specific example of the force on a diamagnetic particle, which in general can be writen Ri 1 3B/0s=—u YB (2.38) where ds saline element along B. Note that the definition (2-36) is the same as the usual definition for the magnetic moment of a current loop with area A and curvent F: 4 = IA. In the case ofa singly charged ion, Tis generated by a charge ¢ coming around w,/2a times a second: 1 = ean. The area A is rr. = avi /o?, Thus [As the particle moves into regions of stronger or weaker B, its Larmor radius changes, but remaine invariant. To prove this, consider the component of the equation of motion along B: det. 8 Ben? te-a9) Multiplying by 2 on the left and its equivalent ds/dt on the right, we have slen)- Here dB/dt is the variation of 8 a5 scen by the particle; B itsll is ‘constant, The particle's energy must be conserved, so we have 8 7 (240) Comat st) =£(boof ent) <0 a neko) =o sos swe naa “The invariance of sis the basis for one ofthe primary schemes for plarma confinement: the magnetic mirror. As a particle moves from a weaksfield region to a strong-feld region in the course of its thermal motion, it sees an increasing B, and therefore its vs must increase in order to keep 4 constant. Since its total energy must remain constant, ‘must necessarily decrease. If B is high enough in the “throat” of the mirror, ay eventually becomes zero: and the particle is “reflected” back to the weak-field region. Its, of course, the force F; which causes the reflection. The nonuniform field of a simple pair of colle forms two magnetic mirrors between which a plasina can be trapped (Fig. 28). This effect works on both ions and electrons. The rapping is not perfect, however. For instance, a particle with 19, =O will have no magnetic moment and will not feel any force along B. A particle with stall v,/0j at the midplane (8 = B) will also escape if the maximum field Bq is not large enough, For given Bo and Bm, which particles will escape? A particle with 9, = #9 and = uj at the midplane will have v2 = vi and vj 0 a its turning point. Let the field be B' there. Then the invariance of yields ev fo/Bo = Sr’? /B es) Conservation of energy requires 4a) Combining Eqs. [2-43] and [2-44], we find Bete B ets) ‘where @ is the pitch angle of the orbit in the weak-feld region. Particles with smaller @ will mirror in regions of higher B. If @ is too small, B’ exceeds Bui and the particle does not mirror at al, Replacing B' by Bu in Eq, (2-45), we see that the smallest 6 of a confined particle is given by sin 0 = Bo/ B® 1/R 45) 33 Single-Porile FIGURE 2834 Ghaper Te FIGURE29 The ots cone, PROBLEMS where Rr is the mirror ratio. Equation [2-46] defines the boundary of 2 region in velocity space in the shape of a cone, called a loss cone (Fig, 2.0) Particles lying within the loss cone are not confined. Consequently 42 mirror-confined plasma is never isotropic, Note that the loss cone is independent of q or m. Without collisions, both ions and electrons are ccqually well confined. When collisions occur, particles are lost when they change their pitch angle in a collision and are scattered into the loss cone. Generally, electronsare lost more easily because they have a higher collision frequency. ‘The magnetic mirror was first proposed by Enrico Fermi as a mechanism for the acceleration of cosmic ray, Protons bouncing between ‘magnetic mirrors approaching each other at high velocity could gain ‘energy at each bounce. How such mirrors could arise is another story A further example of the mirror effect is the confinement of particles in the Van Allen belts. The magnetic field of the earth, being strong at the poles and weak at the equator, forms a natural micror with rather large Rn 2-8. Suppose the earth's magnetic eld is 3 10" at the equator and falls of| 45 1/7, as fora perfect dipole, Let there be an isotopic population of 1-eV protons and 80-keV electrons, each with density = 10" a In the equatorial plane earth rad (@) Compute the ion al electron VB drift velocities. (6) Docs an electron def eastward or westward? (e) How long does i take an electron to encircle the earth? (@) Compute the ring current densi in A fet Note: The curvature dri is not negligible and wil fleet the numerical answer Dut neplet it anya. 129. An electron ties at retin the magnetic Reld of an iafinite straight wire Carryinga current L ACL = 0 the wire is suddealy charged to postive potential without afecing 1 The electron guins energy from the electric Bld and bonins to die (a) Draw a diagram showing the orbit ofthe electron and the relative directions OF T,B, ve Wor and vp (©) Calculate the magnitudes of these drift ata radius of Lom IE = 500, 14 ~ 460 V, and the radiue ofthe wire i Lm. Assume that ¢ is held 2¢0V on ‘he vacuum chamber walle 1em aay Hint: A good inv pictare of the motion speeded ia addition tothe formulas sen inthe vex 2-10. A 20-LeV deuteron ina large mirror fusion device has a pitch angle 8 of ‘45a: che midplane, where B= 07. Compute its Larmor radius. 2-11, A plasma with anisotropic velocity distribution is placed in a magnetic tnior rap with micro ratio Ry ~ 4. There are no collisions, so the particles in the loss cone simply escape, andthe rest remain-trapped. What fraction is wrapped? 2:12. A coum ray proton ie trapped between wo moving magnetic mirrors with 8, — 5 and inaly has W = | keV and, = vjatthe midplane. Each miteor Troves toward the midplane witha velocity a= 10 kn/se (ig 2-10), Acceleration of cori rays. 35 Single Particle FIGURE 2-1036 Chaper Too 24 FIGURE 2-11 (2) Using the loss cone formula and the invariance of find the eneray to ‘which the proton will be accelerated before it escapes. (©) How fong willie wo reach that energy? 1. Treat the mirrors as flat pistons and show that the velocity gained at each bounce is 2 Compute the number of bounces necesary 3. Compute the time Tic takes traverse L that many times. Factor-f-wo accuracy wil sufi. NONUNIFORM E FIELD Now we let the magnetic field be uniform and the electric field be nonuniform. For simplicity, we assume E to bein the x direction and to vary sinusoidally in the x direction (Fig. 2-11) Ee Exfcoseyt ean ‘This fcld distribution has a wavelength A» 2n/k and isthe result of a sinusoidal distribution of charges, which we need not specify In practice, such a charge distribution can arise ina plasma during 2 wave motion, The equation of motion is (deat) = (Efe) + ¥ xB] as @s Drift of «gyrating partite in nonuniform electri fel whose transverse components are we i Bartel oo é ~wla 2 feat ea = wits, ~ ot (2-51 5, -wl, “wt om Here Es(+) is the electvic Field atthe positon of the particle. To evaluate this, we need to know the partide’s orbit, which we are trying to solve for in the first place. If the electric Field is weak, we may, asan approxima- tion, use the undisturbed erbit to evaluate E,(x). The orbit in the absence of the £ Feld was given in Eq. [2-7 xexotrsinad west From Eqs, [2-51] and [247], we now have oEo mole, ~ 0222 cosktsa tr.sin wa) ess, [Anticipating the result, we look for a solution which is the sum of a fyration at @, and a steady drift ve, Since we are interested in finding, an expression for ve, we take out the gyratory motion by averaging over a eyde, Equation (2-50] then gives ¥, = 0. In Eq, [2-53], the oscillating term 3, dearly averages to zero, and we have = 0- wi, ~ Ba) Expanding the cosine, we have cos Klxa +r. sin at) = c08 (bx) os (resin end) sin (lze)sin (irysinae) 58) Te will ulfce to treat the small Larmor radius case, fr, « 1. The Taylor (2561 37 ‘Single-Portle Maton38 Ghoper fe allow us to write cos (to + FL sin at) = (Cos bra) — HAF sin?) = Gin kebr. sine The last term vanis 1s upon averaging over time, and Eq, (2-54) gives =~ FXtoste (1H) = Bt as ‘ ‘Thus the usual E x B drife is modified by the inhomogeneity to read. we A(I jer) e581 TThe physical reason for this is easy to see, An ion with its guiding. center at a maximum of E actually spends 2 good deal of its time in regions of weaker E. Its average drift, therefore i less than £/ evalu ated at the guiding center. In a linearly varying E field, the ion would be in a stronger feld on one side of the orbit and in a field weaker by the same amount on the other side; the correction to ve then cancels ‘out. From this i is clear that the correction term depends on the second derivative of B. For the siausoidal distibution we assumed, the second derivative is always negative with respect to F. For an arbitrary variation Of E, we need only replace sk by V and write Eq. [2-58] as (vstaw x? wn ‘The second term is called the fnite-Larmor-radiue effect, What isthe significance of this correction? Since ris much larger for ions than for electrons, vy is no longer independent of species. Ifa density clump ‘occurs in a plasma, an electric field can cause the ions and electrons to separate, generating another electric fcld If there isa feedback mechan: ism that causes the second electric feld to enhance the first one, E grows indefinitely, and the plasma is unstable. Such an instability, called a drift instability, wil be discussed ina later chapter. The grad-B deift, of course, is also a fnite-Larmor-radius effect and also causes charges to separate. According to Eq. [2-24], however, wea is proportional to ir, whereas the correction term in Eq. (2-58) is proportional to Ari. The nonuni form-E-feld effec, therefore, i important at relatively large 4, oF small scale lengths of the inhomogeneity. For this reason, drift instabilities belong to.2 more general clas called microinsabilitis ‘TIME-VARYING E FIELD Let us now take E and B to be uniform in space but varying in time. First, consider the casein which E alone varies sinusoidally in time, and let tlie long the x axis: Foes 2-60) Since B, = iw, we can write Bq, [2-50] as a> -ol( eo Let us define where the tilde has been added merely to emphasize that the drift is oscillating. The upper lower) sign, as usual, denotes positive (negative) 4. Now Eqs. [2-50] and (2-51) become en -wiu-H) (2-63) = -w2(e, Ge) By analogy with Eq, (2-12), we try a solution which is the sum of a drift and a gyratory mation: 04) If we now differentiate twice with respect to time, we find B= -wlee + (03 -075 re 12.65) B--wh, +0 oie This is not the same as Faq. [2.63] unless o* < w!. If we now make the assumption that E varies slowly, x0 that w* « 2, then Eq. (2-04) is the approximate solution to Eq. (2-63) 39 Single Portle Motions40 Ghaper FIGURE 2-12 Bquation [2-64] tell us that the guiding center motion has two components. The y component, perpendicular to B and E, is the usual EX B drift, except that ve now oscillates slowly atthe frequency w. The x component, a new drift along the direction of E, i called the polarization drift. By replacing iw by 8/3, we can generalize Eq [2-62] and define the polarization drift as 66 Since v is in opposite directions for ions and electrons, there is a polarization current; for Z = 1, this is ne dep dE ip = nelug — 0) = BM + mE = By pon where p isthe mass density ‘The physical reason for the polarization currentis simple (Fig. 2-12). Consider an ion at rest in a magnetic field. Ifa field K is suddenly applied, the frst ching the ion does is wo move in the direction of E. Only atter picking up a velocity vdoes the fon feela Lorena force ev xB and begin to move downward in Fig, (2-12). ITB is now kept constant, there is no further vp drift but only a ve drift. However, if Eis reversed there is again a momentary drift, this ime othe left. Thus vp isa startup drift due to inertia and occurs only in the first half-0 for electrons. Then Eq. (270] becomes 26 41 Single-Potle Motions42 Ghoper Two FIGURE 2.13 > ‘Two-stage adiabatic compression of plasma, ‘The quantity 2x8/a, = B/f. i jus the change 8B during one period of syration, Thus bcdmo2) = 5B em Since the lefwhand side is 5 (12), we have the desired result fu=0 273) ‘The magnetic moment is invariant in slowly varying magnetic felis AAs the B field varies in strength, the Larmor orbits expand and contract, and the particles lose and gain transverse energy. This exchange of energy between the particles and the field is described very simply by Eq, [2-73]. The invariance of 1 allows us to prove easily the following well-known theorem ‘The magnetic flux through a Larmor orbit is constant, The flux @ is given by BS, with $= ari. Thus = Br d= Be oe eB re 7 Therefore, © is constant if wis constant ‘This property is used in a method of plasma heating known 35 adiabatic compression, Figure 2-13 shows a schematic of how this is done, A plasma is injected into the region between the mirrors A and B. Coils A and B are then pulsed toincrease B and hence v}. The heated plasma can then be transferred to the region C-D by a further pulse in A, increasing the mirror ratio there. The coils G and D are then pulsed to further compress and heat the plasina. Early magnetic mirror fusion devices employed this type of heating. Adiabatic compression has also been used successfully on toroidal plasmas and is an essential element of laser-triven fusion schemes using either magnetic or inertial confinement. SUMMARY OF GUIDING CENTER DRIFTS 2.7 LPxB eneral force wat ES eam Ganeral force F ar ean EXB lectrie fl eo 15 Bcc fel a rs) Gravitational eld Nonusiform B: 2-50) Nonuniform B feld 1, BxvB Grade dif won =e ton BAY eas Curvature ari 220) Curved vaca fl 2-90) Polarisation dif 12-6 ADIABATIC INVARIANTS is well known in clasical mechanics that whenever a system has a periodic mation, the action integral # da taken overa period isa constant fof the motion, Here and are the generalized momentum and coordin- tte which repeat themselves in the motion, If slow ehange is made in the system, so that the motion is not quite periodic, the constant of the ‘motion does not change and is then called an adiabatic invariant By slow here we mean slow compared with the period of motion, so that the integral {dq is well defined even though it is strictly no longer an 28 8 Single Porte Motions44 Choper Too 2.81 integral over a closed path. Adiabatic invariants play an important role in plasma physics; they allow us to obtain simple answers in many instances involving complicated motions. There are three adiabatic invariants, each corresponding toa different type of periodic motion, "The First Adiabatic Invariant, We have already met the quantity w= mol (28 daemon mean become fod fsride torn <2 tae ga ‘Thus 1 a constant of the motion as long as g/m is not changed. We have proved the invariance of only with the implicit assumption fea, © 1, where w is frequency characterizing the rate of change of B as seen by the particle. A proof exists, however, that wis invariant even when @ Sy In theorist’ language, 4 is invariant “to all orders in an expansion in w/o.” What this means in practice is that. pains much ‘more nearly constant than B does during one period of gyration. {is just as important to know when an adiabatic invariant does not exist as to know when it does. Adiabatic invariance of u is violated when ‘isnot small compared with w,. We give three examples ofthis, (A) Magnetic Pumping. If the strength of B in a mirror confinement system is varied sinusoidally, the particles’ v, would oscillate; but there would be no gain of energy in the long run. However, if the particles make collisions, the invariance of jis violated, and the plasma can be hheated. In particular, a particle making a collision during the compres- sion phase can transfer part of its gyration energy into % energy, and this s not taken out again in the expansion phase. (8) Cyelotron Heating. Now imagine thatthe B felis oscillated at the frequency «,. The induced electric Feld will shen rotate in phase with some of the particles and accelerate their Larmor motion continuously. The condition w « a, is violated wis not conserved, and the plasma can be heated, Jf mano ORDINARY MIRROR AXIS OF SYMMETRY a Plasma confinement ina cusped magnetic fil Oe a es aaa su somes Pig 218). This aan ee gin totus tors aspindlecsp aston aa adhe pre motions nonadiabte Since he 3 eld rai mor rn ear the ete rae han aa orate veal angular somentam fatty ean Thi Stil they make sali. ‘The Second Adiabatic Invariant, J Fa athe guidngcente) slong ed ne. However ve ota tenon becomes =n ain Fea una nan Jas dete Fora halle 45 Singl-Pertle FIGURE 244 282 Mion:46 Chapter Two FIGURE 2-15. FIGURE 2.16 Motion of «charged particle in the earth's magn A particle bouacing between turning points land in a magnet Geld between the two turning points (Fig 76) We shall prove that J is invariant in a static, nonuniform B field; the result is also true for a slowly time-varying B field Before embarking on this somewhat lengthy proof, let us consider an example ofthe type of problem in which a theorem on the invariance of J would be useful. As we have already seen, the earth's magnetic feld mirror-raps charged particles, which slowly drift in longitude around the earth (Problem 2-8; see Fig, 2-16). 1f the magnetic fld were perfectly symmetric, the particle would eventually drift back to the same line of force. However, the actual field is distorted by such effects as the solar wind. In that case, will a particle ever come back to the same line of force? Since the particle's energy is conserved and is equal to imu at the turning point, the invariance of indicates that || remains the same at the turning point. However, upon drifting back to the same fed, longitude, a particle may find itself on another line of force ata different altivude. This cannot happen if J is conserved. J determines the length of the line of force between twrning points, and no two lines have the same length between points with the same| B]. Consequently, the particle returns tothe same line of force even ina slightly arymmettc Bld ‘To prove the invariance of J, we first consider the invariance of v8, where 6s is a segment of che path along B (Fig. 2-17). Because of suiding center drifts, a particle on 5 will find itself on another line of force 51" aftera time Bt. The length of it defined by passing planes perpendicular to B through the end points of 8s. The length of & is ‘obviously proportional to the radius of curvature 7 ROR sothat Bn 6 RIOR am Bras, “AIR, : The “radial” component of wis vst mn From Eqs. (2-24) and [2.26), we have yy BAVB, mR ven we tues br EAE “The las term has no component along R.. Using Eqs. [2-78] and we can write Eq. (2-77) as 1d R, e ims R, asa RT 2g BY (BX VB) Fi (2-80, ‘This is the rate of change of 6: as seen by the particle. We must now set the rate of change of u 38 seen by the particle. The parallel and 47 ‘Single Potle Motions Proof of the invariance of J. FIGURE 2.17