Arxiv:Physics/0606109V1

arXiv:physics/0606109v1 [physics.gen-ph] 13 Jun 2006 quasistatic hsgpi oreo h ee ffreapeGriffith’s example for of fill level can the the that on material and course textbook. present a textbooks to in most gap is in this paper neglected this less of or purpose more is ics plctoso lcrdnmc a ei the in be may electrodynamics of Applications .Tese rmsaist eea lcrdnmc is electrodynamics general to statics from step The 1. .Mn elwrdapiain nov h numeri- the involve applications world real Many 3. within phenomena interesting of plenty are There 2. easn lk usd ogslni rtoroidal current). or time-varying solenoid a long to with a seems outside coil field (like magnetic absent electric the be dynamic where of regions in appearance addressed. fields the be also should should So situations non-static laws Biot-Savart’s in Ques- and deriva- Coulomb’s avoided. concerning standard be tions may condensed equations Maxwell’s very of tion the that from issues confusing result several levels Thereby intermediate Quasistatic theory. providing the content. in by physical useful its are of models terms in huge tion currents in and waves charges should why free But in space. waves and electromagnetic remarkable particular is in term includes this of conservation. impact elec- charge physical of The equations with the consistent make trodynamics to is order it in where introduced current displacement textbook the standard for the motivation from follows issue confusing hs n ayohrquestions. other many and these naayi a edffiutadi o eesr in necessary not is and Such situations. difficult quasistatic be frequency. may high analysis or an of time-scales short set involving phenomena the full wave describe electromagnetic intricate can the most equations these unnecessarily Using since times complicated many are equations regime. Maxwell’s quasistatic the 1 ? 10 hr r odraost os including: so do to reasons good are There rgeneral or o h motneo ussais hsatvt a lom also may activity This electrodynamics. basic quasistatics. of of presentation simulat importance poss electromagnetic of the is use today for The that activity potential. riche an pedagogical t far modeling, similar applications, electromagnetic more Max in important ways of and some set physics in full interesting and i the electrodynamics theory and general Quasistatic c statics than without between Journal. not in is this level it intermediate to that contributions however many reveals by look electro general closer to A statics from simple. path usual The trodynamics. ussaisi sfli h icsinof discussion the in useful is Quasistatics ussaisi nrdcds hti t mohyit h s the into smoothly fits it that so introduced is Quasistatics .INTRODUCTION I. ihfrequency high lcrmgeisfo ussai perspective quasistatic a from Electromagnetics eamn fPyis m˚ nvriy E917Ume˚a, Sw SE-90187 Ume˚a University, Physics, of Deparment 11 olwfrom follow Rcie etae eie etae cetdtextdate) Accepted textdate; Revised textdate; (Received eie Quasistat- regime. resaewithout space free hreconserva- charge 5,6,7,8,9 Another static oa Larsson Jonas 2,3,4 , nesadn fbscter u ti loimportant also is it but theory basic of understanding h l usin# fti ora hc ocrsthe current. time-varying concerns a which with on Journal solenoid comments this a contains of outside V #6 fields Section question of old validity given. the exact are the laws for these conditions necessary and ficient a.I eto Vw osdrtelw fCuoband Coulomb of theory. laws the time-dependent consider general Jour- we within this Biot-Savart IV to Section contributions In many nal. in discussed been have from III Section perspective. in quasistatic discussed a are derivation equations textbook Maxwell’s standard the- of the Poynting to quasistatic Alternatives the and orems. from charge fields point the moving quasistatic potentials, a include electromagnetic F the for and theory E D, Subsections quasistatics. ina itne nsbeto h corresponding the C subsection In interac- terms distance. instantaneous in c a is other at quasistatics the tion of and A laws feature equations. differential force salient Maxwell’s of in quasistat- approximations terms of of in formulations one equivalent two ics, and A the Subsections contain II. B Section in presented is quasistatics ∞ → ussaisi sflntol o rvdn improved providing for only not useful is Quasistatics etosI n ocr oecnuigise that issues confusing some concern V and IV Sections hsppri raie sflos ai hoyof theory Basic follows. as organized is paper This .Bscqaittcter a eitoue na el- an in introduced be may theory quasistatic Basic 4. hnsais ussaisi uhused much is Quasistatics statics. than r uaini ntrso ipeapoiain in approximations simple for- of equations. second differential terms Maxwell’s The in includ- static is laws, laws. the mulation force Biot-Savart of of and formula- terms Coulomb generalizations in equivalent straightforward is two first ing qua- given The also be theory tions. static may Like sistatics courses basic electrodynamics. for in suitable way simple and ementary eco- more in (or solved elliptic be an ways. can with nomic which model one hyperbolic replace- parabolic) the an to of amounts ment This quasistatic sufficient. some is where analysis phenomena one in often interested quite equa- However, is Maxwell dimensions. three time-dependent in the tions of solution cal osi eciggvsadtoa support additional gives teaching in ions yaisi ahrsotadsurprisingly and short rather is dynamics ii fgnrleetoyaisi sdt derive to used is electrodynamics general of limit beo Cadwihas a great has also which and PC a on ible elseutos ussaisi easier is Quasistatics equations. well’s ocpulyueu ypoiigan providing by useful conceptually s adr etokpeetto felec- of presentation textbook tandard nuigise shsbe illustrated been has as issues onfusing tvt oecag ffcsi the in focus of change some otivate ttc.I shwvr ntrsof terms in however, is It statics. o 12,13,14 eden 2 Suf- 5 2 for applications, in particular for numerical simulations. The three equations (1), (2) and (4) constitute the EQS In Section VI we discuss the possibility of using elec- model that includes capacitive but not inductive effects. tromagnetic simulations in basic courses and exemplify Charge may be accumulated in this model but this re- with quasistatic equations for eddy currents. A couple of quire work and energy may as usual be associated with PDE-solvers which have been used for electromagnetic the electric field. Magnetic energy is however outside the modeling on a PC is mentioned. A summary is given in scope of EQS because there is no magnetically induced Section VII. electric field and accordingly no back emf. No work is then required to create a magnetic field by starting an electric current. II. QUASISTATICS We now like to include electromagnetic induction. Then also ∂B/∂t acts as a source of electric fields and Quasistatics within electrodynamics refers to a regime the total electric field becomes the sum of two parts where ”the system is small compared with the electromagnetic wavelength associated with the dominant time- E = EC +EF (5) scale of the problem”.15 The fields are then propagated instantaneously so we are dealing with a kind of c The first term is the Coulomb electric field limit. We consider in this paper three major quasistatic→ ∞ 11 models. These are EQS (electroquasistatics) , MQS r′ Rˆ 11 12,13,14 1 ρ ( ,t) ′ (magnetoquasistatics) and the Darwin model . EC (r,t)= dτ (6) 4πε R2 EQS includes capacitive but not inductive effects, MQS 0 ZZZ includes inductive but not capacitive effects while the and the second term is the Faraday electric field which Darwin model includes both capacitive and inductive ef- may be defined by a Biot-Savart like integral expression fects. The Biot-Savart law is valid in all three models as (cf. equation 2)19 while the Coulomb law is valid only in EQS. In MQS ′ and Darwin there is also, besides the Coulomb field, an 1 ∂B (r ,t) Rˆ ′ EF (r,t)= dτ (7) additional contribution to the electric field due to mag- −4π ZZZ ∂t × R2 netic induction. The source of this electric field is thus 16 ∂B/∂t in Faraday’s law. We note that EF solves the equations ∂B ▽ E = , ▽ E = 0 (8) A. From the laws of Coulomb, Biot-Savart and × F − ∂t · F Faraday to the EQS, MQS and Darwin models. while the corresponding equations for EC is Dynamical systems that proceed from one state to an- ρ ▽ EC =0, ▽ EC = (9) other as though they are static (at each fixed time) are × · ε0 commonly said to be quasistatic. For electromagnetism the static theory builds on the Coulomb and Biot-Savart The expressions (6) and (7) are the unique solutions of laws together with the static continuity equation. Qua- (8) and (9) provided appropriate boundary conditions at 20 sistatics would then simply be obtained by the allowance infinity are used. for time-dependence in the first two otherwise unchanged The Darwin model may be defined by the equations laws (2) and (4)-(7). This is a quasistatic model that includes both capacitive and inductive phenomena. Note that for given current and charge densities we directly obtain the 1 ρ (r′,t) Rˆ E r ′ electromagnetic fields in terms of integrals without the ( ,t)= 2 dτ (1) 4πε0 ZZZ R appearance of any time-retardation. Instead the integral expressions in the laws of Coulomb and Biot-Savart play ′ an important role in this dynamical model. µ J (r ,t) Rˆ B r 0 ′ The MQS model is obtained from Darwin if the usual ( ,t)= 2 × dτ (2) 4π ZZZ R continuity equation is replaced by the static one where we use the notations17 ′ R = r r , R = R , Rˆ = R/R (3) ▽ J = 0 (10) − | | · It would however appear strange to keep the static conti- Thus the MQS model may be defined by the equations nuity equation unchanged in this time-dependent situa- (2), (5), (6), (7) and (10). It is different from both EQS tion so we replace it with the usual continuity equation18 and Darwin by including inductive but not capacitive effects. A confusing feature of MQS is that the very fun- ∂ρ damental continuity equation may be violated by MQS- + ▽ J =0, (4) ∂t · solutions. Only stationary currents are allowed in MQS 3 and these cannot explain changes in the charge density, thus one cannot interpret the currents in MQS in terms of TABLE I: This table give two equivalent definitions of each of the quasistatic models EQS, MQS and Darwin charge transportation. Of course, only sufficiently good approximations of solutions to the Maxwell equations are Model Def.A,eq.# Def.B,eq.# interesting in the real world so the continuity equation EQS (1),(2),(4) (11),(12),(14),(15) must still be almost true in some sense. Ampére’s law, MQS (2),(5),(6),(7),(10) (11),(12),(13),(16) which imply equation (10), is valid in MQS but not in EQS or Darwin (see the next subsection). Thus Ampére’s Darwin (2),(4),(5),(6),(7) (11),(12),(13),(17), (9) law is not always valid in quasistatics. Griffiths and Heald remark that ”The application of Ampére’s law in quasistatic situations can be an extremely delicate and B-fields will be used in the present paper but it is matter”.2 straightforward to modify the expressions so that a for- The textbook of Haus and Melcher11 builds up an un- mulation with D- and H-fields is obtained. derstanding of electrodynamics by using both EQS and Let us now formulate the quasistatic models EQS, MQS. This is of particular significance in the relation MQS and Darwin as approximations of Maxwell’s equa- between electromagnetic field theory and circuit theory. tions (11)-(14). The relations to the integral formulas Then EQS involves capacitance features and MQS induc- in subsection A above will also be considered. The EQS tance features. For systems involving capacitance and model is obtained from the Maxwell equations simply by inductance both models are needed. For such applica- neglecting ∂B/∂t in Faraday’s law while MQS is obtained tions it is crucial that capacitive and inductive aspects by instead neglecting ∂E/∂t in the Ampére-Maxwell law. are not both important in the same spatial place.21 The Thus in EQS we have instead of (13) the equation use of two complementary quasistatic models in the same physical system is clearly a complicating feature if we like ▽ E = 0 (15) to model the whole system numerically. We would then × have to divide the whole spatial region into EQS and and in MQS the equation (14) is replaced by the usual MQS subregions with appropriate continuity conditions Ampére law22 at the interfaces. A better alternative may be to use the Darwin model which embrace all the physics contained in EQS and MQS, still being quasistatic. ▽ B = µ J (16) × 0 To get the Darwin model we do not neglect all of ∂E/∂t B. From Maxwell’s equations to EQS, MQS and but keep the Coulomb part of the E-field (defined by Darwin. equation (6) above) and replaces the Ampére-Maxwell law with the Ampére-Darwin equation23 Let us now consider the formulation of quasistatics in terms of differential equations. The starting point is gen- ∂E ▽ B = µ J + ε C (17) eral electrodynamics with the Maxwell equations: × 0 0 ∂t Note an important difference between these two last ▽ E = (1/ε0) ρ (11) equations. The Ampére law (16) implies equation (10) · and may violate charge conservation in dynamic situations. The Ampére-Darwin equation, like the Ampére- Maxwell equation, is consistent with the continuity equa- ▽ B = 0 (12) · tion (4).

∂B The electrostatic models may thus be defined with fo- ▽ E = (13) × − ∂t cus on the laws of Coulomb and Biot-Savart (like in the previous subsection) or alternatively as approximations of Maxwell’s equations. These definitions are summa- ∂E ▽ B = µ J + ε (14) rized in Table I. The proof that the A and B columns × 0 0 ∂t define the same models involves only standard procedures found in basic textbooks. In electrostatics one start These equations are complete and general as they stand. from the Coulomb law and obtain the divergence and However, in the presence of polarizable/magnetizable curl of the electric field. By the Helmholtz theorem we media it is in practice very convenient to write them in have the equivalence (assuming appropriate conditions at a different way by introducing the D- and the H-fields. infinity20) Then only the free charge and current densities appear explicitly in the equations. For formal simplicity just E- (1) (11) and (15) (18) ⇐⇒ 4

straightforward way. An alternative procedure is to TABLE II: The table includes several familiar laws from elec- follow the standard textbook derivation of Maxwell’s tromagnetics as well as the Maxwell-Darwin equation and 1 state if they are valid within the EQS, MQS and Darwin mod- equations but in the final step, when the displacement els. current is added to Ampére’s law for consistency with the continuity equation, find the Ampére-Darwin equa- Eq.# Equation EQS MQS Darwin tion (17) rather than the Ampére-Maxwell equation (14). (1) Coulomb’s law yes no no This replaces the usual maximal assumption in the cor- (2) Biot-Savart’s law yes yes yes recting term10 with a kind of minimal assumption. Both (4) Continuity eq. yes no yes these derivations of Darwin’s model may however seem (10) ∇ · J = 0 no yes no rather superficial and arbitrary. It would be nice to obtain (11) Gauss law yes yes yes it from Maxwell’s equations by using some more system- atic method. A basic feature of quasistatic approxima- (12) ∇ · B = 0 yes yes yes tions to Maxwell’s equations is the instantaneous prop- (13) Faraday’s law no yes yes agation of fields. Thus it should be possible to consider (14) Ampére-Maxwell yes no no quasistatics as some limit c of Maxwell’s equations. (15) ∇× E = 0 yes no no This is however quite a singular→ ∞ limit and the procedure (16) Ampére’s law no yes no must be further specified. We take the absence of time- (17) Ampére-Darwin yes no yes retardation as being the most salient property of quasistatics. Let us express Maxwell’s equations in terms of the potentials (V, A) so that the retarded time appears In the same way we of course have (6) (9). In magneto- explicitly. We choose to use the Coulomb gauge statics one starts from the law of Biot-Savart⇔ and obtain the divergence and curl of the magnetic field. One finds A = 0 (22) the equivalence ∇ · and follow the derivation in Nielson and Lewis12 (the rea- (2) and (10) (12) and (16) (19) ⇐⇒ son for not using the Lorentz gauge is considered soon). The equations for the potentials become (see Griffiths1 The above relations trivially remain valid if we allow for p. 421), time-dependence where time appears only as a parameter. However, in a time-dependent situation it is logical ρ 2V = (23) to use the general continuity equation (4) instead of the ∇ −ε0 static one (10). Then instead of (19) we find

(2) and (4) (12) and (17) (20) ∂2A ∂V ⇐⇒ 2A µ ε = µ J + µ ε (24) ∇ − 0 0 ∂t2 − 0 0 0∇ ∂t Thus the Ampére-Darwin equation is a mathematical implication of Biot-Savart’s law combined with the continu- Time-retardation appears explicitly if we solve equation ity equation. The relation (20) is not included in most (24) for the vector potential in terms of an integral in the standard textbooks but has (more or less explicitly) ap- usual way. The omission of retarded time in this integral peared in several contributions to this Journal.24,25,26. is the same as excluding the second order time-derivative The equation (7) for the Faraday electric field is for- so that mally analogous to the Biot-Savart law (2) for the magnetic field. Formally similar to (19) is the equivalence ∂V 2A = µ J + µ ε (25) ∇ − 0 0 0∇ ∂t (7) and (12) (8) and (12) (21) ⇐⇒ The model so obtained consists of the equations (22), (23) and (25). These constitute the Darwin model in Table II includes some laws of electromagnetics and 27 states if they are valid within EQS, MQS or Darwin. All terms of potentials. The usual Darwin equations (11), the included equations are familiar from standard electro- (12), (13) and (17) is obtained by use of magnetics with the exception of Ampére-Darwin’s equa- ∂A tion (17). E = V , B = A, EC = V (26) −∇ − ∂t ∇× −∇ Why didn’t we use the Lorentz gauge? It is straight- C. A limit of instantaneous propagation in forward to make the corresponding calculations also in Maxwell’s equations that case. We then get a model approximating Maxwell’s equations which seems pretty close to the Darwin model In subsection A above we obtained the Darwin model with the Ampére-Darwin law as one of the equations. from EQS by including Faraday’s law in a simple and However, Gauss law (11) is not obtained but a new term 5 appears in the corresponding equation. Certainly we pre- 1. The force between two point charges depends not fer not to change Gauss law and this problem, as we have only on their separation but also on both their ve- seen, does not appear if we use the Coulomb gauge in the locities and accelerations approximation procedure. Using the Coulomb gauge in the Darwin model also have the nice consequence that 2. Furthermore, it is the position, velocity and accel- the Coulomb part of the E-field is given by the scalar po- eration at retarded time of the other particle that tential E = V and the Faraday part by the vector matters. C −∇ potential EF = ∂A/∂t. − The second point is far more challenging than the first one. However, time-retardation vanishes within qua- D. The potential representations of EQS, MQS and sistatics and the approach with interacting point charges Darwin become quite simple and instructive. In the point charge approach to electrodynamic models we automatically include the continuity equation (4). Therefore this section We use the Coulomb gauge. The magnetic field is writ- concerns EQS and Darwin but not MQS. Let us start ten B = A for all models. For MQS the vector with the EQS model. The fields from a point charge Q potential then∇ × satisfy within EQS is 2A = µ J (27) ∇ − 0 Q Rˆ E r ( ,t)= 2 (33) or in integral form 4πε0 R

µ0 J ′ A = dτ (28) ˆ 4π ZZZ R µ0Q v R B (r,t)= × (34) 4π R2 For EQS and Darwin the corresponding equations are (25) or where the notations in (3) are used. The position of Q is r′ which is a function of time and v = dr′/dt. From µ (J + J Rˆ Rˆ ) ′ these equations we obtain (1) and (2) while the continuity A = 0 · dτ (29) 4π ZZZ 2R equation (4) is implied by the point charge description. Let us now consider the Darwin model. Then equation It takes some manipulation involving partial integration (34) remains valid while (33) only gives the Coulomb part and the continuity equation to derive (29) from (25). The of the electric field scalar potential is just the Coulomb potential for EQS, Q Rˆ MQS and Darwin, E r C ( ,t)= 2 (35) 4πε0 R 1 ρ ′ V = dτ (30) The magnetically induced part E must now also be in- 4πε0 ZZZ R F cluded. However, it is then convenient to start all over The electric field for EQS is conservative and is just the again using the electromagnetic potentials. The poten- Coulomb field tials are Q E = V (31) V (r,t)= (36) −∇ 4πε0R while the electric field for MQS and Darwin also include the magnetically induced electric field so that ˆ ˆ µ0Q v + v RR ∂A A (r,t)= · (37) E = V (32) 4π 2R −∇ − ∂t The scalar potential (36) is just the Coulomb potential and the vector potential (37) may be found from (29). E. Fields of a moving point charge An alternative derivation is by the ansatz

µ Q v We will now consider the quasistatic fields from a point A (r,t)= 0 + φ (38) charge in general motion. Coulomb’s law is often taken 4π R ∇ as a starting point for electrostatics. It is conceptually where the scalar field φ is determined by the Coulomb simple to build on interactions between point charges in gauge condition (22). It takes a little algebra to find this way, so why not develop all of electrodynamics by generalizing this approach? This is discussed by Griffiths µ0Q 1 φ = v Rˆ (chapter 2). Two major problems are − 8π · 6 and thus (37).28 The electromagnetic fields is related to and for the Darwin model the potentials in the usual way (26). The magnetically ∂ 1 2 1 2 1 ∂V ∂A induced part of the electric field is ( ε0EC + B )+ ( E B + ε0 )= E J ∂t 2 2µ0 ∇· µ0 × ∂t ∂t − · (42) EF = ∂A/∂t (39) − In EQS only the electric field is associated with energy. This expression will obviously involve the acceleration Building up a magnetic field costs no energy in this model a = dv/dt of Q. The Darwin model should thus follow due to the absence of a counteracting induced E-field. In from the potentials (36) and (37) of a point charge and MQS only the magnetic field has energy. Changes of the it is sufficient to check (2)-(7). Only equation (7) is not electric field is associated with changes in the charge den- obvious and the easiest approach is to use the equiva- sity. However, the continuity equation is not satisfied and lent differential equations (8). But these equations are there is no charge transport for which we could calculate trivially satisfied by (39) using the Coulomb gauge and the required energy. Finally the Darwin model includes B = A. both magnetic and electric energy. Note that only the The∇× standard textbook approach to electrostatics be- Coulomb-part of the E-field is associated with energy. gins with Coulomb’s law (33) for a charge at rest. Some- An important qualitative difference between EQS and times the corresponding approach to magnetostatics is MQS on the one side and Darwin’s model on the other is used by taking (34) as a starting point but now, of course, the possibility of natural resonances in the latter. Since with a moving charge. This may in principle seem wrong Darwin includes both capacitive and inductive features since this is a time-dependent system. Most textbooks we may in principle use these equations to model, for ex- take this seriously and starts instead from the Biot-Savart ample, some field theoretic manifestation of a LC circuit. law with stationary current. However, leaving the point particle point of view also makes magnetostatics more difficult than electrostatics. It comes as a relief that the III. IF MAXWELL HAD WORKED IN BETWEEN AMPERE´ AND FARADAY magnetostatic differential equations (12) and (16) are formally similar to, and not much more difficult than, the electrostatic equations (11) and (15). Quasistatics is eas- We consider in this section various procedures to find ier than statics in this respect; it allows for the use of the full set of Maxwell equations starting from the laws of (34). Coulomb and Biot-Savart. The quasistatic perspective is A conventional Lagrangian description of the electro- useful and the models EQS, MQS and Darwin will appear magnetic interaction between two or more charged par- as intermediate stages. The title of this section refers to ticles is possible only if time-retardation may be ne- the possibility of introducing the displacement current 34 glected. It is straightforward to give such a formulation before Farday’s law. constructed from the potentials (36) and (37).30 This re- Let us start with the standard textbook procedure of 1 sults in the so called Darwin Lagrangian that was first finding Maxwell’s equations. From the static laws of obtained by Oliver Heaviside in 1891.31,32 It was found Coulomb and Biot-Savart we find, by use of the static again by C.G. Darwin in 1920 by an expansion of the continuity equation, the static limit of Maxwell’s equa- Lienard-Wiechert potentials.33 Rather surprisingly it tions. By introducing magnetic induction we then obtain then turns out that that terms up to the order (v/c)2 MQS (i.e. ”Electrodynamics before Maxwell” in Griffiths inclusive are to be kept in the Darwin Lagrangian (see textbook). The final step, motivated by the need for con- Jackson15 p.596). sistency with charge conservation (4), is to introduce the displacement current. At this point a question mentioned already in the introduction may surface: Maxwell’s equa- F. The Poynting theorem tions includes waves in free space but why should these follow from charge conservation? Quasistatics is useful for discussing this issue. It is indeed not necessary to Let us follow the standard derivation of Poynting’s the- introduce all of the displacement current to save charge orem while using the quasistatic models. I.e. we add the conservation. It is sufficient to include the Coulomb part two equations obtained by taking the scalar product of of it10 and then we get the Darwin model. At this point the curl E equation with the B-field and the scalar prod- no surprising new physics appears. The step from Dar- uct of the curl B equation with the E-field. We the get win to the Maxwell’s equations may then, to begin with, (after the usual manipulations) for EQS be motivated by symmetry and beauty of equations. The explicit appearance of the Coulomb electric field in the ∂ 1 2 1 ( ε0E )+ ( E B)= E J (40) Darwin model is a rather unsatisfactory feature and it ∂t 2 ∇ · µ × − · 0 is so easily fixed by just replacing it with the total elec- for MQS we obtain tric field. Amazing new physics now appears and this is accompanied with beautiful new mathematical structure ∂ 1 1 like the Lorentz invariance of both the Maxwell equations ( B2)+ ( E B)= E J (41) ∂t 2µ0 ∇ · µ0 × − · and of the trajectories of test charges. This is of course 7 surprising but, at least, the previous rather mystifying -Savart be generalized to time-dependent theory? motivation for it has now been removed. Jammer and Stachel discuss what might have hap- 2. Consider the trivial generalization given by the pened if Maxwell had worked in between Ampére and equations (1) and (2) where time is included as a Faraday.34 As in the standard textbook derivation of parameter. What is the significance of these for- Maxwell’s equations one may first derive the static limit mulas? of Maxwell’s equations from the laws of Coulomb and These questions are usually not addressed in the text- Biot-Savart combined with the static continuity equa- books. The first question is considered by Jefimenko and tion. At this stage Maxwell might have added the dis- the resulting formulas involves (as must be expected) in- placement current before the discovery of Faraday’s law. tegrals where the retarded time appears.29 Jefimenko’s This would result in the EQS model which, according to generalized laws of Coulomb and Biot-Savart are used by Jammer and Stachel, is exactly Galilei invariant. They Griffiths and Heald to address the second question. They suggest in the article abstract that the discovery of Fara- find that2 day ”would have confronted physicists with the dilemma: give up the Galilean relativity principle for electromag- (a) The generalized Biot-Savart law reduces to the netism (ether hypothesis), or modify it (special relativ- standard one if ∂2J/∂t2 = 0 (then, from the conti- ity). This suggest a new pedagogical approach to elec- nuity equation, also ∂3ρ/∂t3 = 0). tromagnetic theory, in which the displacement current and the Galilean relativity principle are introduced be- (b) The generalized Coulomb law reduces to the stan- fore the induction term is discussed.” This approach is dard one if ∂J/∂t = 0 (and thus also ∂2ρ/∂t2 = 0). however less striking than it first seems. We like to in- (c) The law of Ampére holds if ∂2J/∂t2 = 0 and clude the appropriate invariance structure not only for ∂ρ/∂t = 0. the field equations but also for the trajectories of test charges. Here one get problems with the Galilei invari- These results are all sufficient conditions for the laws ance of EQS while the Lorentz invariance of Maxwell’s but only (c) is also necessary. Below necessary and suf- equations is perfect. The problem of finding satisfac- ficient conditions are formulated. It is natural to con- tory quasistatic and Galilei invariant approximations to sider the validity of Coulomb and Biot-Savart in time- 35 Maxwell’s equations is not an easy one. dependent theory from the perspective of quasistatics. A third approach to Maxwell equations is of more qua- We may divide electromagnetics into three regimes: stat- sistatic nature. We avoid in this derivation the static ics, quasistatics and high frequency. However, statics is limit of Maxwell’s equation by allowing for (trivial) time- just a particular case of the general Maxwell equations dependence in the laws of Coulomb (1) and Biot-Savart while quasistatics is an approximation. This motivate a (2). We obtain the differential equations of EQS (i.e. the third question : equations (11), (12), (14)) and (15) by pure mathematics if we also use the (time-dependent) continuity equation 3. When does it happen that a solution in quasistatics (4). Maxwell’s equations now follow directly when the is an exact solution to Maxwell’s equations? Faraday law is introduced. Also in this case (like in the standard textbook derivation) a question may appear. It will be shown that a solution of the Darwin model Why do the magnetic induction experiments of Faraday also solves Maxwell’s equations if and only if the current imply electromagnetic waves and time-retardation? The has the form answer is somewhat more hidden in this case. The electric field in EQS is a pure Coulomb field EC but when ∂EC (r,t) we include magnetic induction we also get the Faraday J (r,t)= a (r)t + b (r) ε0 (43) − ∂t electric field EF . It is now not clear if we should have E or EC in (14). In the latter case (the minimal as- where the Coulomb field is defined by (6) and the vec- sumption) we get the Darwin model and no qualitatively torfields a and b satisfy a = b = 0. Formally we new physics appears (only what is needed to explain the write this ∇ · ∇ · experiments of Faraday). The step from Ampére-Darwin Maxwell + Biot-Savart Darwin + Eq.(43) (44) to the Ampére-Maxwell law has already been discussed. ⇐⇒ Let us compare this with (a) above. The last term in IV. THE LAWS OF COULOMB AND (43) may appear new and unfamiliar in the context of BIOT-SAVART IN TIME-DEPENDENT THEORY being a part of the true current (this term but with op- posite sign is familiar as a part of the displacement cur- The laws of Coulomb and Biot-Savart provide a start- rent). However, this term is just the solenoidal part of J ing point for the static theory of electromagnetics. Let the current. Thus consider any current density and us consider the following two questions: write it uniquely in accordance with Helmholtz theorem as the sum of two parts J = JT + JL where JT is irrota- 20,36 1. How should the static laws of Coulomb and Biot tional and JL is solenoidal. Then it follows from the 8 continuity equation and the time-derivative of (9) that but also necessary. Consider a current density of the form JL = ε0∂EC /∂t. The result (a) may be reformulated in a way− similar to (44) but then we only get the left im- J (r,t)= a (r)t + b (r) (49) plication (i.e.” ”) and the last term in (43) is then as- ⇐ where the vectorfields a and b satisfy a = b = 0. sumed to be linear in time. The leaking capacitor is one ∇ · ∇ · of a few examples below showing that quasistatics may Then apply exactly also with a nonlinear time-dependence of Maxwell + Ampére MQS + Eq.(49) + Eq.(4) the current. ⇐⇒ (50) Let us now prove (44) and start with the right im- Consider first the right implication. Both Ampére’s law plication. That Darwin is satisfied follows directly from and Ampére-Maxwell are valid so ∂E/∂t = 0. By us- the equivalence (20). Then both Ampére-Maxwell and ing the time-derivative of Faraday’s law we then find Ampére-Darwin are satisfied implying ∂E /∂t = 0. F ∂2B/∂t2 = 0 so from the second order time-derivated From (8) it then follows that ∂2B/∂t2 = 0. From the Ampére’s law we get ∂2J/∂t2 = 0. Then (49) follows if second order time-derivative of Ampére-Darwin and (12) we also use that the current has zero divergence. The we now get left implication in (47) is obtained by just reversing the 2 procedure above. ∂ ∂EC J + ε0 = 0 (45) We will now consider a few examples involving the re- ∂t2 ∂t sults above. The first one will also be used in the next and (43) easily follows (including the conditions a = subsection. b = 0). Consider next the left implication in (44).∇ · By Example 1 substituting∇· (43) in the Maxwell-Darwin equation we find Consider an (infinitely) long solenoid or a B = µ (at+b) so that B is at most linear in t. From toroidal coil with time-varying current I (t). We assume ∇× 0 the coils are winded so that we may neglect the axial cur- the time-derivative of (8) we then get ∂EF /∂t = 0. In this case the Ampére-Darwin and the Ampére-Maxwell rent in the solenoid and the poloidal current in the toroid. laws are the same and the solution to Darwin also solves Within quasistatic we get (exactly) zero B-field out- Maxwell’s equations. side the coils (from the Biot-Savart law and symmetry). Above we found a necessary and sufficient condition The MQS and Darwin models coincide because there is (44) for the exact validity of Biot-Savart’s law within no charge density and accordingly E = 0. There is Maxwell’s equations. Thereby the result (a) of Griffiths C a magnetically induced electric field E = EF outside and Heald was generalized. Let us now consider the cor- the coils in spite the fact that the B-field vanishes ex- responding result for Coulomb’s law and allow for a cur- actly. In general this is only an approximative solution rent of the form to the Maxwell equations since the Ampére law (and of ∂E (r,t) course the Ampére-Darwin law) rather than the Ampére- J (r,t)= b (r) ε C (46) − 0 ∂t Maxwell law is satisfied outside the coils. An exact solution is obtained, as we have seen, when the current is where b = 0. We will prove that linear in time. The electric field is then constant in time. ∇ · The second example shows that quasistatics may agree Maxwell + Coulomb EQS + Eq.(46) (47) ⇐⇒ exactly with Maxwell’s equations even when he time- variation is quite arbirary. Let us start with the right implication in (47). From Coulomb’s law it follows the E-field is conservative so Example 2 Consider any given spherically symmetric that EQS is obtained from Maxwell. From Faraday’s charge and current density satisfying the continuity equa- law it then follows that ∂B/∂t = 0 and from the time- tion (4). derivative of Ampére-Maxwell we now get (using E = EC ) By the spherical symmetry we get E = B = 0. The solution to EQS, Darwin and Maxwell’s∇× equations∇× is ∂ ∂E E = E , B = 0 and J = ε ∂E /∂t which is valid for J + ε C = 0 (48) C − 0 C ∂t 0 ∂t any prescribed charge density ρ = ρ (r, t) in (6). The equivalences (44) and (47) apply with a = b =0. A and (46) easily follows. To obtain the left implication particular case is a leaking spherical capacitor. We as- of (47) we first substitute (46) in Ampére-Maxwell and sume that the medium in between the spherical plates obtain. B = µ0b. This equation together with (12) has some conductance and does not violate the spheri- results in∇× a time-independent B-field and thus Faraday’s cal symmetry. In between the plates there is, during the law is satisfied. This is what is needed for the Maxwell slow discharge, zero charge density. The Ampére law is equations to be valid for an EQS solution. The proof of clearly not satisfied because B = 0 while J = 0. (47) is now completed and the result (b) is generalized. The examples 3-5 below illustrates that6 in a linear, Let us finally consider the result (c) involving Ampére’s isotropic and homogeneous conductor the conduction law. This condition turns out to be not only sufficient current (assuming Ohm’s law), in certain cases, does not 9 create any magnetic field. This corresponds to exact so- The vanishing of the B-field is a static phenomena lutions of Maxwell’s equations that also solves EQS and • while in the nonstatic case there is a small magnetic Darwin. field outside the solenoid.

Example 3 It is not good to consider the time-varying elec- Let us now consider an example without • symmetry in the charge distribution. Space is assumed to tromagnetic field as a source for the induced elec- be linear, isotropic, conducting and homogeneous charac- tric field, the source is rather the time-varying 16 terized by (ε0, µ0, σ). Assume that we at time t =0 know current. the charge density ρ (r, 0) = ρ (r). 0 However, even though both these statements are cor- From J = σE in the continuity equation and by use of rect in view of the Maxwell equations, there may remain Gauss law we find some uncomfortable feelings. The way we in practice calculate the induced electric field from Faraday’s law make −(σ/ε0)t ρ (r,t)= e ρ0 (r) (51) it natural to think in terms of the time-varying B-field as a source and, even if there is a small B-field outside the The corresponding solution to Maxwell’s equation is solenoid, the main part of this source is well separated in space from the effect we consider (i.e. the induced out- −(σ/ε0 )t E (r,t)= e E0 (r) , B = 0 (52) side E-field). This apparent action at a distance seems to contradict the local action quality of Maxwell’s equa- where E0 (r) is the Coulomb field due to the charge den- tions. But of course, a careful analysis shows that there sity ρ0 (r). is no true contradiction involved. Let us now consider this problem from a different point Example 4 The last example above may be modified so of view. If we use a quasistatic model in the calcula- that we are dealing with a finite conductor surrounded tions it may be favorable to also think in terms of this by free space. Let the conductor contain all given charge model and not in terms of the full set of Maxwell’s equa- ρ0 (r) at time t = 0 and let the surface of the conductor tions. This is analogous to what we often do in other be an equipotential surface of V0 (r) defined by fields of physics. For example we use classical mechanics ′ without worrying about quantum mechanics, Newtonian 1 ρ0 (r ) ′ V0 (r)= dτ (53) kinematics without referring to special relativity or New- 4πε0 ZZZ R tons law of gravitation without thinking about general Inside the conductor both the charge distribution and relativity. It is in the solenoid example easy to calcu- the solution to Maxwell’s equations is the same as in late the magnetic and electric fields using quasistatics. the previous example. Outside the fields are static with One gets an exact solution within the quasistatic model and a good approximation to the true Maxwell solution. E = E0 (r) = V0 and B = 0. The surface charge density on the conductor−∇ is This quasistatic solution (of MQS and Darwin) has no magnetic field at all outside the solenoid. This solution −(σ/ε0)t is fundamentally inconsistent with Maxwell’s equations σS (r,t)= ε0 1 e ˆn (r) E0 (r) (54) − · since there is a time-varying E-field in free space where the magnetic field vanishes exactly. However, the so- where ˆn is the outward normal. lution is of course consistent with quasistatics. This is Example 5 Related to the last example is the leaking an indication of the quite different nature of quasistatic capacitor of arbitrary geometry and with homogeneous, models and Maxwell’s equations. The PDE’s appearing isotropic and weakly conducting material in between the in quasistatics are often elliptic or parabolic while the two perfect conductors.4 Maxwell equations are hyperbolic. The conceptual problems have their roots in using quasistatic calculations but

In this case we must define V0 (r) by Laplace equation interpreting the result using Maxwell’s equations. with appropiate boundary conditions on the two perfect conductors. The solution may be then expressed in terms of E = E (r)= V as in the two last examples above. VI. ELECTROMAGNETIC SIMULATIONS 0 −∇ 0 A. The PC in basic courses V. A QUASISTATIC PERSPECTIVE ON QUESTION #6 Electromagnetic simulations are essential to electrical and electronic product designs in many industries. A A time-varying current of a long solenoid causes an broad range of important applications are within the induced electric field at the outside. But how can that quasistatic regime like motors, sensors, power genera- be? There is no magnetic field at the outside! A.P French tors, transformer systems and Micro Electro Mechanical asked a similar question and it was addressed in several Systems (MEMS). Industrial and scientific applications papers.5,6,7,8,9 The answers included two essential points may involve extensive calculations and accordingly the 10 need for much computational power. However, the usual B. Equations for eddy currents PC has developed enormously and so has the software for simulations with increasingly user friendly interfaces. Eddy currents and associated phenomena like induc- Today the PC is a sufficiently powerful tool for model- tive heating, magnetic shielding and magnetic diffusion ing many electromagnetic phenomena within just seconds are in practice very common since the basic ingredients of computational time. This makes it potentially useful are just a conductor and a time-varying electromagnetic as an instrument for teaching basic electrodynamics. In field. Still eddy currents are not much discussed in the the PC lab students may simulate various electromag- textbooks. An exception is the one by Smythe where a netic phenomena. Examples within statics or quasistat- whole chapter is devoted to eddy currents and analytical ics include the edge effects in the parallel plate capacitor, solutions are given in terms of series and integrals.37 An distributed currents in conductors of various shapes, the interesting method of images may be used to calculate Hall effect, magneto resistance, the appearance of non eddy currents in a thin conducting sheet. The theory vanishing charge density in inhomogeneous conductors, was developed by Maxwell and reformulated in modern the electric field outside and surface charge on a conduc- terms by Saslow.38 However, the analytical theory for tor with currents, various objects placed in a given ex- eddy currents is still comparatively difficult. Electromag- ternal static or time-varying electromagnetic fields, eddy netic modeling on a PC is today an attractive and easy currents, inductive heating, magnetic diffusion, magnetic way to include more about eddy currents in basic courses. shielding and more. Eddy current problems fall into two classes, steady- state and transient. In steady state analysis (also called time-harmonic analysis) we simply replace ∂/∂t with the factor jω and allow for complex valued fields. Maxwell The use of electromagnetic simulations as an effec- equations becomes time-independent and much easier to tive pedagogical tool also presupposes some support from solve. Physically we may obtain a steady-state condition theory. One should in particular focus more on how sufficiently long time after the start of a time-harmonic Maxwell’s equations are used to formulate well-posed source of the fields. The initial transient behavior is not PDE-problems for various physical situations. Most considered in this analysis. The steady state Maxwell textboks, at least the basic ones, are not much influenced equations may be further simplified in quasistatic situa- by the appearance of computers. Their guiding principle tions, this may be useful for analytical theory but is not of is to find analytical solutions in simple and instructive much interest for electromagnetic simulations. However, cases. The underlying PDE-problem may not be needed in time-dependent (transient) analysis an initial value explicitly in some examples like when integral laws are PDE must be solved. Then a quasistatic approximation used in very symmetric cases or when tricks like mirror may imply a major numerical simplification changing a charges or mirror currents work. When the PDEs are ac- hyperbolic PDE into an parabolic or elliptic one. tually used, and this is much more the case in the more Let us assume that the time-variation is slow enough advanced texts, one finds solutions in terms of integrals so that quasistatic theory applies. But what quasistatic and series by comparatively tedious calculations involv- model should we use in the study of eddy currents; EQS, ing, for example, special functions and variable separa- MQS or Darwin? Obviously not EQS since magnetic tion. induction (Faraday’s law) is outside the scope of that model. What about MQS? This seems to be the standard model for eddy currents calculation but it may in fact 15,37,39 With a numerical PDE-solver available the situation is only be used in some particular cases. A quite common misunderstanding concerning the quasistatic ap- somewhat different. The formulation of well posed PDE- 27 problems for various phenomena is now motivated with- proximation may partly explain the popularity of MQS. out the intention of finding analytical solutions. This Let us now consider situations where the use of MQS is an easy part of the PDE theory for electromagnet- may be justified. This happens in certain symmetric ics and it is instructive by providing increased insight cases when MQS and Darwin are equivalent models. We into the structure of Maxwell’s equations. It is in par- consider below two examples of such symmetries. The ticular important and sometimes straightforward to find first one include, as a special case, the situation when Maxwell’s theory for eddy currents in thin conducting how Maxwell’s equations reduce thanks to various sym- 38 metries. For example, the independence of one Cartesian sheets applies. coordinate or alternatively an axial symmetry results in Example 6 Consider a body in which the conductivity a decoupling of Maxwell’s equations. A simple inspec- σ = σ(z) only varies in the z-direction. In the xy di- tion of the equations written in component form reveals rections the body is homogeneous and of ”infinite” extent.− this structure. Such symmetries explain, for example, The electromagnetic field is created by the use of an ex- why the magnetic field of a toroidal coil only has an az- e e e ternal current density J = Jx(t,x,y,z)ˆx+Jy (t,x,y,z)ˆy imuthal component or why there are TE and TM modes with vanishing divergence; Je =0. in planar wave-guides. In the next subsection we illustate ∇ · this by formulating PDE problems for eddy currents. In this example there will appear no charge density 11 and the Darwin model is equivalent to MQS. The reason The traditional use of Darwin’s model is not eddy cur- is that electric fields will only appear in directions of rents but concerns charged particle beams and plasma constant conductivity (i.e. they have no z component) simulations. and they cause no pile up of charge. The− system may Let us write equations for the Darwin model in term be described in terms of the potentials Ax and Ay while of potentials. The conducticity of a possibly inhomoge- e V = Az = 0. From equation (25) with J = J +σ( V neous conductor is σ = σ(r) and that the time-dependent −∇ − e e ∂A/∂t) we get the following equations for Ax and Ay external current J = J (t, r) is prescribed The total current may be written J = Je + σ( V ∂A/∂t). We −∇ − ∂Ax 2 e use, as always in the Darwin model, the Coulomb gauge µ0σ + Ax = µ0J (55) − ∂t ∇ − x A = 0. The dynamics of the potentials are determined ∇· ∂Ay by the continuity equation µ σ + 2A = µ J e 0 ∂t y 0 y − ∇ − ∂ ε 2V (σ V + σ∂A/∂t)= Je (57) These 3D equations determine the dynamics and may − 0 ∂t∇ − ∇ · ∇ −∇ · be used to model physical systems with the above sym- and the Ampére-Darwin equation metry. A similar 2D axi-symmetric case may also be formu- ∂ ε µ V µ (σ V + σ∂A/∂t)+ 2A = µ Je lated. − 0 0 ∂t∇ − 0 ∇ ∇ − 0 (58) Example 7 We use cylindrical coordinates (r, φ, z) and These equations have a non-standard appearance by a rotationally symmetric conducting body with conduc- containing mixed time and spatial derivatives. They may tivity σ = σ(r, z). The external current density is Je = however be solved using Comsol Multiphysics. Actu- J e(t,r,z)φˆ. ally there is a considerable freedom for the user to en- ter non-standard equations in this PDE-solver. This is Also in this case there will appear no charge density. achieved by making the equations available also on the Some potentials vanishes, V = Ar = Az = 0, and the ”weak form” level which is the natural form for the fi- system may be described in terms of the vector potential nite element method. The numerical solutions appears, ˆ A = A(t,r,z)φ. This time we get the equation for the at least qualitatively, to behave in an expected way for dynamics as the few examples that we have considered. The practical usefulness of the Darwin model for eddy current calcula- ∂A 1 ∂ ∂A ∂2A e tions remains however to be proven. µ0σ + (r )+ 2 = µ0J (56) − ∂t r ∂r ∂r ∂z − A qualitatively new feature of Darwin’s model (as op- The equations may easily be studied by using a PC and posed to MQS or EQS) is the possibility of resonance. some PDE-solver.40 For example, a straight copper wire Since the Darwin model includes both capacitive and in- above a copper plate is considered by Backstrom using ductive phenomena it may in principle be used to model equation (55) and the solver FlexPDE.41 He also con- systems where the energy oscillates between the electric siders a similar example using (56) for a circular copper and magnetic fields. Consider, for example, a field theo- wire above a circular copper plate. The transient fields retic version of a LC-circuit. A suitable design for such appearing when a current is turned on in the wire is stud- an application may be a resonator in the form of a short ied (pages 75 and 99 in Backstrom’s book). The same coaxial cable where one end is short circuited by a metal models are also easily studied using Comsols Mulphysics plate and the other end is still electrically open but with with the Electromagnetic Module. These problems are increased capacitance created by the use of two close par- numerically simple and, of course, there are many other allel plates, connected to the inner and outer conduc- PDE solvers that may be used. Such numerical tools tor, respectively. This kind of resonator has applications make it possible to consider many cases for which there in connection with electron beam devices at microwave are no analytical solutions. frequencies (example 3.4.1 in the textbook of Haus and 11 The examples above were carefully designed in order to Melcher). The resonance frequency of a LC-circuit is avoid charge density. This is also true for all eddy current f = 1/√LC but the frequency must be low enough not examples in Saslows’ paper (see in particular Appendix to violate the quasistatic assumption. In the above de- B of that paper).38 Usually there will however appear sign of a resonator the most essential method to achieve time-varying surface charge on conductors with eddy cur- this (in the absence of dielectrics or magnetic materials) rents and, in case of an inhomogeneous conductor, there is to make the capacitance large and thus take the two will also appear charge density inside the conductor. In capacitor plates very close to each other. order to describe the physics we then need to include the scalar potential in the analysis and MQS cannot be used An interesting possibility is then to use the Dar- VII. SUMMARY win model. Such applications of Darwin’s model have been suggested in connection with eddy currents in the Maxwell’s equations are fundamental for the descrip- human body caused by high voltage transmission lines.42 tion of electromagnetic phenomena and valid for an enor- 12 mous range of spatial and temporal scales. The static cludes the Coulomb part of electric field. limit of the theory is well defined and of course much easier. The electric and magnetic fields may in this limit (3) EQS and MQS may be considered as approxima- be given by the laws of Coulomb and Biot-Savart. How- tions of the Darwin model. EQS includes capaci- ever, as soon as there is any time-dependence we should tance but not inductance while MQS includes in- in principle use the full set of Maxwell’s equations with ductance but not capacitance. Poynting’s theorems all their complexities. Time-retardation is a fundamen- for these models show that there are only electric tally important but also a complicating feature. Using energy in EQS and only magnetic energy in MQS. Maxwell’s equations means in analytical theory that even if the effect is small it will not vanish and this makes the (4) The law of Biot-Savart is valid within EQS, MQS theory unnecessary complicated. In numerical analysis and the Darwin model. these effects, however small, may force us to use smaller timesteps (for numerical stability) and expensive calcu- (5) The law of Coulomb is of general significance for lations. It is therefore useful to introduce quasistatic quasistatics (EQS, MQS and Darwin) as is obvious approximations in Maxwell’s equations. The quasistatic when we use the formulation in terms of force laws models are also useful for a better understanding both (subsection IIA) or use potentials (subsection IID). of low frequency electrodynamics and for explaining the transition from statics to the general high frequency elec- (6) The law of Ampére is not of general significance within quasistatics. It is valid only in MQS but trodynamics. This have been discussed in the present 27 paper and below we list a few major points. not within EQS and Darwin.

(1) The quasistatic limit of Maxwell’s equations is a (7) Galilei invariance of quasistatics is a somewhat del- kind of c limit obtained by neglecting time- icate issue. Galilei invariance structures may be retardation.→ ∞ The Darwin model is obtained if we deﬁned for EQS and MQS but if we like the force use the Coulomb gauge. to have the corresponding invariance then only the electric force is included in EQS and only the mag- (2) The Darwin model involves both capacitive and in- netic force in MQS.35 ductive features but there is no radiation and the interactions are instantaneous. Poynting’s theorem (8) Quasistatics has important applications in electro- for this model shows that there is both electric and magnetic modeling of transient phenomena. magnetic energy, but the electric energy only in-

1 David J. Griffiths, Introduction to Electrodynamics (http://web.mit.edu/6.013 book/www/). (Prentice-Hall Englewood Cliffs, NJ. 1999) 3rd ed. 12 C.W. Nielson and H.R. Lewis, Methods in Computational 2 David J. Griffiths and Mark A. Heald, ”Time-dependent Physics (Academic Press, New York, 1976), Vol.16, pp generalizations of the Biot-Savart and Coulomb laws” Am. 367-388. J. Phys. 59, 111-117 (1991). 13 Dennis W. Hewett and John K. Boyd, ”Streamlined Dar- 3 A.P. French and Jack R. Tessman, ”Displacement currents win simulation of nonneutral plasmas” J. Comput. Phys. and magnetic fields” Am. J. Phys. 31, 201-204 (1963) 29, 166-181 (1987) 4 D.F. Bartlett, ”Conduction current and the magnetic field 14 Mathew R. Gibbons and Dennis W. Hewett ”The Darwin in a circular capacitor” Am. J. Phys. 58, 1168-1172 (1990) direct implicit particle-in-cell (DADIPIC) method for sim- 5 A.P. French ”Question #6. Farady’s law” Am. J. Phys. 62, ulation of low frequency plasma phenomena” J. Comput. 972 (1994) Phys. 120, 231–247 (1995) 6 R. Mills ”Answer to Question #6” Am. J. Phys. 64, 681 15 John David Jackson, Classical Electrodynamics. (John Wi- (1996) ley & Sons, 1999) 3rd ed. 7 Allan Walstad ”Answer to Question #6” Am. J. Phys. 65, 16 It has always been tempting to regard ∂B/∂t in Faraday’s 462 (1997) law as a source of an electric field. This has sometimes 8 Jacques.D. Templin ”Exact solution to the field equations caused confusion and objections; currents and charges in the case of an ideal, infinite solenoid” Am. J. Phys. 63, should be considered as the only sources of electromagnetic 916 (1995) fields. The integrals expressing the fields in terms of these 9 R.J. Protheroe and D. Koks, ”The transient magnetic field sources involves in general the retarded time so that the outside an infinite solenoid” Am. J. Phys. 64, 681 (1996) news from the sources propagates with the finite velocity c 10 H.S. Zapolsky, ”Does charge conservation imply the dis- Howevever,˙ in quasistatic approximations the interactions placement current” Am. J. Phys. 55, 1140 (1987) are instantaneous so this objection is less valid. 11 Herman A. Haus and James R. Melcher, Elec- 17 The notations in Ref.1 has been the preferred choice in this tromagnetic Fields and Energy, (Prentice-Hall, paper but here is one exception. Griffiths use a script lower Englewood Cliffs, NJ, 1989). The textbook is case ”r” instead of ”R” but unfortunately I have not found also freely available from MIT OpenCourseWare a nice representation of this symbol on my computer. 13

18 The static continuity equation together with (1) and (2) Century-Crofts, New-York, 1966) constitute a quasistatic model without capacitive or induc- 30 Allan N. Kaufman and Peter S. Rostler, ”The Darwin tive effects. Model as a Tool for Electromagnetic Plasma Simulation.” 19 It is important to note that the integral (7) converges. This Phys. Fluids 14, 446-448 (1971) is evident from the law of Biot-Savart which implies that 31 J. D. Jackson, “From Lorenz to Coulomb and other explicit the B-field goes to zero as 1/r2 when r →∞. gauge transformations.” Am. J. Phys. 70, 917-928 (2002) 20 We consider physical systems where the charges and cur- 32 O. Heaviside, ”On the electromagnetic effect due to the rents do not extend to spatial infinity and may use as motion of electrification through a dielectric” Philos. Mag. boundary conditions that the fields approach zero suffi- Ser. 5 27, 324-339 (1889). The Darwin Lagrangian in Ref. ciently fast far away. 19 was, as mentioned in Ref. 20, obtained already here 21 This is never a problem according to Ref. 11 but I have 33 C. G. Darwin, “The dynamical motion of charged parti- not been able to convince myself that this is true under all cles.” Philos. Mag. Ser. 6 39, 537-551 (1920) circumstances. 34 Max Jammer and John Stachel, ”If Maxwell had worked 22 Ampére’s law ∇×H = Jf differs from (16) if there is a non- in between Ampére and Faraday: An historical fable with vanishing polarisation current ∂P/∂t. However, Ampére’s a pedagogical moral” Am. J. Phys. 48, 5-7 (1980) law is strictly speaking a static equation and we prefer to 35 M. Le Bellac and J.-M. Lévy-Leblond ”Galilean electro- use (16) as a time-dependent generalization with the same magnetism” NuovoCimento 14B, 217-233 (1973) 36 3 name. The term ∂EC /∂t goes to zero as 1/r when r →∞. 23 Like in Ref. 2 we use the names Ampere’s law for equation 37 William R. Smythe, Static and Dynamic Electricity (16) and Ampere-Maxwell for (14). Then it seems natural (McGraw-Hill, New York, 1968), 3rd ed. to use the name Ampere-Darwin for equation (17). 38 W. M. Saslow, ”Maxwell’s theory of eddy currents in thin 24 P.A. Mello, ”A remark on Maxwell’s displacement cur- conducting sheets, applications to electromagnetic shield- rent,” Am. J. Phys. 40, 1010-1013 (1972) ing and MAGLEV.” Am. J. Phys. 60, 693-711 (1992) 25 T. Biswas, ”Displacement current-A direct derivation” 39 Eckhard Baum and Otto Erb, ”Eddy currents and elec- Am. J. Phys. 56, 373-374 (1988) trical surface charge” Int. J. Numer. Model. 16, 199-218 26 N. Gauthier, ”A direct derivation of the displacement cur- (2003). rent”, Am. J. Phys. 56, 871-872 (1988) 40 There are many PDE-solvers that are suitable for electro- 27 There seems to be a quite widespread misunderstand- magnetic simulations in basic courses. In the textbook by ing that quasistatics, defined by the omission of time- Gunnar Backstrom (Ref. 41) the program FlexPDE is used. retardation, includes Ampére’s law (16) as a valid equa- I am using COMSOL Multiphysics with the Electromag- tion. This is stated or implied by many textbooks. See for netic module. Information about these programs may be example Jackson (Ref. 15, p. 218) or Smythe (Ref. 37, p. found on the homepages http://www.pdesolutions.com/ 368). However, this is not correct for an obvious reason. and http://www.comsol.com/. Ampére’s law implies stationary current and thus constant 41 Gunnar Backstrom, Waves by Finite Element Analysis charge density. But certainly, a slowly varying charge den- (Studentlitteratur, Lund, 1999). sity cannot force us to include time-retardation. 42 Pierre-Arnaud. Raviart and Eric Sonnendrücker, ”Approx- 28 Ernst Breitenberger, “Magnetic Interactions between imate models for the Maxwell equations” J. Comput. Charged Particles.” Am. J. Phys. 36, 505-515 (1968) Appl. Math. 63, 69-81 (1994) 29 Oleg D. Jefimenko, Electricity and Magnetism (Appleton-