SOLUTIONS MANUAL to accompany ELECTRICITY AND MAGNETISM MUNIR H. NAYFEH MORTON K. BRUSSEL University of Illinois JOHN WILEY & SONS New York Chichester Brisbane Toronto SingaporeCopyright © 1986 by John Wiley & Sons, Inc. This material may be reproduced for testing or instructional purposes by people using the text. ISBN 0 471 80692 7 Printed in the United States of America 10987654321CHAPTER 1 * * 1.1 The cross product A x # is perpendicular to both & and 3. Since we want a unit vector, we just divide by the magnitude: A = Ax B/|& x 3 keys Kxb-=]2 -6 -3] = 15% - 1loy + 302 fo ot [a x B] = As? + 107 + 30? = 35 ~ _ 15e - Loy + 302 _ 3x - 2y + 62 SR a ee 1.2 The position vectors of these three points are B= aK - y+ 2, By = aK + Oy - » and Py= x + ay + 22. ‘The position vector of an arbitrary point is f= xx + yy + 22. If all vectors lay in the plane, then the following triple cross product vanishes. (B= EF, - Fp e(F, - Fy) = 0 which gives llx + Sy + 132 - 30 = 0. 1.3 The position vectors of these points and an arbitrary point x - 2y - 42, and are #, = 3k + + 22, #, ® = xx - yy + 22. The equation of the plane is governed by thecondition (# - #,)+(# - #,) = 0 which gives the equation 2x + 3y + 6z + 28 = 0 for the plane. de"/de = ner™! = net 2F 4 > 1.4 a) vr" b) Venlz|) = 2 denlé|/dr = t/r Vis) =f d(i/r)/dr = -2/r? 1.5 Consider the surface defined by £ where £(x,y,z) = 2x22 - 3xy - 4x - 7 = 0. Recall that Vf is normal to surface f(x,y,z) = 0, then yy : Ve = RR + RW + ge B= (22) - By — ADK - Bay + bed At (1,-1,2) we have: VE = (8 + 3 - 4)x - 3y + 82 = Tx - 3y + 82. The unit vector normal to the surface at this point is: EE vel : fates ¥122 1.6 By definition we have dé/ds|. = vel. Now Ve = 2xyzdx + x23) + 3x2y22Z, thus at x = 2, y = 1, z = -l we have vo = -4x - 4y + 122. Hence, d@/ds is maximum along (-08 = 45 + 122)/Va? + 4? + 12 direction and [vel = Wir + reir? = Te.1.7 Using Eq. 1.57 WCE /e?) = W/E ye + Wee/e = BEE + 7 1.8 Using the vector identity given in Eq. (1.60) we have GK x B = BCG xR - Red x B = 0, since Vx k= VK B= 0. 1.9 A is irrotational if V x A which gives K(c + 1) - $(a - 4) + 2(b - 2) = 0. Therefore a = 4, b = 2, > - . « Hence A= (x + 2y + 4z)x + (2x - 3y - z)y + ca on abs dbs BOs (ax -y + 22)2. Now A= = k++ ae z. Thus a # =x + 2y + 4z . Partially integrate with respect to x: 2 o= ae Qxy + 4xz +£(y,z). Now - = 2x - 3y - z, thus partially integrating with respect to y gives = 2xy - 3y2/2 - ye + g(x,2)- Now 32 = 4x - y + 2z, thus partially integrating with respect to z gives: % = xz - yz + 22 + h(x,y). Comparing the 36's, we see that if F(y,2) = ~3y2/2 + 22 - yz, g(x,z) = x2/2 + xz + 22, (xy) = > x2/2 + oxy - 3y2/2 then we get a @ where V6 =A. That is @ = x2/2 + 2xy + 4xz - 3y2/2 + 22 - yz + a constant. 1.10 VeA = 1+ 1+ a= 2+ a = 0 which gives a = -2. Lal = 2/r? = f/r is conservative (spherical coordinates are > * z implied) if ¥x#= 0. Note that B= EY.roof $.i-— oo a Ez. oO 0 ex - 3 (x 9B /36 - reine § a£,/20) = 0. Me can quickly r’sing see that any spherically symmetric (6,4 independent) radial field is conservative. Thus V x (r/r) = 0. 2 al ao 3 L # = t/e = -v6 = -( 5, ot Ostet § aime #) - Thus I/r = - 3¢/ar = -dd/dr. Or -d@ = dr/r which integrates to = -tnr +c. Applying %(r = a) = 0 gives c = gna; thus ° Qna - gnr = -£n(r/a)- 1.12 Notice that if we can find a scalar function $(x,y) such that A= Vo as required in the second part of this problem, then it would be a necessary and sufficient condition that A be conservative. However, we will go ahead and explicitly show > + that ¥V x A= 0 to show that A is conservative: v x = 4h[2(3x22 - y)/ay - a(3x? - 2)/ a2] “Jf a(3xz2 = y)/ax - a(6xy + 2°)/ 22] 42 [3(3x2 - 2)/ax - a(6xy - 2°)/ay) = R-1 = (-1)) - (322 = (32) + 26x - (Ox)) = 0. Therefore + > vx A= 0 +A is conservative. Now we find such that A= Vo.R= (oxy + 238 + (3x2 - 29 + (axe? - yé = 00/ax K+ 20/ay ¥ + 30/az 2 Integrating 36/2 = 6xy + es gives = ax°y + xz? + gly,z)- Integrating 34/2y = 3x7 - 2 gives ¢= 3x7y - zy + h(x,z)- Integrating 36/82 = 3x2" - y gives @= a - zy + f(x,y). By inspection of these three relations we can easily show that (x,y,z) = 3x2y ad oe -zy+C. + 32-3. The surface integral le g Fenda can be written in terms of a volume integral using the divergence theorem I = [Ver dv = {3dv = 3V where V is the v v volume of the surface. t aA, 3A. 3a, : oh = e+ ahs Bo 4 - by + 22. 1.14 a) Vv meta ta tte b) From here on, it is easiest to switch to cylindrical coordinates where VeA = 4 - 4psing + 2z. In cylindrical coordinates, x2 + y” = 4 is the same as p= 2. From the figure the unit vectors normal to Sy is p, normal to SQ is z, and normal to $3 is_-2. The divergence theorem states J vehdv = § Redd = [ (4 - Spsing + 22) pdodgdz ¥ v f ee = f dz [do f pdp(4 - 4psing + 2z) = 84m 0 0. 0 Now we calculate the surface integral g Read = f Read Tf Read, + [ Reda, using Ss s S. 5. * i: 3 52s oy : B = Rcosg + Jsing, and A = 4xx - 2y2y + 222 Ready = (ax 2 - 2y°¥ + 272) (cos + Fsind) pd ddz = (Apcos”4 = 2psin?4)pdédz This gives Reda, = B(cos”$ - 2sin*4)dodz at p = 2 Read, = (272) +(2)pdgdp which is equal to Yoded} at z = 3. Read, = (22) +(-€)pdedp which is zero at z= 0. Adding the three contributions we get G keda = 9 t odo i do +8 a (cos?$ - 2sin?4)d6 i dz s 0 0 0 which gives 841, the same as [V-Adv. 1.15 a) We first find the unit normal. The cylindrical surface is described by f(x,y) = x2 + y2 - 16-0. We use f= Vf/|Vf£|. The gradient of f is VE = 2xx + 2yy. The magnitude of the gradient is |V£| = %? + y%. But everywhere on the surface we have vx? + y* = 4, therefore |v£| = 8. Hence n= (xx + yy)/4. b) Substituting for A and & in Aen wa get Kel = (xz + yx)/4. Writing this in cylindrical coordinates we get Coscong + p2singcos)/4. Thus fAenda = fAenpdédz = We 5 J J (2cos$ + 2singcos$)dgedz = s(x + 1)/2. OG: 0x y z 1.16 v xk = Jasax a/ay a/ az 2 2 ax -y -yz -y°z Vx A= x(-2yz + 2yz) + y(0 - 0) + 200 + 1) = 2. Using xsin@cos$ + ysin@sing + zcos®, B> " n> 4 x = rsin@cos$, y = rsin@sing, and z = rcos®, we have (v x A)enda = (2)+(sindcoss + Jsindsing + Zcos8)r“sinddédd = cos6+dgdOsiné at the surface of the sphere (r = 1). ‘Therefore oS n/2 W/2 {0 x Kenda = 2" [ cos@sinedé@ = -2n cos’6/2 | = 1. Now we 0 0 soy o calculate § Aedr. In this case df is along a unit circle on the x,y plane = its direction is along $. ‘Therefore Rea? = [(2x - y)x - y2?¥ - y222]-[$rag]. at r= 1 and using spherical coordinates we have Kea? = [(ax-y)x - yz2¥]+[-%sing + yoosslag = -(2sin6cos$ - sin@sind)sind - (sindsin$)(cos“6)coss > : n/2 along C, therefore Aed® = 2 cosssing + sin@ Now, 6 = 7 oe 2 2) G Aedt = [ (-2costsing + sin’o)d¢ = cos'o | + 7= 0 0 0 thus f x Kenda = f Kea? Ss c 1.17 a) [nda = 0. From problem 1.1% we have [v@ dv = 66 n da. Let @ be a constant function such that 7vo = U, and we can take 6 outside the surface integral. Thus 0 = fn da or Jn da = 0. b) §® xn da = 0. From problem 1.18 we have fv x av = fa x Bada. Taking B= #, the radius vector, then vx f= 0 and we have §f x nda = 0. 118 a) [vx Badv= fax Bda. Let A= bX C in the divergence theorem where C is a constant vector. Then es + + 4 [¥(B x C) = §(B x C)en da oe > os Since Ve(B x C) = C*(V x B), and (B x C)en = ae na > > * * Ce(n x B) then C+fv x B dv = Coif n x Bda or : ae b) fvodv = [¢ ada. Let A= C in the divergence theorem + where C is a constant vector. Then fveat dv = [oten da rs Now we have V«(6C) = C*V@, then > oe 7 Cef¥e dv = C+fa nda or [Vo dv = fon da1.19 It is by far the easiest to solve these problems by using spherical coordinates, thereby taking advantage of the symmetries of these functions. 2 dénr lod 2 1 Eigora) (oa dee) 2 a(n + 1)r ale di 2 #l TF ael™ xx + yy + Zz 1.20 Given that Fe(Gek) = Coe + v9 + 22) (CR + Hy + 22) (xx + yy + 22)] (#+o)-#. There is no ambiguity in writing ‘ ee ee ee 1.21 We use Eq. 1.76 with V= FS + ¥ oe zs, and Ba xx + yy + 22; thus VE = xx + yy + 22.CHAPTER 2 2.1 The total force is the vector sum of four individual forces of equal magnitudes but different directions. The magnitude of each force is F = qgq/(4negr”) where r? = 17. Now only F cos® contributes, where 6 is the angle each force makes with the normal to the plane of the charges. ‘Thus F = 1.73 2Ne total 2.2 Consider a charge element dq = Ads. We first notice that due to the symmetry of the problem, the x-component of the & = (1/4neg) dq'sind/R?. Substituting as field cancels, and dEy ij dey > dq! = Ads = ARd@, then |E| = (A/4me,R) fsinede = ' ° -N(4ne,R)cos8 | = A/2me,R. But A= q/mR, therefore + 4 ° E = qy/2n7e er. A d2( 28 - 22) _ 2.3 df = —2-——__—~ ._ Now the z integrals cancels, and dve (4 + 22) 5/2 0 z=0 z=-5 220 2x Adz : Bee fe de te (ede = 2a T3Ty = 05 x LO AV/ me 25, z=-0 25 Ane (4 + 2°) ° where Xo is in C/m. 10Ajdz * 5 ° Ro - zz Ane, cr? a 22y3/2 integrating from z = -d to 2 = d, makes the z component vanish 2.4 Using Coulomb's law we write di = ‘Thus because of symmetry, and thus . z RP § dz a 4, Re Zz i ame a (Re + 22) 3/2 ane x? Ce + 22) 2g 2rd 8 é —— (ane, R)(R? + ai? + 2.5 From symmetry, E has a component only along the axis of the / 22.2988, neve disk. Hence E, = [dicos@ = Oe -ar? dq = oda = opded$. Hence dp woah S or ine, ee eye o z poe ] 2” Ze je ey? 2.6 By symmetry E has only a component in the z direction, the axis of the disk. ‘Thus dE, = —L— 922288 ynere dq = oda and 2 Tre Tp a? da = 2npdp, thus .-L. 7 sg cosa(2npdn | % i cos8 do aoe 2” Gre + 3/2 de > So ee] fo lee P| = h tan8, dp = h sec*ed0, and |? - #*| = h sec@ then ee=a sin@ | or E, ° note that the total charge on the disk is ag q=foda=f 3 =2 (2mp)dp heer ° Now for h >> a, (a2 + h2)7!/2 » 1/n thus E| = a,a/2e.n” = a/4negn which is the E field due to a point charge q = 2mo,a. Thus, far from the disk, its shape becomes unimportant and it looks like a point charge. 2.7 We use c' for the point of observation and r for the charge. dq = 0 dx dy, #' = (0, 0, a), ® = (x, y, 0) s+ 1 4% 4% @-toaxa Be fp See le etl Te aa o x20 y=0 |r - rf (x2 + y2 + 41/2. Thus ey ere eee ae) B= Tye J ax Say eons]? Pe Or? + v2 47) 22 a Ee pee fdx f dy 2x(xx + yy - oot ol 2.8 By inspection we find that only the component along the axis of the disk survive. That is E, = [dE cos® = 1 dq foe f ——+— cosé, where dq = oda = a sin“ pdp dd ee if oeny do sin"ép db, E -—_b___ z “00 © (e" +n?) fe 4 2 o,h 2m a 2 pdp ogh 1 1 sin’ ¢d¢of = f 7 ] ~ te ole 23/2 4e, (a2 + n2yl/2 2.9 a) Consider the o case. From Gauss' law we have |E| = 0/e,- Between the planes the fields substract, and outside they add. Thus B= ox/e, x<-l B=0 “1é x1 gE = -ox/ey oe b) For the o' = -o case, the fields add between the planes and subtract outside the planes B=0 x ¢-l and x> 1 B= ox/eg -1< x <1 oo Ee _ lo -y for y > 2 sheet “ 2e) ‘45 for y < 2 ~ ALG + DY + | - 292) ie anes (y + 1)? + (2 - 2)° Thus check is ee The 2 component can only be zero for > z= 2. Hence for z= 2, Eyiie + Esneet = My + DY + Tp - 0 requires y + 1 = 2 or 2nes(y + 1) 0 IY x =4 + Thus y +1 = +4are solutions. 132.11 Let the Gaussian surface be a cylinder of length L, and radius r- E will be radial by symmetry. Then c £ § Bead = ee(amre) = L 20 fore" 2? erdx faz o o ° ° = 12 ' 2.12. a) For r < R we have [| gee (amt )dr' o. 2) r r + > | ana ae eouG aes fieda = 2 fart => = 0 0 0 R 2 ' b) For r>R, fieaa = 2 f 4 pe . 4nak/e, = 4ar°e. Thus oo or E = aR/egr’. 2.13 Flux penetrating is equal to Q/e, where Q is inside S. tae a) (ay +4) + 45)/€, = 1el x 10°" Coul/e, Qn 2 2 vy L foia=t ff SE paoag = 2 cout/e . € ° ° ° 00 0 Fi , 2m 2 c) — foda ae {do sing fodp = 0 & oo ° 4) (20 - 20) x 1077 Cle = 0 14Q Since total flux is fied = 7° 2. A © D = egE = q/4nr? = 9.55 x 107! ! ¢/m2 then flux density is 2.15 To find the charge densities, we use the differential form + of Gauss' law: VE = p/ey + . * a) £ = 10 sin@r + 2 cosed Pe Cor’ a VeE = eB oe (10r"siné) + Psind 96 Q cos@sin@) or + 1 13 rs 2, b) WE = >a 555 (p - a’) = a. Thus charge density ts . Se te ee p= ae, for a

b. Thus charge density p = 0 for p> be a « But dq = oda = ordr® so le- 3] (pe, + W)/2 4 ° an al rodrd@ _ oW ' eae f J 7 = Ter + It doesn't depend on py ° o/2 ° ° ‘0 a bax 2 2.17 a) o= tre, ee « Note that x = rytan@, dx = rysec*6d6 and r = r,sec8, thus ¢ = 7 1 ome, 15terete aera: ones eta ast remnant tn =n rhe earn eam ener don r+ constant b) In the £ >> r, limit we have @ = - 1 2ne, 1 c) For ry, ry << £, we have Vjy = r,) - (ry) = i . This is, as we would expect, the same as the ° potential difference of two points in the presence of an infinite line of charge. eG 30° rsind a6" + 2.18 a) We use E = -V where V = spherical coordinates. Thus E = t(2ar~>cos® + br@2) + 8(ar~3sine). b) We use a = VeE. Consider first ? # 0. °° z_ 1 38 (2 ec) WE = “5 oe (EL) * reine 96 (2198 Eg Pea ie -1 toe oe €, "(28 (lar “cos® + b) + 755 gq (ar sin’) = -2ar “cos® + 2ar “cosé = 0 Thus there is no charge at f # 0. We now inspect the equation as @ +0. Also we note that Ve#/r? = -V*(1/r) = 4n6(#) where from example 1.4, the Dirac 6-function 6(f) = 0 for r # 0 and [s(2)43F = 1 if volume includes %= 0. This would have been obvious if we noticed % = b/r is a point charge potential. We should also recognize (a/r2)cos® as a dipole potential of dipole moment 4nega, in the 2 direction and placed at r = 0. See 16example 2.9. ©) We use the following in one step: 2= df = §.(-¥6) = -v7s. (See Ex. 3-20). z é € oe a °o > 2.19 a) Using Gauss' law § Bead = nt ives LA “get. Q b) Using Gauss' law gives i, = om r 0 “2 €) Take pf} = -p, gives the cavity, thus #) = 7-27". Adding +e + MY gives B= Ey + Ey = 52 zoe 0 ee ES ea = ly a) @=-ffedt =—22 f az= =a 2 r' cose’. One 3e, 70 36. 70 ° oo ° can also find the potential by finding the potential due to the > sphere and the cavity. (Same procedure used for E.) 2.20 We notice that this problem has symmetry about the y axis, so it is natural to set up the problem as in example 2.9, with the axis redefined: : P 17-L @- The potential at point P is clearly ¢, Tre, & da}. Now, we note that Vr? + (2/2)? = recosa, x" = fx? + (2/2)? + recosa We are to assume £ cosa 0 2k: 2 2 i Cee a Cee as seen in example 2.9. We now need only to transpose this in terms of spherical coordinates, i.e. what is cosa in terms of ele 6,6? Note cosa = — == x rs no Sets oe pcs = but 2 = sina, £ = cos(F - 4) ee ~ 1 at sinesing sin $. Therefore cosa = sin@sing, and 4, ine, 2 spherical coordinates. 2.21 We can superimpose the potentials of the four charges and expand assuming that r >> a just as we did in Ex. 2.9. In here, however, we will use Eq. 2.69 just as we used it in Ex. 2.16. +4 +% -} “2g 34 agp TO tee OO 434 Thus we write (see Fig.) 663) = -8(3).ye(2) = -az-vol2), where 9) = 88 13 cos70 - 1] (Ex. 2-16). Now 18(2) = pedaa” 2 5 paae ve) = ¢ [92 [3 cos’ - 11] + 6 [42 [-6 cosesine)] « 4neg 4negr Using zer = cos®, z*6 = -sin® (Eq. 1.3) we get 2 4). 39a 4ne rc? 0 [3 cos0(3 cos”8 - 1) + 6 cosé(cos6 - 1) 7 = 893 p,(cos) « ee Ane or 2.22 We use the multipole expansion of Eq. 2.61 since we are interested in the potential at large distances. We note that o(2) = 0 and 6(1) = 0. Thus because charge is in x-y plane 92) 2 LL pp 3¢8.3)? - pt ]dq. Let # be at (r, 8, 0); ane, 33 then £+3' = p'sin@cos$! and hence 3(r+3')2 - p'2 = [3 sin?ecos?4! - 1]p'7. Now dq = Gh [6(0" - a) - 6(p' - b)]p'dp'd¢'. Thus the integral becomes: 2am] ff {t3sin2ecos2@* - 11% SIRF IIE 5¢ ot = a) - 8(p' - b)]dp'as! 000 ae Qn = Qa Ps =f [3 stn2ecos%¢* - ngage - fF 13 sin’ @cos*6" - np ae’ ° ° u e, 2 2_ 2 22 Sealed ne Olea C2) (ieee big) 730 one vA sin sat) so ¢ coy ee (F sin’e - 1) im = 2) (3 cos28 - 1) 19This is the potential of a linear quadrupole where two charges 1 of -q each are at a distance of (a2 - b*) /2 from the central charge of 2q- 2.23 The dipole moment is defined as (taking the z axis to be the axis of symmetry) § = [f'dq. Taking dq = oda = oR*sinededd where R is the radius, and using #' = x sinécos$+ y sin@sing + 2 cos0, give 6 = ok? [(x sindcoss + y sindsing + 2 cos®)sinédéd$. Only the 2 component survives, and we have U2 = 3 = -2n0r2z [ cose deos@ = roRz. : 7 2.24 We first calculate E and VE. + 86 2 ab ad eee B= -to = -« 3 Fy 7 Bae 7 OO + OX) VE = -ayee, is a diadic. > a) U = -p*E is the potential energy of a dipole in an external field: U = poxt(ajx + a7)X = po(ajx + a). At x = U = pom: + b) f + j oe : BeVE = (Pox) *(-0) xx) = -a)pox = U- > > > px E+ #x £ where # x f is the usual torque due to the ce) 7 distance from the origin (or the point the torque is being measured around). Here, # = 0 (at the origin) so %= p x E = (PX) x (-Cayx + ay)%) = . 2.25 a) One can solve this by calclating f = ~4Eqipoj_ OF by > > f£ GE harge* We will use the second for illustration of > x + yy + 22: the method: i ee oe cake - Ge es yx + ysl?B= pz then $7 = - p 2 and h fap? Ce . P gz and hence f =P ay Echarge og [3k + yy + 222 _ 1 a ey 2,372 ag pRprcos F ane, 5 me Fspole b) The couple is given by t = § x E + # x f according to Eq. 2467+ : to ed ee p_sind $ 2 Gaegr r 4neor 2.26 Dipole $, at the origin, dipole p, at #. ‘his must be Ane ° : done carefully, expanding all products. If V*E = 0 which is the + * + + case at r#0, then F = (P+V)E = WBE) = Be( VE). a) The electric field of the dipole p, at a point of observation F is given by Eq. 2.4 ta ee ae Ae 4ne,B, = 35 "(B,+F)E + BL, thus p, rByis oe 3% yt vty 8 et 4ne py ck, = 3r (Py r)(P, ee ree (OTD 5) 1 which is symmetric in 8), By- Now fae WB yf) = 31 OB, VB, + 37 DG, “F) Be a ae + 3x (p+) Vp, or) - Wr “Cp, *P,) 21= 30 BG DGB + 3 OVE G2 + 36GB, BUR, «FB = VP, «BD F = [-15(8, PR, PE + 38,8, +8) + 3B, (8, 98) + 308, BE Vane e! : b) In the form above, all terms are in the form (8) +r), (®, +t) or (B,+8))+ ALL terms are proportional to pypy+ Lf Pp Poe a then # = +4. The interdependence shows this is the 4ner © maximum. + 2.27 a) We first determine the E field of the two charges at the position of the dipole: (x = 4x + 22 eae i -a(x - 4) tax + 4) 4nek. 2 Sor ESS on ee ee ee 4me, d6,/8x = 0 (by symmetry) at x = 0. Amey By = a2[ (Cx ~ a)? + 2297/2 - (x + a)? + 223/21 0 224ne,9E,/ax = ~6qde(d? + 2*)~9/? at x = 0. Note that 9E,/ax = 2E,/92. Thus : fe — VE(2 = 2) = bade(d? + 22)75/2¢x2 + 2x)/4ne, and Be Book = oaanca? + 02) 9/2 0p 2 + pM /4ne, We use the results of problem 2.26. Taking r= 2, By = XCBy et) = 0, then F = —1L [32(f,+2)432(, of) 1. Hquating 4 Fs 2 Ane the forces, we get: B, = 2qd(1 + d2/ 92)-5/ 2g, 2.28 We use Eq. 2.62. * 1 dq, + or) = tae, ll tt re - Sslea] r Since r' = R = constant, then the first term gives 0?) = Ro/eg+ Because of symmetry one can show the other two integrals vanish. Thus 6(#) = Ro/eg- 2.29 a) The monopole moment is equal to the total charge = q. The dipole moment is p = [#edq = 2, Zz (J dedzyd = $0 2 = ° 0 2 = 0. The quadrupole moment = a% 2 7 Na f 2'"dz" = yqet ad ri")da" = 5d ee ° b) The monopole moment is the total charge = q + q' = 2nRA ~ 2nRA = 0. We note, firstly, that the point charge is added at ?" = 0 and thus makes no contribution to any moments higher than the monopole. Thus, the dipole and quadrupole moments are those calculated in Ex. 2.15 an Qn or EAL Gxt? et?yrao = an f Gar7c086 - R7)a8 = ° ° Qn 5 ar? f [3€1 + cos26)/2 - 1]do = ar2/6 ° by = qk2/6, and Of, = -qRr2/3. 2.30 We use kq. 2.69, 062) = -8(2).9@(1) and 9) = * : qa cos6/4negr”, and §(2) = -az. Thus (2) (a coser - a singd)-(24-2 c088 7 4 da sind 5) 4neyr 4ne,t 24CHAPTER 3 3.1 From the results of Ex. 3.1, and using @, = 1/2, we have @ = V £n(tand/2)/en(tane,/2). The charge density o = egi,- Now Se 7 E = -0(d¢/d6)/r = -V8/[rsinogn(tan6)/2)]. Thus 0 = -eQV/ran(tan6,/2) at @ = 1/2. 2 3.2 a) =: B® _ 9 which gives ® = ab + b where the boundary pe a¢' conditions are: 6(¢ = 0) = 0 and ($= 8) = V thus $= 5 6 .+ - 1 6 b) From Gauss’ law arf = ofeg, but E= 5, 6 or a E= z $, thus o = #V/8p. The total charge is foda. Taking sve, °2 p 4Veqh da = hdp we get Q = — fe BT 2M Py!) + 1 3.3 From Ex. 3.4 we have tan(@,/2) > > - tan( 0/2) 1 e - 8 ae @ = Ven (ance, fay)! (antB, 7D + Now -E= + V8= 235 or z= eae Sve rsind tan(6)/2) r siné gn(- tan(8,/2) f b) Inside the metal of the plates, E + At the surface, there is a surface charge density, 0, such that eee 0 tan(6,/2) r siné cater) eewhere ® stands for 8, or 8,, and + is used for 6, and - is used for 3.4 a) 4 (r,0) = ar + Br YP (cose) eo(rsa) = Tan ee Bt =") yp (cose) n=0 b) ‘The following are boundary conditions that must be satisfied: 1) ¢ is finite at r = 0 which gives B, = 0 for n > 0 2) $52 0 which requires AY = 0 for n > 0 3) = 6) = V, cos@ at r = K requires that A, = B, = OU for 0 n #1, AJR = Vg, and By? = Vy. Therefore (1,6) r cos® r ¢ Rand 6)(r,6) = Vo (4)7cos8 x > Re -Vv yoo -_1 3 ce . ©) Ep = -¥e) = GP-S Spb or, be =P 2. Now > -¥0, = (V_/R) (2 cosér + sin® 8). (Note that the ° tangential component is continuous). It cannot be conducting because there is a tangential component. 2. 4) Eq - Eyn 7 O/eg+ Now Ep, = cos and Ey, = 8. z cos Thus o( 6) = (3V,€,/R) e) We can calculate it from the following expression: > + + p = fo(@) rda where da = R2 sinOdédé and r = Rr. Alternatively we can read it off from the expression for the potential outside the sphere, since this potential is a dipole potential. Comparing with Eq. 2.44 we get § = 4aR” voe, 2. > 3.5 To get E, just get -V@ where 6 is given in Hq. 3.35 - 26a * 36 * 1 36 Eorort °t ¢ : E = -r[—t + —a— 4 7 2 3 4nesr 4ne, «& - RD) "Loe ("4 - psine ane,(Ro - RP) °° The charge density on the smaller sphere is o = €,£,(R),0) = qa. sl _ 38 cose an U2 3 _ ye Be ee The charge density on the outer sphere 61 +R FR) is calculated in a similar way: o = - ary onl: RD For oe total charge on the outer sphere“we havé 0 =! fot (Ry, 8) da = for, 10) amr? Ja(cos) = -q. me 3.6 We start by writing expansions for the potentials according to Eq. 3.28. For rok: @) = Siar" pr "ly Pp (cose). For ° n=0 r>R: 6&=J (cr dr” “hy P (cose) « Now B, = 0 for all n ie n=0 n a n since oy is finite at r = 0, and Cc, = 0 for all n since % is finite at r = ©. Next we match boundary conditions: Vo c0s20 = J Arp (cos8). Now cos(28) = 2cos%(@) - 1 n=0 arse eee = 5 GF cs oo) 5 4 Peg ee i 2 4 ele “3,4 ee 1 eee? seo ae Oe - FV ZUR (F cos 9 - 5) and Clas ad 2 ~aVor R+Z Vor R°(G cos“@ - 5) 27. 5 B23) on Gee? 3 WE, = 5 Vi2rk G cos’ O-1)r + 7 Vor R 6 cos@ sine)6 and eal -2) 2 4 44 Ey = Vcr ORE + FV, (3 3 we cos*e-1)é + tye te cos®@ siné)6 and 0(8) a - ee 1/20 (3 a. avels = = 1BACR) - ECR} er = evr 3 & cos"e-1) 3) 3.7 Using the fact that the potential is finite at r = 0 and r = @ then we write the following for the potentials inside and outside the shell. 0, = Tale (cose) and ¢ = Sar (le (cose). Now use the n a nlrex information given about the charge at the surface, Ey, ~ 2 9, (cos 6-1) 36, = ——__ where E, ==. Note € an or 22 : : (cos@ - 1)° = 5(P,(cos8) - 3 P, (cose) + 2e)(cos8)). ‘Thus yoni) BR Coosa) + F TA ea a n P (cos) = 20,/3€, * (Py - 21 + 2Po). This gives 3B)R4 + 2A,R = 20,/3€55 °, 40. -3 sia eatan a —(n+2) n= BBR? + AY = = Gs BR gees and Co + DBR = -nA,R for n > 2. We also have the boundary conditions %|, = %|x which gives 23J a,r™e (cose) = J ae) po(ane) n n P(cos8) or AL = Bi Solving these and the previous relations simultaneously gives for the constants: a sero 0 for n> 2. Thus 20k By = ° oR + and Ay = ° 6) = GE (2 - & cose +4)? 1 Geos*e1)), ° 20 R o R, fe 1 R31 2 as 3e, (“2 - (D* cos8 + SQ)” 7 Geos @1)). 3.8 Since the potential should be finite at p = 0 and at p + », then we write the following expansions for 4 inside and outside cylinder: (0,4) = J ™(C cosmé + D_sinmg) and =0 "(Hcosmé + F sin m6). We solve for the (0,6) = J, m=0 coefficients by applying the boundary conditions. First the potential is continuous at p = py, that is 4 = %, es . s | = a + . y: bs (C cos mp + D sin mo) = J a (Ecos mg + F sin mg) m=0 m=0 9), Secondly the boundary condition Ey,-Ej, = 0/€, at p = py gives Sg le coe a + Fo sin ne) ) nm peste cos mee on eT En ™ ym OS m=0 n=0 ‘0 6 p,sin mé) -< cos 34. Note that on RHS of this relation, one ° has cos 3 that is m = 3 only and nothing else, thus E, = aofor m # 3 and Fa = Da = 0 for all m. Solving relations from both boundary conditions give E3 = 0909//6e, and Cy = o4/6€gp7+ Thus 3 4 op Po 6 = cos 3¢, 6, = cos36 be 0% 2 bee? ‘00 0 ise . BL = - Sy (6 cos3e - § sin3¢) 2€ 9 ‘0° oe i ty = $0%, (8 cos3e + § sin34) Epp) 3.9 Since there are no charges, separate variables by taking #(y,z) = Y(y)Z(z) which when az ay substituted gives —> /Z = - —z/Y. Both sides of last dz dy 2 2 equation must equal a constant ts 422.4 2.1 #¥, az 2, ay 2 oe ay SS = + a°z, a’y = 0 which gives Z= A cosh oz + B sinh az, dz ay and Y = D cos ay + C sin ay. The boundary conditions imply: 20) = 0 gives Z = B'sinhaz, ¥(o) = 0 gives ¥ = C sinay, Y(y,) = 0 gives Y = C sin nny/y, where n= 1,2,... or a= nn/y,- The most general solution is (y,z) = Ta, sinh(nmz/y,) sin(nty/y,). To get A, we use the information about the fourth side, that is (y,2,) = V, or V) = Tay sinh(naz,/y,) sin(nay/y,)- Multiply by sin(nay/y,) and integrate from o to Yor 30%o mm paao %o nv mm vf sin ™2Z ay = JA sinh ( ) f sin 2% sin BW ay. ° y, n Y, ve ° O a eo ie oe ’ Now f sin “sin 2 ay = 6mn —, and ° Yo Yo a mW [ sin ™ ay = 2y /mm (m odd). ; ie 0 wey o%o 2 1 Thus A, = {2S --——___— , n odd; therefore sinh ¢ Yo Mo 1 nz Cs =) sin ep. ™ a odd nwz, n sinh (—2) Yo 3.10 The angle between the plates is 45° = 180°/4. Thus we need 2n - 1 = 7 image charges. ‘They are distributed as shown in accompanying figure. 3.11 We use the method of images in this problem. we replace the sphere by placing an effective charge q', a distance £! away where &' = 2R2/2 and q' = ae Taking the charge to be on the z axis, then 9(r,0) = a ea 4 Taq Pe By / where = 22/2, 6 = E'%. oy 31- { q 2Rq/k 1 ame 2+ (4/2) teeosey/? (+ ae 2 AR cos0y!/? We now determine o(6). Now 4 =e es +8 = : Een| = yr Rives Es(n) ea & at r = R which gives = —4 fe o( 8) zi 4m Now for & << R, we can neglect terms of order ay. Thus o(8) = 4; { 1 anx? (1 - Gcose)*/? The force between the charge, q, and the sphere is just the force between the charge, q, and its image, q' = -2Rq/2. From general results + ee Fl = aye ae ana ° 0 b? . a 2. aR?) |Fl = Te for 2 << R. 3.12 eae, a) We want to find the image charges which will make the potential on the boss and plate zero. To make 6 = 0 on the boss asurface, we must have an image charge q' = -qR/D placed at y = R2/D, as shown above. To make = 0 on the flat part, we must have two image charges -q' at -y (below the plate) and -q at -D (below the plate) as shown. b) The potential on the side of the plate is o 4 a6 anid —+ B/D —- 4ne|t - bz| 4ne, lt - (R /D)2| 4ne |r + (R°/D)2| ——4———__ or explicitly 4ne |? + dz] O o (r,@) = 4 f. 1 a R/D fne, “(2 + D% 2rveose)!/? (x2 R4/p? - (22R?/D)cos8) 1/2 “1 : R/D. ] (r2 + 0% 2rveose)/2 (re Ryn? + (2rR2/d)cos0)!/2 +. ©) Charge induced on boss is given by Em = o/e, and Q= f oda - * boss = -e, [VoendA is the total charge. Now, on boss, E must be 2 7 ae purely radial, thus Q = -21e,R J Fel pagsined®. Evaluating 26) or and carrying out the integration gives =R d) Force between the charge and the plate is just the force between q and the three image charges: a =R/D : R/D 1 co - R270)? w+ RD)? (20) EE3.13 a) We use the results of the example in the book on the method of images on a sphere. We need an image dipole located at R°/2, and of moment -pr3/73. The distance between the dipole > and image is (Z, - R2/Z,)+ Note that ®,+r = 0, pyr = 0, thus using the results of problem 2.26 we get > . P2812) == r=——°*— (toward the sphere). RYy4 4ne,(Z, - 7) 0 ae b) E = qr/4ner> is the electric field from a point charge at eae =peseaae > (q/4ne evr) + rWr73)). Now Wr) = T= 2 > r = 0; thus VE Re + $9 + 22, W073) = -3r-4e = -3r73r, thus VE = (q/4neg)(-3e 328 + e731). Now r = %, and (Bee) = 0, then fe Bove = ap/ ane, 73 (see problem 2.25). This force is added to the force in part a to give the total force. Note that the force on a point charge due to a dipole is obvious, and there must be an equal and opposite forces. es > 3.14 ‘The torque acting is t= r x F + p x E where F is the + force acting on it and E is the electric field acting on it. > Taking E to be that of the image dipole, we find that it is + along §- Moreover F (from the previous problem 13.13) is along ?, thus t= 0. 13.15 Using the results of Ex. 3.14, and taking a = 0, we find S that the image dipole needed is p' = 2(Rj/Z)> p and located at b = R3/z,. This is an addition to a charge of Ryp/ 22 located in 34the same place. Using procedures similar to those used in problems 2.25 and 2.26, we find that the force and torque are ea 2 Fe-—2°2 __ (2? 4 or”) z. The torque here is zero Pata Ce tae 2nes(Z) - Ro) +> because #, F, E and § are along each other. In general for any angle a between dipole and the line joining the center of the sphere and the dipole we have = 2 7 ri PRLZ, [cree R2) cosa + 3R,)] Fe-—22 4 2 me, (2 = nyt eR 2 sin2a 00 2 2,3 3ne,(Z, ae RY) 3.16 We use the method of images. We need to find the location of the line of charge and its image. Given x, = d, and the radius of the cylinder R. Thus the line charges are located at 21/2, 4a where a = (R? + x3 Thus the potential a= = an(p-/pt) = = an (ators), Now the charge no fo (Goo ase ye density is 0 = €, i evaluated at x = 0, thus = vit (x + a) a (y) = €y (39/90) |qe9 = ZF (tn ¢ ie (x + a)" 5) ge y+ (x - a)? which gives o =3.17 We use the method of images. Since the cylinders are identical, then we have symmetry. The locations of the lines 1 are at x = 4 a where a = ((A/2)2 + R2)/2. The potential is given by d en[& + ays y ] ee ee a) The charge density is o = -e,V@ evaluated at the surface (x - A/2)2 + y? = Rk. This gives after some algebra: ont (Oc) y? = p4)? + axey where b = 2x4/(A2¥a°— 4x7) b) ‘The force between the cylinders will be that between the image line charges. Thus from Chapter 2: os Ge oreyce Rel ee 3.18 We will solve the problem by solving for the potentials directly. In the region outside the cylinder there is no charge, so the potential satisfies Laplace's Eq.: V°6, = 0. by symmetry, the potential has only p dependence: 1 a a6. A : : 3 Fp (P Gp) = Owhich gives # = Atnp + Be inside the cylinder satisfies Poisson's Eq.: » the potential z e (o 3%) = 2 which after two integrations gives a= p> +A! np + BY. Since the potential at = is not 0 36equal to zero, then we take it to be zero at p= 0. The requirement that ,(0) = 0 requires A’ = B! = 0. Applying Gauss’ law on a cylinder,that encloses the charge distribution a evaulates A: Q = fadv = [a 2xptdp = natn ° ~Ae 2 mp2/p which 2 ean oe 0 gives A= -ap2/2e,. Thus 6) = Te, met B. Applying the boundary condition 9, = ¢) at p = p, gives Best 2, 2 fnp_ and hence %,(p) = Werte es aro ee : 0 0 2 -ap? $,(0) 7 te, {24n(p/p,) + 1}. 3.19 For r < Ry and r > Ry, V2@ = U which gives 6) = A/r +b, and 6,5 G/r + D. For R Ry, 3 + 0 at @ thus D = 0. Now continuity of at R; and Ry gives ~BR ~BR. = a c 20 cE a i, +P, Oe et Fe We cam get the constant 1 3 ° i C by Gauss' law, or by realizing that C/r should be the potential if all the charge was concentrated at the center of the shell. Thus C = (8/2€,)(R - R{)+ Applying Gauss' law inside the shell gives [ Een da = em Ears few 0 r 7 °: rv ° 37sateen iemeaeedeeteeaeeeeneeneeeieneieenceemenenneeinmmnmemnninmmmeennennmememmnremememnaeeeneameneen Ris and hence F = 8Ry/ey, and B= B(Ry - Ri)/eg- € 0 Therefore %) = (B/€,)(Ry - Ry), 2 2. fo er ee ray 2 cere fener, pea ee: ‘ e © ° 0 3.20 See part (b) of Ex. 2.12. It deals with a uniformly charged sphere using Gauss! law to get £ and then integrating to get ¢. The direct boundary value or integrational method can be used. 3.21 The potentials in the z < -2,, z > 2, and Rg Gay 44%) regions satisfy the equations = =0, Z 2 45% dz dz s- = -0/e,+ The boundary conditions are & = 9, F pat dz Z = 29s % = %, y at z= zo, % az, Ey * constant as Z+-@, a z, E, > constant as z ++", E,(o) = 0 at z= 0. Now integrating the equation for %, gives ho 22 °o%o oe ere ae (m2/2z,) + Ae ate oa where A and B are 0 do 20,2. constants. © sin (12/225) - Ala. Since dz ° 4p,2 E,(o) = 0, then A, = 0. Thus 6 = 7 cos(wz/22,) + BL and £5 20525 . Ey = zo sim(nz/2z,)z. Now %) = Cyz + Cy and Ey = But ° 2PaZo E, = Ep at 2 = 29 gives Ey = -C) = —2* . Now 4) = ¢ gives 0 383.22 We guess a solution of the form ¢ = A sin a,x sin ayy sin ajz. Substituting in Poisson's equation sin ax sin a,y sin a,z gives A(at + a} + a3) = p9/e, a and p+ ap sin a,x sin ayy sin a,z. 2 1 3 eo oes 0 (ay + a) + a3) 3.23 We can write the volume charge distrubtion as (x,y,z) = 9, cos(a)x) cos(ay) 6(z). We solve Poisson's equation V6 = p/e, or 2 2 2 one ae ae axe ay? az” jo 6 = cos(a,x) cos (a,y) 8(z)- Assume ° | 4(x,y,z) = F(z) cos(a,x) cos(ay)- If we can find an F(z) and Poisson's equation is satisfied, then, by uniqueness theorem, we get our . Plug ® into the differential equation to get -a} cosajx cosagy F(z) - aj cos(ajx) cos(agy) F(z) + oF cos a,x cos a, cos(a, x) cos(a,y) 8(z)+ el 1 1 2 For z #0, then we get -(at + a3) F(z) fe ey ce fe m dz F=A en!) + a) 2 for z > 0 and ovat ap forz <0. 7 FeA 39do do Now evaluate A. At z= 0, [Fo - Fling or + oe 12 1 2 € 1 2 °, 0 Thus A = ——2—— and alee ab io [oe (x,y,z) = 2e var + a5 eel 9) @ coe a\x cos ayy, z > 0. oe a Zz (x,y,z) = 2 cos ajx cos ayy, z < 0+ 3 3.24 We use V7 = -p/e, with (r) = (1/4ne,)eO* and i 3 wat sc 2). tus p= -e, vo = FE (4 - tarde ™ r oe se 3-25 Now @ = aye Ge tg) exp (29 + Gee ~ Teer 0 ° a -2r/a ee a ee 0 0 -2r/a -ar/a, = -e,v26 = (-qv2 Pe eee!) gives p = -€,V° = (-qV°/4n) te ie + 2 ]= -2r/a Fd + PE — 4] Be), the first two terms 2 e2tla give q8(r) - (q/4n)(2/a) /r (see Ex. 3.20). The last terms give V(e2t/4) = (-a/ae"2¥/@[1/r - 1/a]. Thus ole) = q 8(r) ~ (q/nad)e2t/4, 403426 a) We use d@/dz” = 0 to find that % = Ajz + B, and O) = Ayz + Boe Now at z = 0 we have %) = 0 and at z = a we have 4 = 0, therefore $) = Ajz and a = Ag(z - a). Now as z + -o oS and as z + @ we have E = E,z. Therefore % be Ez and %) = -E,(z - a). a+ es b) o(z = 0) = -e,z+E(z = 0) = ~egE,, (2 = a) = egz*E(z = a) = ££, ‘oO ©) . d) o(z = 0) = -eg£), o(z = a) = 0. Moreover E = 0 in the z > a region as predicted by Gauss’ law since o(z = a) = 0. Thus > . E = E,z in the 2 < 0 region as predicted by Gauss" law. e) pu a2 ee ae -4 ee z +5 eB 2 =0. After grounding weLD LLL ALLL LCL CN CILLA A A CC CCE LL CCT CC CTT 2. The factor of /) has not been yet explained. It is explained in Chapter 4 as due to self fields (Eq. 117). 42CHAPTER 4 * * 4.1 We are given that P = Pz * : +B oso p = -¥-Pz = 0. Now on upper > a) We know p = -¥B, o = f+B = Pcos@. On lower surface n = -z surface i = 7, thus o thus 0 = -P. b) q = foda + fodv = P fcos@ da - P fda. Now L 8 v u 2 2m W/2 2 P fcoséda = PR f[d@ { cos@sinéd@ = MRP, and u Oso P f da = 1p. Therefore q= aR’p - RP = 0. L c) Since the material is neutral then net charge 4) By integration: p = fr dq = ftdq + [rdq. Let be the L u projection of # on x, y plane, i.e. % = rcos®z + rsin@ p = + $. Now on upper surface we have dq = P cos® Rd sinodé My while on lower surface we have dq = -Ppdpd$. Thus > 7% 20 . : ead : p= fasinode § dgPcos@(coséz + sin@p) - § do fae Pp’ pe By symmetry we see that contribution proportional to p vanish > 2a im 2 Pp = -20 R°z [Pcos’ @sinéde 0 Now by definition of B: p RPZ. We see the two answers are equal. 434.2 Dipole moment: p= P fdv = Par” 1/2. aa Polarization charge densities are: o, = Pen = -P, for + a rectangular surface. o, = Pen = P,x+p = P,cos¢ for curved surface. o, = Pen = 0 for top and bottom. p, = -V+P = 0. 4.3 Given x2 + y2 = R2 and P = (ax? + b + cy + a)xx + pxy Now o = Ben, n= THT with f(x,y) = 0 is the equation of surface. VE = 2x% + 2yf and [vE| = 2x2 + y2)/2 = oR Thus 0 = Set = eed 2007 + y") : Thus we have o = Ben = ((ax* + b + cy + ax + pxy)+ CR* YW) R 0, = (ax4 + bx? + cyx? + x2 + pxy /R and -Vep - -(3ax? + bd + cy + a). 4.4 a) Using Causs' law in the region p < 9, gives 0. In the region p, < 9 < pp it gives o od sa 1- t_ a, Pls b= Ho pe and & a oerce 7 Yo. In the region py < p< p3 og 2 2n g2 3 oe oe D= a5 (6) - 91) and "* Rey Tye, (2 e °) * Pinaily in pp recion ce nave pe Bo ee eye 3 ee 2pes 2 eal 0 ° > + b) P= 0 for p > 93 and for p< pp- But P = (1 - 1/K) x a2 _ 02s czati 2p (65 pie for py < p < p+ From the polarization we get the 44polarization charge. The volume density poss = Pp = -Veb = 5 S-(oP) = 0, while the surface density is =n ey ee (pepe 7 oy = a8 givin Ce.) (9) ~ 9)) at p = py and 5 K" 2p) Geo? - 9, = (1 - 1/K) %, (py ~ oy) at p= 63+ Finally 9, = 0 for all other supfaces + eo Ps ©) Vs filrydr = f bdr + f Ear 2 a 2 Ga 22 Po 2 _ 2 2 Bence? ~ p})/2 + gn a (Cp) - 0) )/K ~ 9))) 4.5 a) Take the z-axis perpendicular to the dielectric interface, and the origin of coordinate system at the center of ee E,t- Thus the electric field at the interfaces is purely tangential to it, hence, equating the electric fields at the boundary of the dielectric gives E,(r) = Ey(r) = E(r). Applying Gauss' law to a spherical surface gives i Qn a= 1? fa(cose)f doe(s)|E| = 2mr2(e, + €)E -1 0 SF b) We use of = Den. This gives the following for the vacuum € SS) and dielectric regions: 5 eee 2na’ 0 ad (ee °%f a ee) omc Go c) We use P= D-e,E. Thus at r = a we have 45és t P= ae (e - €,)/(e + €4) and 20a ee o, = Pn =~ (e- eet ede 2ta 4.6 a) ‘There are three regions in all of which the electric field is normal to the plates (say along x). We use Gauss in all three cases. In the vacuum next to 9, plate, we have (Bed = foda or B = o,- In the charged slab we have hcxcnte: fiedd = foda + [odx) da or > : h D = (0, + p(x - h))x. htt In vacuum above slab we have fdeda = foda + [ pdxda or > ix h D = (oy + tp)x. b) 1) On the top plate we have og = Ben = ~(o, + to) 2) In the dielectric we have § = 3 - ee = 3S z 1) or ed = P= (CQ oy + p(x - h))x i) Volume polarization p, = -vee = - z o ii) On the lower surface (x = h) we have o, = Bea = - iii) On the upper surface (x = h + t) we have eee op = Eo (a, + pt) > 4.7 We use Gauss' law for D. u a) fd-da = q which gives BD af/(4ne,r7) c+ ar)t. ‘Thus 4460 -¢ + ar)) or r 46* 4.8 The differential form of Gauss' law gives V+D = p, = 0- Therefore dD/dx = 0 or D = G& and E = D/e,k = Gi/e,K where C is a constant. But we know 3 -c 4 +C ; v = -fEdx = eae fexp(-ax)dx = aoe (exp(-ax) | 0 oO 0 0 *C- (exp(-ad)-1). Thus C = -Vae,k/(I-exp(-ad)) and ae k = E = -v a & exp(-ax)/(1-exp(-ad))« 4.9 a) From Gauss’ law E = 1 5 2npe, + b) If we let © = a/p with a constant we see that E will be constant. * 4.10 a) Since there are no free charge at x = 0 then Dd) Dyef or Dyy = Dyy Which gives Dy, = 1-5, and hence Ey, = +6/€4+ Also continuity of tangential component of E jgives en 4 Eyet = Bet or Eyy = ‘ay? and Ei, = Boos Thus Eyy = -4/€,, and Ey, = 6/e,+ Thus Ey = (l/eg)( 6x - 49 + 62)¢/n2. b) Db, = 1.5% - 27 + 32 C/m®. D, projected onto yz plane gives ae Sey -2y + 3z. Thus cos®, = |-2y + 3z|/|D,| which gives 0) = 1, V3 V2.25+13 cos = 22.60 > « a“ “ ~ a Since Dy = 1.5x - l0y + 15z = -10y + 15z so ea ¥335 ¥33542 225 = 4,70 = cos 474.11 Assume there are no free charge at the interface. ‘The parallel E component is continuous thus in@, = E,sin®,. The normal D is continuous thus €f)K)cos®; = eg )K,cos@, where 6), @, are the angles the field makes with the normal to the interface in regions 1] and 2. These two equations give K, tan, = ytan6y + 4.12 a) Let us take the field and potential in the region 0 Vv = -fEed? gives -v = £)(1) + Fy(1)- Moreover the continuity ot + the perpendicular component of D gives Solving for Ey and simultaneously gives -V = E,(1 + K)/K,)- ‘Thus " " O) = HME, = XVKg/(Ky+K,). Nowd)(x = 1) = -VKo/(Ky+K)) and = Oe = 1) + Oe = DEy =Levgy + VOR, - RK WOG + Ky) + eerie pce te b) at x = 0 we have Ey = -VK,#/(K)+ K)) and D = Ke k) = ~VK)KyeQX/(K, + Kp)+ Thus the free charge = B> «DB 6, = ~VK,K,€,/(K, + Ky)» and the bound charge = 2 é o> he ° . 2 =O, = VEE (Ky = D/C, + Ke ace > ¢) The bound charge is o, = x*(P) - Py) = net p = nD) = € ae + oF > x*(Dy — €4D) - Dy + Egy) = EgV(Ky - Ky)/ (Ky + Ky)- 2 > Aa’8/r. Now Py = Dy - € 4.13 0, = Ar®, o) ) = -A( er + 6). ‘Therefore 48P) = aCe, - €:)(4r+ 4). Similarly Py = ACey ~ €9)(-O(a/r)?r + (a/r)26). Thus = 0B = -ace,- e4 £ (2%0) + C/rsiney & (sine 1 peg ECS ae rsing) 45 (sine) = Ae, - €))(20/r + coté/r). 0) = -VePy = -Ale, - €y)(a2/r3)cot®. The surface polarization oo. charge densities at r = a, and r = b are o) = Per - Pyer = a ACey - €,)8 + ACey ~ €y)y and oy = Pyet = -AB(e, - €y)(a/d)*. The free charge density at r = a is of = Dy_ - Djy 80 of = Ley(de,/dr) pig - €)(d%)/dr),24] = Abe, + €y)- 4.14 We have for a solution from Eq. 3.65 o= Cinp + D+) (Eo + F p")(A sinmd + 5 cosmé). We have oo ine cen in three regions #)(p) in p < a, (p) in a

b. We also have the boundary conditions: * : Pease x gives o3(*) Seba ~E pcos 2) At p = bwe have (hb) = &(b) and D,(b)-B = 4(h) +8 3) At p = a we have (a) = (a) oe constant Thus we write the following expansions for A c + $6, 0 = VV + (S- Sb. eV ¢ bo)cosd, %, = V ¢ & p)cosé. We now apply these conditions. Conditions two and three give gb) » e(h - a/b”) = e,(Ey + A/b?), (Bd + A/D) = (Clb = BG) V, = (A/a + ba)cosé + V, or A/a = -ba. Solving these equations simultaneously sive 49B= 28,b2/(a2(1 - k) - b%(k + 1)) A= -28,b2a?/(a(1 - k) - b%(k + 1) wEjb'(a2 - b? + k(a? + b2))/(a2(1 - k) + b*(K + 1) b) 946) = Dea. From (a) of = ~e(B ~ A/a”)cosd = -2eBcos¢ with B given in (a). 4.15 a) First consider the case €, > €)+ The electric field will exert a torque on the cylinder due to the induced polarization: the cylinder will orientate itself to create zero net torque. The two such positions are with the axis of the cylinder aligned (1) and perpendicular (2) to the field. In (1) the cylinder will be in a stable equilibrium. In (2), the cylinder will be in unstable equilibrium. For a thin disk, its axis will be perpendicular to the field in a stable equilibrium. b) For the case €, < €2, we have stable equilibrium when the axis of the cylinder is perpendicular to the field, and the axis of the thin disk parallel to the field. 4.16 a) Since the polarization of the material is permanent, then the permeability of the unpolarized material is irrelevant. The field inside the cavity is dictated by the polarization only. We can use the polarization charge ea te technique. Now pp, = -vep = 0, Oye Ee Per = Pcos@ on the surface of the cavity. ‘There are no other charges because the material is infinite. ‘The electric field corresponding to this 50charge is uniform inside the cavity and it is equal to E= P/3e,. [See Exs. 4.2, 2.17, and 3.6-] b) When the polarization is not permanent, then we have to solve the problem in an external field with the permitivity e of the material becoming relevant. We have already solved a similar problem in Ex. 4.8. In this example we have a dielectric sphere of permitivity ¢, place in an infinite medium + s of permitivity €, and with an electric field E = Ej2. Now we take ey €, = €, and E, = P/(e- €,). Thus from tq. 4.80 we have : : _ 3 7 “sphere — + 2€, ‘o” changing this result to the present case gives > eee eeere 2 + 2e Ce - €,) z = hole — Ce, 4.17 a) See Ex. 3.5 to get # = Asse 47/44? b) The potential is (x,y,z) = (1/4ne,) {a/(x - d)* + Vey. vhus yrs at) ee q(x + ay? + y+ = -(36/ax)| = oe for x=0 “n (alan) (x = d)/(Cx = a)? + y2 + 24) 91? + Ge + a CCx ay? + yh & 22)! HL -(2qd/4m) A/a? + y* + 22)3/2) ow ~Cqd/2my/(o? + d2)3/2, Now = om da where da = 2npdp, thus 0 5122 2,2 y- ad yp _2nede ge tt We do the Bre, (0 a2)? M0 Cp? + a3 integrat?on by letting u = p* + d*, hence du = 2pdp and the ae 242 ‘i ee force becomes 7—— f Stee ao | -—15% 0 0 Qu (p24 de) 0 lene d which is the same as in (a). 4.18 Because the potential has to be finite at r = 0 and r = and because the normal component of P on the surface of the sphere P, = P,cos@ involves cos@ only, we keep only the cos® terms in the expansion of the potential inside and outside %) and $5: +9) = Ajrcosé, Taking ) = ) at r= R gives A)R = B/R2, Matching Dj, at r= R, we get Sse ooo 3 (ek, + P)en = e,Eyen or -(Ay + Po/eg) = 2B,/R°- Solving for Ay and By we get Ay = P,/3e,, By = R’P,/3e,, hence 2 = R°P,cos@/3e,r° + = Pyreos8/3€,5 4.19 We use expansions for the potentials of the form of Eq. 3-65. The potential inside the cylinder @, has to be finite 52at p = 0, and the potential outside the cylinder $) has to vanish at p = *. Moreover because at p= a, the polarization is Pycoss, then we will keep the cos$ terms only. ‘Thus @) = Apcosng, 4. Atcos¢/p Taking ¢) = 6) at p = a gives Aja = A'/a. Matching Din = Yon at 2n 8 a 5 p= a gives (c.f) + Py)en = e,£yen or -(e,A + Py) = egAt/a*. Solving for A and At gives: A -P,/2e, and 2 At = -a*p,/2e,, and hence PoP a’P 4, = - go cose, ~~ Tap con Px a’p . . [pcos + gsing] 2e,e 4.20 We use the results of section 4.7 to introduce the charge q' and q” as images of q, and the charges -q' and -q" as images of the charge » as shown in the figure. &5€,Thus the images are: an image dipole $' in material €, at distance d, and an image dipole $" in material €, at the location of the real dipole. ‘The moments of the image dipoles are & -€ 2e. 2a) ‘f 2 ee re Dl De econ Pla te it % 4.21 a) , or * X+Ax ae 1 - a Co. ee Je xn ‘dx x a n xt+Ax = Xp’ x nn ecko =F] = Ge + wo" - x] = SP. The operations of averaging and differentiation are indeed interchangeable. + * > is > > 4.22 a) Since f = x°x then E = x*x and D = ek since 1 Coulomb. At one face of the differential cube, we have, + : cE = ex2x, and at the opposite face we have oro u * > e(E + (8E/dx]Ax. Thus the net flux of 54Bee = AxsAydz = 2xe AxdyAz. This net flux must be equated to the total charge inside the cube, that is p¢ AxAyAz = 2xe AxAyAz or p = 2xe+C/m. : b) pe = VoD. Thus p¢ = €8x2/ax = 2ex C/m>. 4.23 a) Because of the symmetry, the fields and the potentials > + will depend on x only. ‘Thus VeD = p and VeD = 0 give > a > « D) = (ox + b)x for x ¢d, Dy = fx for d as ae Force/Area = dF/da = ofx/2e = b°x/2e. 55CHAPTER 5 5.1 Given p = 1 g/em?, K = 4, and molecular weight = 59. Thus N = 6.02 x 1023/59.07 = 1.02 x 102? molecules/em?. From Eq. 5.5 3e we have a = 5% at + Thus a= 1.3 x 10739 coul?m/N, where Nz is the Newton, the unit of force. 5.2 a) Taking E along the z axis, then using Eq. 5.23 = + pk/kT = 1.3 x 107%. b) B= N ve = 7 x 107! C/m, B= N

2 = 7 x 1077 c/n? ©) = 2.6 x 1073 double the average dipole moment, thus we need T= 150° Ke 4) If saturation is achieved, then = p, thus p of the block is N pV = 4.95 x 1079 C/m 5-3 a) Baipore = (p/4mR?) (2e0s er + sind 8}. ‘The induced dipole is 8, = af = (pa/4ne,k>) (2cosor + sindd} b) u = Bef = - P22 cose + sine8}).—P-{2coser + sindd} a) 3 me SR 4neR = -a( 25)? (4eos?9 + sin®e) = -a(p/4negk?)7(1 + 3eos70) 4ne.R 562 3,2 7 2 : poa(2n)/(4ne R°) J (1 + 3cos“@)sinede U(8)da c) = =- Jaa 7 (4m) = -0p2/ar2e2n6 5-4 a) Let 6; and 0) be the angles py and py make with a line joining the molecules. Then we know the interaction energy = (1/413) (308) *5)(By*2) ~ By eB). Now Pitt = cos®), pyr = c088, Pit by = sin®,cos$,sin8,cosd, + sin®)sing,sind,sing, + €080,cos8, = sind,sin8,(cos(4, - $)) + cos®,cos9,. U= (-P1Py/4ne,R>)(2cos8 ,cos0.y ~ sind,sin8,(cos($, - $)))« i exp(-U/kT) Bb) P(81,8946),6)) = —————— 2 om 7 f sind 140! fsindsdey, J 4g) f ds,exp(-u/kt) ° ° ° ° c) Let $ = ¢) - ¢) 1 Qu Jind 49, Jatne, 485 ! dé U exp(-U/kT) p = w Ta fsino 40, Jsind,d0, [ do exp(-U/kT) ° ° ° Taking U/kT << 1 then exp (-U/kT) * 1 - U/kT. Let us call Denominator = D and Numerator = N of the above integral . 7 20 3 D= tear fsind de, J do(1 + pypy/(4ne RKT) x (2cos8,cos8y - eanestneacet)) Qn Note that [ cos¢do = 0, and Joind cos0d0 = 0, then D = 81 " oon 2n 0 N= fsind 49, if sind,d8, ! ds[U-U7/kT]. Note that we have ° ° already calculated the fire term in this integral, so we need to calculate the second term. Remember that U* is proportional 57to 4cos20,cos*8, - 4cos8,cos#,sin8,sin@,cos¢ + sin?@,sin26,cos”. The middle term goes to zero since fTcos sae = 0, but [™ cos*@sined® = 2/3, and fsin2ed0 = 4/3, therefore N = 4+2/3+2n + 4/3+4/3+n = 485/9 = 10n/3 so pe p? pe _ 16n/3 Pe fie ee cam, 226. oy nee ep lone RRT = 24 eR KT 0 ° 5.5 a) Using Eq. 5-17, we write py = 2a Eo/[1-(2a/4negR3)] « Solving for 2a = Po! Laney? Py + EQ). b) The molecular polarizability for all R is given by Eq. 5-17 at = 2a/[1-(2a/4negR?) ]- 5.6 Consider the figure where it shows the atoms and the oe > external field. From Eq. 5.14 § = a(E, + E") where E' is an A B —— additional electric field produced by each atom at the site of 4 for 9 = 90° we get E'(8 = 90°) = the other. Using Eq. 2+ -p2/AnegR?. Thus & = -aE + BY) /4ne, 3 Solving for £', and hence for p and & gives 58* + fr = - —_,_—_"______ , p= 4ne R[L + a/dne R>] 1 + a/4neR ° 0 3. at = la/(l + a/ane,k?) 5.7 Using Eqs. 5.9 and 5.21 we get a +a’ ta = L : ane + Po(coth n=) where n= pyE/kT. At high temperatures ue have a = 4neyx? + po2/3KT. ‘The average dipole moment is

= Chnegr? + ng2/3kT) EG 5.8 using Hq. 5.36 we write K - 1 = 3/6(i-T,). Since K is very large, then K - 1 = K, and we have I/k = 4(1 «+ The points ra) (i/", T): (0035, 133) and (.0105, 153) define the Line and we met 6 = 1.05 x 1079 and Ty = 123°K. 5.9 the ferroelectric condition is given by Eq. 5.34: Na/3e, = 1. sut from Hq. 5.5 we have ha/3e, “= 1V/(K + 2) = 290090 which is different from 1; thus it is not ferroelectric. ror tne liquid we have ha/3e, = .367 which is not ferroelectric. In this case 5.10 In Eq. 4.4 we nave o, = - (K-1) of/K. For ferroelectric naterials x >> 1, hence o, = o,- 59CHAPTER 6 6.1 See example 6.1. 6.2 (4) = Aexp(-ar 2 2 a) From Poisson equation we have V#= -p/eor p= ~e Vo. Note @ has only radial dependence, thus Peed ee. tor Gb ertoun san shaceep(-at?> (net). Thus the density is 0 = 3Ae,ar(4-3ar)exp(-ar’). b) v= (1/2) fotdv = (3/2) Ae a frexp(-2ar?)(4-3ar>) dv = 4m (3/2) A2ega fexp(-2ar9)(4-3or yr Jar. We know fx"lexp(-x) dx = T(n) with T(n) the gamma function 0) T(nt1) = n(n), and T(n+l) = n!, if n is a positive integer. 3 ang put the integral in the form of the f integral Ane) A’ we get V * ras 1(4/3). oe ae caer ae 6.3 We use Eq. 6.9: 0) = us P,Q, where the coefficients of ik potential are: Pj; = eae Pjq = Waneyt for j # k since the spheres are identical and they are at the same distance from each other. Grounding sphere | changes the charge on it but does not affect the charge on the other two. Thus 6) = 0 = 60(qy/a) + (q/2) + (g/t). Solving for q) gives q) = -2qa/2. Now, ungrounding sphere 1 and grounding 2 gives %) = 0 = (q)/&) + 2a e (ag/a) + (q/2) which gives qy = - = (1 - 24). Now ungrounding 2, and grounding 3 gives % = 0 = (a\/2) + a2/%) + (q3/a)- This pol ves| for qo 2 se i =) te i e ie to allowing the charge on each sphere to adjust to the external “until equilibrium” refers potential. You do not have to keep grounding spheres. 6-4 Since the force between the spheres is given by F = -8U/ar where U is the electrostatic energy of the system, and r is the distance between then, then we need to calculate U as a function of r. We use the coefficients of potential to solve the problem. We use Eqs. 6.8 and 6 olde PLO P 4, ; HOt Pyy!ys P50, + Poot, where My te oe Dil 202 j Py, = P92» and Py) = Pyy because of the symmetry. Now when © = qy, % = 0, and ) = V. These equations give V = Py,9q and y = Po14, = Pyyqy vinich can be solved for Py, = V/qy+ ‘the 2 energy of the systen is 1 = 1/0] Py,, and hence from the force ap, rs laze Yo aq (=) we get dP),/dr = 2F/ay-+ In the second F operation, we have *) = Pyyqy + Pyydy, and V = Pyyqy + Pyyay which gives Py) = Y(qy - a2)/aq, and hence Uy =p Py (ay + 49) + Pyya,4y+ From the force between the spneres we can calculate aP,o/dr as follows: ap 1 + 03) a 614Py/dr = -F(2qf + 43)/ 439+ Now that we know Pj), Pyy and their derivatives with r we can calculate the properties of the last operation. In the last operation sphere 1 is grounded, thus %, = U, while the charge on the second does not change, therefore 0) = Pi Q + Pid, oF Qy = -a9€ay - 47)/q,+ The force between them can be calculated from F = ~U3/r where Uz = My Py (QT + 43) + Py2Qazy OF F = qp(2ay - 43) Flaps 6.5 a) See Ex. 6.8. By symmetry Py, = P97 = P33 = Pays Now =i 4ne a °° % = Pij4+ thus Py] = I/énegae By Symmetry: Py2 = Pa] = Pyy = Pay = Pay = Pay = W4negt = Pog = P32+ Also by symmetry: Pig = P31 = Pan = Pou = MW4ne,(v28). c) (1) When we connect 1 and 2 then the charge merely splits because of symmetry, that is q) = 4/2, q) = @/2+ (2) When we connect 1 and 3, then because spheres 1 and 3 are symmetric with respect to sphere 2, the charge splits equally q, = 4/4, and a3 = 4/4. (3) Similarly there is symmetry in the last operation and we get q, = 4/8 and q, = q/8. Thus the final charges are q/8, 4/2, q/2, q/8- 6.6 a) See example 6.6. b) From above we have C = 4ne,ab/(b - a). Now if b - a= d << a,b then C* 4ne,a’/d = Ae,/d with & = 4nd the surface area of the sphere. 626.7 Let us call the fields in the regions r < a, a< 4 + + . b c as Ey, Ej, Ez and E, respectively. coe a hg 20 a) é- 0 8 = Calaneg Dee. oe, (Q/4ne se bE b) 9 = -[Bedk = -[B, +aF - (E,-af - ff, sat ° ° c b =—2@,@ (lil free ger a! = = yy ce) C=QVe= -p) d) We know that capacitors in series add inversely ee : Ted feo ee eee eR ane le ce > 1 GGG i = 4nec and ° 6-8 a) U=Ny wv? = Y(egald) v2- b) ‘Treat the system as two capacitors in series. Let the shect of metal be a distance b away from one side of the capacitor, then 1/Cp = 1/C, + 1/Cy with C, = Aeg/b and Cy = Aeg/(d - b - a). ‘Thus Cp = Aeg/(d - a). Now during the operation the charge stays constant. ‘Thus UI = 07/20, - Q2/2Cg- Using G, = Acg/d and Cp = Aeg/(d-a) gives U = Aega v2/d?. ©) Since y= constant, then Q = GjVy = GV_ which gives Ve = GV/Ce = Vid - ad/de 6.9 We use the results of Ex. 3.5. g- a l-lL),—_ 4 _¢, = ero ee ane (R? - 1) r 2 ond eal lcos8 to first order in 6. Now it is important to realize that this result was derived from the boundary condition 4| 0. (Indeed, if you sphere #2 = 63plug r = Ry + 6 cos® in the above result, after much algebra you will get 6) = 0). Therefore we have oe 7 = rele 7) 82 = | _ a - = Ue ere ee ee Cour) 4ne oR Ry c=4 = — 21 4 which is the capacitance of a concentric ay RD Ry spherical conductor. ‘hus, there is no change in the capacitance in the non-concentric case to first order in 6. For a larger 6, we have to include higher order terms to $. The next term is the P(cos6) dependence which would be proportional to 62. In this case the potential difference will depend on 6”, thus giving a change in C. THe next correction is found to be Re RZ ie ceases 2) ee ee (Ry - R)"(R - RY) 6-10 a) W = Qv/2 = Q2/8neR. + | * o b) Since E = q/4ner* r and D = q/4mr* r, then 2 wel [EB r -2a which is the same as Oe 8 4) The energy goes into mechanical work. Consider the sphere =z (AR/RR') « at a radius r. The force acting on an element of surface area + : is dF = otda n/2e where o = Q/4r7. The total pressure acting 2 on the surface is then F, = [dF = (1/2e) ecu xan? = 2, (ame?) ber 64R' 2R 2 7 ere .2 dil, The mechanical work is Wy ee wae : 2” Bre Qu: 6-11 a) ‘he large conducting sphere is grounded but has charge Qy> thus 6) = P),Q) + PyoQ) = 0 where Pj, is the “self” coefficient of potential 1/4ne,R;- Due to symmetry we have Piy = Poy = I/4negd. Now v thus P))Q) = -Py2Q) which gives 0 = -qR\/d. b) If the sphere is then neutral, Q, = 0, and hence 4) = PyQ) + PyyQy = a/4negd. ©) From Ex. 3.52, the image method gives the same results as above. 6.12 ‘he field at r inside the sphere is found to be equivalent to all charge in the region inside a sphere of radius r, all at the origin. This is only true for spherical symmetry (no 6, is E a( &, consider all charge at the origin. ‘Thus for r > kK. Wow ek dv or U = 52 fal(r) amar + ° 20K ° 9 wr) amr? ar _—— > ee : te Roo Hboea ee ° a Ry, @ yl ey 5” Brey“ K” 20nEDRIareeeiee ee enneneiaeeeemieenseteneademeneamemeemasmaaaniamnesneteeenmmemememeteemeamninmmintnmemenmmenenenmenatennenesa nent 3 i or of. Thus 3 w= (1/2) f e.8? dv = 180 WaPegan f et ar 3 6613 6 = Ae”, B= -V0 = 3A0r7e™ 4ane ae °. = T (4/3) as in 6.2. 24/341/3 6-14 1) In the first case the outer shell is neutral. because the outershell is neutral; then from Gauss' law there will be qi 4neor sphere.° With charge + q on the inner sphere (outside surface), E- around each sphere, including the outside outer a charge -q will appear on the inner surface of the outer sphere and +q on the outer surface of the outer sphere. ‘Thus from Eq. 6-8 the energy of the system is Uspe™ (y+ 9h)/Bne,F, independent of r,- 2) When the outer spheres are connected, we find because of symmetry, that the charges on the very outer surfaces of the outer spheres equalize, while the charges on their inner surfaces or the charges on the inner sphere are unaffected. ‘Thus a} = 4) = q' + (ay + 42)/2, corresponding to a flow of charge #(qy) - 42)/2- ed lel hae Now Urinar = (ay + 45)/(8ne,) ee - 7 + ner, 0 2 2 (q, - 4) 2 de lene ry 4 bre,ry change comes from the charge that flowed. + Thus the energy aU = - 6.15 We are given that # = -(6¢/m) V,- 666v. a) E- vee LER § = 2 Fu = (1/2) gk? = 1BegVQ2/o2nt 2,2 1/6 1. .6 b) U= fudv = 18e,v"/n" f de [dz [do/p = 3€Q( tn6)V.2/n. ° eee 4,4. qd 1 oe nee Oe) ey Gace ane ee o 21 But ry) = tp) = d, thus U = q)q9/4ne,4- b) Let us take the z axis to be along the line joining the e qt CMe goa ae | charges U = 52 ferav. pot # -=—L— 4 —2 4ner a 4 ve RIGS * r lowe The first two terms constitute the self energy, that is the energy needed to assemble the charges themselves. This is a divergent quantity. The last term, on the other hand, constitute the interaction energy. This integral can be evaluated easily if one recognizes that dv = dadr where aa da = r’sinededs and da = r da. (See Eqs. 1.22 and 1.27). Thus Midi pide + Now the area integral is U > ane lore, rate (es just 4n singe the integrand is df. (See Eqs. 1.23 and 2.23.) 444 444, oe Ode i Thus Uige “Teer | SF Tee” Thus the difference between the two methods is the self energy and the minus sign. We have a sign difference because what we just calculated is the energy stored. 672 2 : : 6.17 V = 3x2 + 4y?, E = -W = (6xx + Byy). U = fudv where u = € iT 2/2 = e((18x" day”). Thus 1 u=e f dz f dx f ay (1824 32y") = 50 €./3J 200 Oo : 6.18 a) First we use U = (1/2) fo @a where o = surface charge s density = Q/4mR*, but Q = CV = 4ne,RV, thus o = €,V4/R and therefore U = 2nReyVa- + b) Now for r < R we have E = 0, and for r > R we have z 2 bag VoR/r + Thus 22 coeds eee 2 ° 2 U = (€9/2) fevdv = (e,/2) f dv = 2nRe VO. © ‘ovo re 6-19 a) From Problem 3.2 we have @=% 6. b) From Problem 3.2 we have at plates $= 0, and $= 8. Q = (-Vegh/B)£n py/p,, 9 = -Veo/B, Q= (Vegh/A)&n 9/0) 5 o Veg/ Bo. oe ©) c= 2 = (ne,/ayen(p,/p,)+ 4) U =¥ cv? = (ne,v2/28)2n(p,/p))+ Consider a virtual rotation of one plate in $ direction, then > au» 22 > T= 3g 2 = (he v"/28")&n( 09/0) 2 6-20 From the results of problem 4.6 which were arrived at using Gauss’ law, we find that the electric field in the slab + using |F| = [ Edq where dq = Apdx, E = [o, + p(x-h)]/e, thus : htt \F| =" f + p(x - h))dx ~ A Pls 686.21 a) The stable configuration is when the rod is aligned with the external field. The unstable one is when the rod is perpendicular to the field. A force arises because a dipole moment is induced in each of the conductors which exerts a force on the other dipole. From Ex. 3.7, the potential field and the moment due to the presence of one sphere are: cosé, f= a Qeoser + sineé}, p= 4ne.a’e 2. 4ne re Because £ >> a, then ofe takes this field to represent the field near each of the two spheres. The force of one dipole on each sphere (we are only concerned with the force in the 8 direction) aE 2 2 fg = (pe oe ee in? is Fy = (p-V)Eg = pg = 7b —2sin2e). ° 2ve_f © re i The total torque is just: |1| = ay (sin26). Que ° > c¢) The work necessary is just W = [t-d8 or 2 4/2 w=—2_ ff sinzede = -8ne a°e2/ 9? . 07 Fo anee o 6.22 We use the notations of Fig. 3.13. ‘The potential energy + > of the dipole is U = -p+E where E is the electric field of its image. Now > Pp a a P “ a E = 2 —,0cosér + sine) = —°——{(2cos6z - sindx). But 4ne (24) ane, (2d) P = py cos0z + pysindx. Thus 692 2 Po 2 2 Po 2 = = —2— (200878 - sin’) = - ——2—(3c08°@ - 1). Thus 4ne,(24) ane, (2a) the energy needed is W = 6.23 a) We replace the sphere by an image charge q' located at 2 Z where q' = =a, and Z, = a as given by the method of images. Thus the force F between the two cite is oer ee ee 7 P3 ane (e - EP ane (r?= by? Abas se or (rt py? 2 ogee Bre (r= v7) b) Yes as position of real charge changes, the induced charge on the sphere changes. In other words, electrons flow to or from ground to change the induced charge on the sphere. c) If we isolate the sphere and place a charge Q on it, it is like placing a charge Q - q' at the center of the sphere (see Eq. 3-116). Thus the force becomes: ba F -bq?r a(Q - 4") -bq2r aa + = Daan ee oaaaea ae 4ne (r2- b”) ner? 4ne (r b*) Aner ° ° ° ° r 2 © + -bi b we f Fear = “, 3 f pare ® Bxe (r= b”) r Se wie (8+ oy) 81e, (ee b’) ane, r 704) The difference is clearly Aw = —4 @+ This is just ane 0 the work done in moving the point charge against the charged sphere. 6.24 The energy density is given by Eq. 6.61: u =l/) E+D = + +e Mp E*(egE + P). From problem 4.19 the potential and the field inside and outside the cylinder $) = (Po/2e,)pcosé, = -Pyx/2e, oe Pp, pe . - Ono 2 oo 9 = Fer F CO8bs By = GE —Xcosdo + sind}). Thus the energy 2e, 2 cee density outside the cylinder is ug = P3p3/Bego*. For the region inside the cylinder we have uy = -P2/2e,- 71CHAPTER 7 7.1 Using Eq. 7.4, J = p<¥>, we get I/A = ne<#>. Thus tA I <0 = ag tel cals. + 7.2 Now equipotentials are spherical shells, and the field E + and the current J are radial. Thus using d@ = dr, and A = 4nr¢ we get: bar lel Re : i 2" Gao ‘a ao 4or 7.3 R= f—t—+1 oe a 4nr?g =e Gara 1 hs L R= =) b b dp el aaa Yb, Ta R= f Soee = Trae 8 | = Teave 2G) a % c a c >__ Be oe i ap 7.5 a) Using VeJ = - 52 and J = o£, we get o.VeE = - FP Integrating over the volume of the capacitor we get + a aoe do, 9, [VeBdv Scop (ede. A > + By the divergence theorem we have [V+Edv = fEeda, and a a, fda = -. 72Oi ee cQe ee dO ce c Ke A dt t Ke * 0. ° ~(o¢/Keg)t The solution is Q(t) = Qe . b) The rate at which heat is generated is dt a 2 9,0 fU-Edv = folk dv = ——_ v (Kea) o,40? 2 -2(0,/Ke,)t ed Total heat = —“2- [fe Ode =>2—. originally, the 2 2AKe, (ke,)"A 0 0 energy is Q7/2C = Q/2Ke,(A/d). ‘Thus the total heat produced is equal to the original energy. = 3.8 x 107*4sec. 7.6 We will use the boundary conditions: the normal J is continuous J,; = Jy2» and the tangential E is continuous: or I 3 tl = ata : t= Ee bat gy = Jy0088), Jug = 3700889, Jey = Jysind, = fe adc Jqitan®, Jeg = Jysin®y = Jqgtan®). Then using Jn) = Jn we get Jq1tan)/o, 6 = Jqytan®,/9 , or q otan®, = g ,tand 7.7 a) From &q. 7.38 with no external electromotive fields we have: Ve(o,V@) = (Vo,)+(¥8) + o,(976). Note that the current of the E field and the potential depend only on x. Using 732 vo = 2% Fo = ax and (v0) = 22, we get ax? "% 2 dx’ 2 afatx) P44 a Heo, 2 ax dx dE, de, ao, ce ax : oo -*+e = Eee b) BE ax? Thus (a + x) a B= Oor Ee Sax whieh integrates to ink, = -£n(a + x) + const or Ey = A/(a + x)+ ° a 2a 6-6 =-f pax = af af Se am(22) = atn2 a 8 aces x a %, 7 oe Qe = oAx, then f = Ja” = aAa’x. 5 c) Since J oe Aine nz T aaa aa’ 7.8 This problem is similar to Ex. 7-11 except that both spheres are completely immersed in the infinite medium. We first use Eqs. 7.58-7.59. Now Eqs. 7.60-7-@ become 1 L al 1 ee a c 72 c 747.9 This problem can be solved by introducing two image currents across the boundary such that the J, across the boundary vanishes. Thus the image of I, = I is taken to equal to I and the image 1, = -I is taken to be equal to -I. Now using a similar procedure to that of Ex. 7.11 we write 1 i ama ) © Ve + a? 7.10 From Eq. 7.36 we have €/Ro,. 7.11 Using the expression & = [d£/o,A given in problem 7.2, we can find the resistance between the plates dz. d Bl aoe ae c c Alternatively we can use Eq. 7.36 RC = e/g, to arrive at the same result. 7.12 Using Eq. 7.72 we have o, = nq?t/m. Thus t= ogm/ng? = 241 x 107! !sec. 7.13 The potential of the ith conductor is given by Eq. 7.57: 75= i Ril: But the heat generated due to the current leaving the ith conductor is Q = 4,Ij- Thus the total heat is You =,5. Ray lylae pot tie Raktite 7.14 We use the equations developed for the circuit of Fig. 7.16: -IgR3 + 14k, = 0, IyR) - TR) = 0 and -e) - IjRy - 14k, = 0. Along with these equations we have Ty = 15 + Ig, Ty = Ig - Tye Ty = Uy t+ Ty and Ty = + Ip Solving these equations for I, we get: =e, (R)/Ry — R,/Ry) ee 5 CT # RAPRAVR, RY + R,CR/R, + RRA) Thus I, = 0 when R)/R, = R,/Ry> 7.15 Differentiate I, with respect to Rj using the result of problem 7.14. Thus 41. (bottom)d(top)/dk, ~ (top) d(bottom)/dR, 3 (bottom)? balance Since the top = 0 at balance, then d(top)/dR, Ce 23 (bottom) 7.16 a) When we take €) as dead, we replace it by a resistanceless wire. Thus the resultant resistance is RK = 1/29, and the power delivered is P) = ef/R = 200 Watt. b) We replace €, by a simple wire, thus one of the 12 1 76resistances gets shorted giving R = 19 and hence Py = €3/K = 100 Watt. ¢) The actual currents in the circuit can be determined from Kirchhoff's law. Assume current qy and Ty to flow clockwise in the loop on the left and on the right respectively. Thus 10 - qy et Iy) = 0 and 10 - yy i 21, = 0. Subtracting the equations gives 1) = 0, and I; = 10 Amps. Thus total power delivered is €,1) + €91) = 100 Watt which is different from the sum of the individual ones in a and b. 7.17 a) Using Kirchhoff's law for the series case we get Jeg - UR,I - RI = 0 or ney ~ nRy - RI = 0 giving I = neg/(R + nRy)- Teeto) b) We use Kirchhoff's law. Assume current I), Ij, + flow in the loops as shown. Set the equations for each loop, for example we have (-21; + 1,)Ry = 0, (-21, + I, + I))Ry = 0, (-21, + 1, + I,)r = 0, (Ry tReet Solve for I,- Take n = 4 as an example and then ger> A “Aw 777.18 We use Kirchhoff's laws to set the equations for circulating currents similar to the analysis of Fig. 7.16. We determine the current I through the source €, then R = e/I. This gives R = 3r/2 for n = 2, R = 13r/7 for n = 3 and R = 47r/22 for n = 4& 78CHAPTER 8 8-1 The equation of motion for a particle of mass m and charge > + q in a region with & and B# 0 is given by Eq. 8.2 oe ae F = q(E + ¥% x B) = ma. Note that the magnetic force + x B is perpendicular to %, and hence for orbital radius r we have mv2/r = qvB sin. The smallest B needed is for sin® = 1; thus B = mv/qr which gives B = 5.69 x 102 Tesla. + > 8.2 To get B we use Ampere's law § Bed2 = uw, [Jeda. For p < p, ° we get: oe § Bead = B20) =u, f fe’? prdptaet oo 2mu, 4720! Beet _ To e Ae ee eer ery B= a5 Sy aD Toll - 20e e P), For p > P, we have ° ' Qn u, -2p -2p B(2mp) =p © P'prapt fdb=i2tl-e °-2e I. o ip ° ou 8.3 We will first find the field produced by the sheet. The procedure is the same used in Ex. 8.6- We calculate contributions from opposite sides of the sheet, then total up to > - give a horizontal component. Hence B = -Bx eeW/2 uk, fe h B= [(dB, + dBy)cos6 = 2f 3 yay, 7, eh Ge ney) ee ney ee wok oh w/2 uk oo. oo” 1 Thx oXo | -1pW_ys : gear (FF) | = - tan Ge Now F=f x By Thus 3 Ink fe - 2 © tan ‘Gy. It is an attractive force. When W becomes large, then F/2 = -Iu,Kgy/2+ = 100u, 4p 2oe + > a o oO mp oO n - 8.4 a) B a z sin 55 te Vx B= ule 6 04 z In cylindrical coordinates: vx B-4/2 2% - 2 oe “p |8e ery az 0 e(B,) 0 2 100y 4p. 200, ao oe me. vx o 213 = sin 55 3 008 1) ‘° ‘0 J = 100 sin 22 205 Py b) I= 100 f sin oe da, da = 2npdp, I = 2007 J o sin zz = ‘° 1 cd e mo 40, I aay sin 2, - aT 608 20! | = room) 2p. 0 ‘ lot OB a8. soe bei 1 _¢ 8.5 Using B 54 and 4-8 2 (op,) +4 3 + » and noting that B only has a $ component and p dependence, so 7 = 0 and : Ae an, 3B, 2B, x Be 5b age - a ae) + GF 52) + i conp - 32) 80The only term we can use is 5 503,) where pB, = u,/2 and Flu Ven) = 0. Now, check p = 0 carefully: a - pap PBy) = For p + 0 we have (* - ©) which is infinite which indicate the presence of a filamentary current at p = 0. 8.6 Let's consider the contribution to the force on the ee diagonal wire due to current 1). &, = 52% = Z- The force on the 1 > Bax > oe + 4 a diagonal wire is given by F = 1, far x Bwhere d& = xdx + ydy- But y = x +c, then dy = dx, and d2 = dxx + dxy. Thus i. te Ve ee te ~ Hor'2 , (dxx + dxy) + io 2%, dxs 8 Fa HSER 8p Gok Saws) 4g Hoe OF Be - 5) x x ° ° a Hotty ar 8-7 From Ex. 8.9 we find that the vector potential of a filamentary current along z is Ay = -vgI@ tn p,/2m. For a filamentary current pointing along -z we have Ay = ugl2 tn py/2n where p, and py are measured from the currents. Thus at any point in space oe A= Ay + Ay = Uol Bn (p9/p1)/2m 81x y 2 7 7 18 a a - 88 a) Aps Bx: VeAs ie oF ORL -By 0 0 a ae aoa a Wh, = Gok + ES + Fy Cow = 0 x y Zz . 7 Pale a a * Ay = Bry: then V x A= |o~ yy S| = 2 0 BK 0 + > > and VeAy = 0. 3 = -'y Ex Br We may choose the B-field along any axis without any loss of generality. Take B = Bz. thus My Bx B= -YVy (By)x +My (Bx)¥- Hence o. > oe > V x Ag = - Yy(-B - B)z = Bz. Moreover we have VrA3 = + + b) To show that A) - Ay is the gradient of a function y, it > suffices to show that Vx (Ay - Ay) = 0. Taking Ay - Ay = -B(yx + xy), then x y 2 i: -ib-lR & B-- = vx (h - 4) = | oO a B+ B= 0. Hence —By -Bx 0 > > Ay - Ag = Vy. It is easily seen that » = -Bxy + constant. +> 8.9 a) Ampere's law gives §B-d% = u,l, where c is a circle of c radius p around the wire. Thus we get B(2mp) = ul, or Beu oli d/ 20. r Bot ca DF = [Bead = J r Geo a0 Ot aoe anc r a> c) We use Eq. gre: [irda = Fe a where c is the loop and s is the surface bounded by c, and A is in the 2 direction, and 82depends only on p. Thus F = [A(r) - A(r,)]t, and hence > > oe r A(t) - Cry) = 2 en 22 > > + a d) The force on ab is dF = Indt x B= I,d2B(-z) or ‘ol aonie2 4 rs an(*)2 The force on be is dF = Ind% x B= Ipd£B(-y) or ec ek wt, t= jn, f Pha, -- Tolle. 5 4 dur 2ur 8.10 a) The magnetic field in the solenoid is B= ujInz. The flux F = BA = uoInxy. + > > a b) Using fAed2 = F and taking Ay = A(x)y, we get Ay(x)y - A) (Oy = Ay(x)y = ugInxy which gives Aj(x) = ujInx and + - A, = uyInxy. €) Take Ay = Ay(y)x+ Then we have ~Ay(y)x + A)(0)x = ~Agly)x = > : ugInxy which gives A)(y) = -ujIny and Ay = -yyInyx. + as ee d) Take Aj = (Ay + Ay)/2 = -ugIn(yx - xy)/2. Check > VeA3 = 0 = -uyIn(ay/ax - dx/ay) = 0. It is clear that the potentials are as plotted. > > 7 8.11 We expect A to be along z. Thus V x A= -$9A,/2p. Now eres > using B= oan ¢ we get from V x A= B: -8A,/8p = ujI/2mp which . wlan integrates to A = -_ya 838.12 a) From example 8-12 and taking the point of observation ul 0 2r at the center of the loop that is z = 0, we get: B= z where r is the radius of the loop. b) Integrate loops from a to b, dI = INdr ae uy2NT b ae. u2NI boy 2NT 2 v oe a wn r| = 5 an(2) a 2, thus Check for b * a: on(2) = en(l + B= uy2NI(b - a)/2a. Now check for b- a << ar b-a= 7 ; thus B = uy2I/2a. a 8.13 From symmetry we have for each side B= Bz. We use V I = ee ab de x = | For che side show: & = -xx - (a/2)y, + s é dQ = -xdx, thus ut a/2 > ‘0 adx B eames Gear Using x = au/2 and dx = adu/2 side Bry) Ge ae/ay le . uote } du v2 wiz side gina. (ue ee tna 2 av yw Iz Bequare ~ “Beige "wa * 848-14 From problem 8.12 we have for a loop of radius b: acy 3, = zp 2+ B for 2 wires from 0 to @ is the same as a wire MeL from -= to *, B= Zap 2+ Thus the total field is patois, tots _tolt ab 2nb > 8.15 Consider a band of loops at @ in the range d@, thus the current in Rd®@ x IN/tR = INd®/m. From Ex. 8.12 the field on the axis of a loop of radius r and current I is dB = ugIr?/26? where & is the distance of any current element from the point of observation. Taking r 1 Rsin®, & = R and integrate we get UNL ow uot B= fat sin?o(Nd®) = ee! sin’ed0 3 7 wIN 1 * : - qGsin 2), Be Ze 8.16 Consider the result of Ex. 8-15 about the B field on the axis of a spinning disc: ae 2 + aa r - B= > uput{ 2\zI} where t is the thickness of the disc, and r is its radius. We consider the above spinning sphere as a collection of spinning discs with variable radii r and thickness t = dz. Since we want the field at the center of the sphere, then for all of the discs we have r2 + 22 = R?. Thus 85> 2 a B= upulreede- 27} | 2 = ° 8.17 From Ex. 8.12, the field on the axis of a current carrying loop: 2 . ugIR : Ec romoumrara/2 a 2(2" + RY If we venture slightly from the 2 axis, we expect > > + B= B(p,z), i-e. we will introduce a B field in the p 3B. oe 12 z direction. But V+B = 0 gives > 5> (pB,) + g>= = 0 or OB. i (0B,)+ If we are not too far off the axis, then: Bo vat hence ot eo therefore z + Rye” Be oe Rye’ A Bu IR’ Su IR?2 5 3p (PB,) ae °B, = ST ie pde = 3y TR 2p 3y,IR 2p c eRe co Si teeae a But at p = 0 we Buy IR20 require B, = 0, thus C= 0 and B, = Gazal : 3ujTR'2p ee Moe) et ahs8-18 From example 8.12, the field on the z axis, from a single Qe IR 2 2)3/2 ° coil is B(z) = z 2(2" + R°) Now Ben Pree? + RP 2(- By(reye + (22 + R)]2. For coils at z = + R/2 we add the two terms. Since 3B/az is an odd function of z, then J 3a + Now R ood pene aR oo: 2002 4 p2y-7/2 Re = z te FZ WyIR (2" + B 5 +é +R)]2=0 2 uot ge 2(R' 2)-3/2- _ Yo /5)-3/25 B= wor’ F +R) z= Ze + = ie ee es 8.19 a) B= -VO, (Bo/tg) FX + 2 + H+ We must make sure > > that V-B = 0 and V x B = 0 for the potential to be reasonable. Both are clearly satisfied. os i fa ce b) F = V(ieB) as given by Eq. 8.107. Now it = na2Iz = na*fez where f is the frequency of revolution which is equal to L v/2na: The speed for a circular orbit is v = (e2/4nega) /2. : Bam : Bim, Tus Fe - va +) = - ke ° ° 8.20 From Eq. 8.97 we get for a small current loop: fee a + ANG; + + At large distances & = (2 - 2") = 2. Thus 87+ Using 2 x f= $sin® we get: > 1 DY 1 $. + 8.21 a) The magnetic moment for each turn is mi, = IA where A= R2z. Hence fl = NInR@z. = b) Force = V(m+B) as given in Eq. 8.107 B s_ oo F = v(Nimr'2 + 7 (x + y)) = 0- c) Torque = mxB as given in Eq. 8.109 NInR7B : Nimk?B ee 2 2 8.22 a) The current is taken counter clockwise. Force due to > os a magnetic field on a current I is given by F = I fdg x B, where e c is the path of current. We apply this to each side: 0 > > : + obviously Fag = 0. For AB, dé = dy(-y) and Fag I|ABIB (-¥ x %) so Fag = +1-98 2Nt. Now Fug = 1B(0.3/0.5)y x x = 1.982Nt. See that net force = 0 on a loop in a uniform magnetic field. b) Dipole moment: for a current loop m = TAs] for counter” clockwise sense of current, looking from top, points upward: so it = 6 xp (0.3)(0.4)2 = .362Aem2. The torque on a = + : dipole = t = ih x B= .396yNem 888-23 a) m = superposition of 2 dipole moments = it, + it, > Ina2 a + n, 2 f= Eh + 8) = 10D + 2) mpm? b) Since 732 + 4? meters >> 1 cm, the loops are effectively dipoles with the net dipole moment given in (a). The field due to a dipole is B = [3(-mt - m]. Since the dipole is at amr origin, f = 3x + 4y, and r = 5m then a i. Ba 2 fo ED Ga wo + HEED - 4ur (wx 1074 x + 2)} B- Fx og - x + 385 - 21 teste 5 > > bg 8.24 Since t = it x B then || = tl = 1.96 x 10724 amps m2. Now I = 52 and I =—"), then w - 7, - 2 x 1016 secm!, ar er 8.25 a) By example 8.19, the magnetic dipole moment of a = charged spinning disc of thickness t is m= ee . Now, we 2 7 replace t by dz and r2 by R* - 2%, then R = 28 fon? - 2?)2ae ° * l®lsphere 5 4npuR” > 4 ee 4 es (R° = 2R°2” + 2 )dz or "ephere™ 15 _ mew oe? one "sphere 89b) Now, consider the sphere as a “collection” of current loops with a magnetic moment dm = AdI where A is the area of the loop, and dI is the current in the loop. Now dI = dq/T, o(2nr)Rdev _ ee oRvd 8. dq = o(2mr)Rd6, and T = 2nr/v. Thus dI = Taking v = wr, dl = oRwrd®. Now r = Rsin®, thus dT = oR2usined®, and dm = (1r2)(oR@usin0d®) = onR4usin>ed0, 1 1 m = onR4y fsin3ed0 = -onR4w [dcos@(1 - cos*8), a a 1 & = omR4u{-1/3(cosd)(sin2e + 2)|}z = 4 our? 6 90CHAPTER 9 9.1 m= (ayy? + by) ¥ + age? x * a) The volume pole density is pm = -Vem = -2ayy - 2ayx. The surface density is o, = men, but n = Vf/|VE| where f is the P22 2 e equation of the sphere = x7 + y? 422 = r%. Using VE = 2xx + 2yy + 2ez, |VE| = 2r, then n = xx + yy + z2/r and On = (U/e)fagx? + (ayy? + dyy- x y z ~ = 3 a 3 b) Jno Vxm= | 5 y | 70 2 2 ax (ayy+ bj) 0 x z Soe 7 2, Komxn= fax” (ay b)) 0 i x 2 r r r Ky = x(z/r)(ayy? + by) ~ ¥le/r)agx? + 2E(y/r) agx? - (xir)(ayy? +b]. a 9.2 This is the magnetic version of problem 4.2. Given M = My where M, = constant. > A a) pq = -¥eM = 0. The surface density 9, = Men is Mery = 0 on top side, -M,z*y = 0 on the bottom side, My*y = M, on front face, and M,y*P = Mysind on curved surface. 91L/2 Rk Qn L/2 4, - if e,dv + f oda = J dz f dx Mi + J Rao f d2M sing v s -L/2 -R 2LM, - 2LRM, = 0. Note: The region is 1 < 6 < 2m by convention. c) No magnetic monopoles have been detected, although there are theories that predict them, and a Stanford experiment may have detected one. We do not know for certain that they do exist. Ro2n + : : : 4) m= fip(%)dv + fEo(E)da = fap f Rdg(op +My 2) 0b Pe v s oO 7 a R an L/2 R Jao J ragcod - EZ) PEM + fae faced + xx of Ace 5 o om ly -L/2 -R 7 L/2 an oe 2 Jo dz f Rdo(zz + RB)oTYS = 0 +0404 LRM f (x cosd + el/2 |e [ a e aR’LM, y sing)M singdd = TT = ore pe 9.3 a) Magnetic pole densities 9, = -VeM = 0. The surface density on flat faces is o, = Men = 0. For circular edge, we have fn = cos® x + sin® y, thus o, = M, cosd. oo b) The magnetic current density is J, = Vx M= 0. ‘he surface current density on the flat surfaces are K, = Mgx * 2 = -Myy on > - : a the upper surface, and K, = Myx x (-2) = Myy on the lower surface. The surface current density for circular edge is Ky = Mot * (cos8x + sindy) = M, sindz. — ioe ©) B=—2 fBida +m. Now for R>> T; xe 0 92Qn M 8 uy, 20 Mcos % » -Rp, thus B = T R“(cos@x + sin@y)d@ fdz + uM cd eg ‘ uTM x 2m uo TH oes 0 0 2 0 fd ogee Mae ae Be Poo + uN. Since Bu,M = H, > + then H = ~UGTM/4R. 9.4 We use the method of magnetic pole density to find %, along axis. Now p, = 0 and o, = M, at z= 2, and -M, at z= 0. Thus 1 ce My an a a a d = aq J er aa = Zel- f ao{ f PR + f ———} ° Ott 6 eae +22 att (2 - 2)? 7 Yo} @ - 2)? " — e = Sa + 2? - vee (2 ~~ Jz] + [2 - el] 9.5 Let us consider the potential inside the sphere first: Fy cet m 1 b) From 9.16 @ = 7 if a J <7 da where p, = -VeM, we can show that p, = -24)(r)/r - My(r)- Moreover o, = M)(r) + ae Myer. We will evaluate the volume integral which we will call I, first. Using Eq. 2.59 we get: c R 1a 2 = {e Jeg dr + [egt dr}. Substituting for p, the volume integral becomes 2 Ae 1 ee 2 R R rs Ine —.] Saye dr + 2fm, (r)ar + JM, (eye dr. r r 93We now perform an integration by parts on the second and fourth terms. 2 Wet own 1 2 le M (re de + 2M (r)r ! ] J Mr) (2r)dr + - fh (rar = M)(r)r + M) (RR ~My (r)r + ae R R 2fmy(r) dr + My (r)e | - r r R R JM, (rar = M) (RR + JM) (r)ar. Now, look at the surface integral r r I,+ First we note that from Ex. 9.7 the contribution to the > >> potential from M = M, is given by %, = (1/3) Myer. The contribution of the constant density M(R) is first from analogy with electric charge problems, a constant potential -MR. Thus 1 R Se %, = My(RR + Jay cree +a Mier - My (R)R 7 leo J M(r)dr + 3 Moore, a) Outside the sphere, the potential satisfies the following: . -(n+1) : = T ar P(cos®). But ¢ = ) at r= R. Thus n=0 ® 1 -2 aes) Ju, (r)dr + 3 MjReos® = AJR “cos®. Thus Ay = 3 MGR” and r 4) = (1/3) MR? cos8/r’. Since @) is independent of M,(R), then the field outside of the sphere must also be independent of + M(R), as By = -u,Voo- 949.6 We can easily show that the fields in this situation are op 2u ‘iven by H) = —4— i, ii, =| if, by showing that these 8 1 ayy vo? 82 wyF wy “0? fields satisfy Laplace's equation and the boundary conditions. By uniqueness, therefore, these would be the only solutions. > + 1) By example 9.4 H, and Hy satisfy Laplace's equation. 2) There is no field parallel to the interface, so clearly ae + (Ho. tee Hie) = Knormal is satisfied. 2uy vy 2uy Uy 3) eee ie eo ites oe satisfied. Thus H, and Hy above give the field equations, as they satisfy all the requisite conditions. + 9-7 a) Find B using Ampere's law + + > i) For 0

p9, we have a field as if all the current is at uot p20: B= 5 + b) By symmetry we have A= A,2 and A, = A(p) only. Thus we > oJ solve V7A, : V2 oo : i) For p< oy: Noa, = pap % ap “ce 7 Oe ches A, C £n p+ D. Now A, must be finite at p = 0; therefore C= 0. Moreover we take the potential to be zero so we set > D= 0. Thus A= 0 for p < p)- 95ii) For 9;

p 9: A, = Cy inp +d B-- 5; but from Ampere's law we have ulz o Qn 2n p+ De . The constant D, can be determined by matching A at p = p9- 2 ior tp ut ot | Mo 1 o ee eo eee 96. uniformly magnetized sphere we have m = f Mdv = + + os = uo(M +H). But f =m x B, thus Sut + ty =f aaty th x + x > + ray Ma sina where a is the angle between M and H. 9-9 Let us label the three regions r < Ry, Ry R, by 1, 2, and 3, with potentials 4, ¢), and 4, respectively. Then o, =) D,r"P.(cos®). Note that ¢, +0 as r+0. % =) [Bart + car) JP (cos8). 03 = -Byrcosé + ah aye) p(cos®). Now we have the following conditions: 1) $3 = $) ater 2) o =) at r= Ry >, 36 2 2 L Do Wy ae Me me at Because of orthogonality of P,(cos®), only the n = 1 term is involved (due to the magnetic field = -Hjrcos@ for large t)+ Thus¢, = Djrcos®, %) = Byrcos® + Cyr~*cos® and 03 = -Hyrcosé + Ayr@2cos® where Hy = By/uy- From (1) -B,Ro + Ay G 3 3 =z = BYR, + my oF -BYRD - Cy + Ay = GRD. RX Ry From (2) -D RP + BYRD + C) = 0. From (3) 2upAy + uyB,R3 - 2C) wy = —vgR3 Hye Ay + 41 ByR2 - 2) uy = —uRZ Ho From (4) uyB) -—g— = Diy or wy RPBy ~ 2u,C) ~ DyugR} = 0. aeSolving these equations for the constants, we get the following for D, and hence for ¢, and By: ~9uy WH R, (2uy + wy Quy + 204) = 2G) (ey -¥,) cat 9 hot z Sup uyH 2 > {(2uy + Wy)(y + 204) = *&) Gy > ¥)") For vy >»> Ug» we rewrite the expression: u 28 | 9G HG? eee u Rig FP 2 13,2 in fay a Ee 2. ae on 1)"} giving fi, = 0. Thus, for a material of high permeability, the H field in the cavity is expelled. Thus, the shell acts like a shield of the magnetic field. > . 9.10 We take H = H.y, with a potential -Hy = -H,psing. We should now see that only n = 1, sin(#) terms can enter the 6's so we rewrite our $'s in the cavity, in the shell, and outside in the form: , = Apsing, ¢) = (Bp + Cp!)sing, 98> $3 = (Hye + vp!)sing. The corresponding H fields are: > : . * : : H, = -Absing - Agcos$, Hy = -0(B - Com*)sing - $(B + Cp72)cos¢, 1 Hy = -B(-Hy - Dp~)sind - $(-H, + De™)eose. We now require either @ or Hy to be continuous at p = a,b; thus A= B+ Ca?, B+ ch = -H) + vb We now require B, to be continuous. Note that By = voll), By = wHy and By = Woy. Thus we get at a and bs uB - uCa UgA MB - ucb? = -uoH, - udb-2. These four equations must be solved for A, B, C, D in terms of Hy, a, by wy Uy siving 2 2 ae ~2ugHb (i + uy) i. a(u Yo), . yey ee (ut uy Ge aya ° uno a= oe uty a’(u - uy) ° _ ‘ In the material we have By = ul) = -uV®) = -pu(B - Cp””)sing - gu(B + cp72)cos¢ = Quy Ho be[(Gutn ) - Crew ))6 sing + (Cutw) + Gru ) )écos¢] 00 ° ° ° 2 a0 2 (Gut 4)? -G- ug)a2 9.11 We have an infinitely long cylinder polarized such that > 2 M = Myx. Note ® will have only cos dependence because of 99Saeemaeenemneenaaentanentenecaeenenetennmeeemenemenieeeaeaemennmetenmeeenmeeneenenetnneee ao matching boundary conditions with Men = M, cos. Thus the potential inside and outside the cylinder are: % = Aj)p, cos} cosd + C, “O** (We take C, = 0 because ¢, is finite as p + 0) &, = App cos¢ + C, 298% (we take Ay = 0 because ¢, is finite as De a ar AQ 2 p +=). Now at p = py we have % = ) therefore Aj = Cp/p2. ee U93%/8p + ugMep then Also at p/p, we have By, = Byy+ But Buy =p Ay + oly = Wy Cy/p2 = which gives Ay = ugMy/(ugtg)s Cy = UgMyP2/(ug + Uy) uM wy B, = - 22 (cosp 6 - sing $) +umx=—~22 a eee oo eee io ee one 2 + _ YollaMePo rcosp = . sing > aE ere ° 2 ° ° Note we did not use u,; there is no need to give this parameter. In fact, once the cylinder is magnetized, then it does not make any sense to introduce w)- 9.12 Because of the linear nature of the electromagnetic field equations, we can consider the field as a superposition, of the field induced by the external magnetic field, and the field due to the current in the wire. Consider first the effect of the external field. The magnetic potentials inside and outside the cylinder due to the external B field, are %) = Ajpcos$ and %) = -H,pcos$ + 100The boundary conditions at p = p, are %) = %), and By, = Boy or a6 2 = 2 Jpn which give Alp, = “Hye, + By/P, and HuyA = ugly + ¥yB,/P2- Solving gives nla Nets AL a ts Home 9) = -Hypcosd + a 72 H cose 1 Hy = me cos$p - a sings ii S)Hcoses -- aos Zeya Now, the magnetic field due to the current is simply calculated using Ampere's law: The field inside and outside the wire is, therefore, given as the sum of the two fields: 2H Ip auyly » : : fi, = = cosop + (. sing)é . wy) ee Ho 1p + My ‘° wer we 2 u iM 2 ioe Go. - I dered no. i, (i+ Ty Fy sDugcosts + los Ty Wy pH sing ]é 101> 7 > - 9-13 a) Small spherical cavity. Take M = M,z and H = Hoz+ From boundary conditions at r = © we conclude that only terms of order cos® survive; thus the potential inside and outside the cavity take the forms: A Cc 1 1 2 cos® + Ayr cos6, % = Hr cos + 2 cos®6 Moreover, since $, must be finite at r = 0, then A) = 0. Now we match the boundary conditions at r = R i) (R) = 4)(R) or AyR = -HgR + C,/R? or Ay = -Hy + C/R? > > 41) Byy(R) = Byq(R)- But By = -u V8 + uoM, thus Bin = ~AgloC0895 Bon = Uglgcos® + (2u4C,/R?) cos® + uoMycos®, hence ~Apug = Ugly + 2ugAy + 2ugH, + gM, 3A) = 3H, + oo Mo/3, thus i te i acs es Se ne: @,(F) = -GH, + 32)reos8 = -(H, +522, 8, = (a, + DE > > b) For a cylinder with axis parallel to M, we have H, is + + . + : continuous. Thus Hy, = Ho, = yz. Therefore B, = uaz. > « c) For a cylinder with axis transverse to M = Mx, then matching at r = © requires that only cos} terms are kept. o 1 = (A;/P) cos + App cosd, % = (A3/p) cosd - Hyp cos$. Since %) is finite as p + 0, then A, = 0. Moreover matching boundary conditions at p = p, gives 1) €1(05) = 7(04) OF Ayes = A3/Pq - HoPy or 3/02 = Ay + Hye > + * ii) Byy(g) = Byg(o,)- But B= -u, VO + uM,, then 102. 2 Ag = Ag/pG + Hy + Mye Thus Ay = -H, - M,/2, and ) = -(H, + M,/2)pcosd, By = ug (Hy + My/2)%- 9.14 The correct boundary conditions at the boundary are first Bj, = By, OF Bycos®@ = B,cos®, where 6, and 6, are the angles In n between B, and By and the normal to the interface, and the condition Hj, = Hy, or B,sin®)/u) = Bysin®,/u,. Eliminating B, and B, from these two conditions by dividing then we get 2 tan®)/u) = tanO9/uy or tand)/tan@y = u)/U9- - - 9.15 a) We use & = lp fé x J dv where J = 44) 6(r - R)sind§ = 2 free x Psind 8(r - R)r2dr sind dodo 3 quae & = - G28" 6 sin? ae dg. Now only the z component is nonzero by symmetry. Thus 8 = 6 cos® - 2 sin®, hence el gahe fans wR? + m = UE feinne z de = 1 2. ° b) To use m= fim dv we need to find the effective > > A - magnetization. K,= Mx n= Mx r = (qu/4nR)sin@g. Thus * qu > i= Ee? (see Eq. 9-20). 8 aos qwz ;, qv 24 93 _ qu” > m= fi dv = TE fav = Paz 5 rae * €) Use form of vector potential (Eq. 9-108) Ay = (¥yquR?/12qr”) . oy omxt Yo sin®$. But Anay dipole "Ge gS 7 Gey 8ined- By comparing : - eee Ay and Anag dipoles We Tead m = 45° 2 é : 1039.16 This problem is solved using the method of images as done in Ex. 9.9. We use the notations of the figure of this example. The force on the wire I is just the force exerted by the image current I). Thus from Eq- 8.28 dF ae ut I,/2n(2d)~ Since I) = Id - ud/ + Uo)s then 9.17 Taking u >> u, in problem 9.16 gives the required answer. 9.18 This problem (u, >> up) is similar to a line of charge and a conducting cylinder, which was solved using the method of images (see Figs. 3.23 and 3.24). We therefore use the method of images. Here, however, we need two image currents: one is located inside the cylinder and it is equal to I. The position of the image current can be determined in terms of a and b using the relation x, x, = R? where x, and x, are the distances of the image and the original currents from the center of the cylinder, and R is the radius of the cylinder. (See discussion on page 111). Thus x, = a2/b. The distance between the two currents is therefore D = b - a2/b = (b2 - a”)/b. The second image current is -I and located at the axis of the cylinder. Thus the force exerted on the current is equal to the force exerted on it by the two image currents: 1042 2 Oe ee ual ual at Qn D ” 2mb 2.2) ~ 2nb 2n(b? - a”) 2nb(b? = a”) To prove that these are correct values of the currents we satisfy the boundary conditions as follows: taking the origin at the axis of the cylinder, then we have I at x = -b, Ij at x = -d, and Ip at x = 0. Using the boundary condition H tangential continuous, and because 7 vit is finite, then Ay > 0 as By > Now the continuity of B normal implies B must be normal outside the cylinder. The vector potential Re tz an((x + b)2 + y2) + : : > + + 12 an((x + a)? + y*) + Iz an(x? + y2). But B= Vx A= fo ie as (K By - 5 FyAae = ¥ cosh - x sing = FD - 28, ee yy, 2D Bg Gt ey tar ay xr ty? Be BA Yep ty (etary x4y = ta Beangential ~ 9 oF B*# = 0 Rives + Began ZL + bx +y? , x + ie a a atgix +7) . (x +b)? +9 ee ay This should be true at any (x,y). We will evaluate at x = -a, +a, and 0 to get three conditions respectively:= 2 a Solving these three equations for I; and I, and d gives exactly the above solutions. In fact if we remove the condition uy >> wg, then the off axis image current is Ty = I(uy - ug)/(uy + ug) and the on axis image current is -I), and the force becomes a’? - ar M2 mand) de ~ “2nb WF, 9.19 Just like the previous problem we use the method of images. The images are the same, one outside the cylinder and one on the axis. However the force is now given only by the current outside the cylinder. Thus we get: 2n(a2 9-20 We attempt to find the force between the dipole, and its image. The magnetic field, due to the image dipole, at the point of the original dipole is {2coser + sindd}. 106The force on the “real” dipole due to a magnetic field is given > > by F = (&*V)B. Now noting that ae A 3.82, 128 a a eV = (m coset + m sindd)(s-7 + 1 558) = m cosd g- + 2 sind $5 (and noting that 91/30 = 6, 90/30 = -r, 96/ar = 0, and : 2. > ar/ar = 0, see Table 1.1), then F = at (1 + cos"6). The ms 64ne 25, torque acting is % = th x B which can be Ga8ily shown to be as given in 9.120. 9.21 For the bar on the right, we have an analogy with a oe solenoid: B = ugl(cos®, - cos@y) where Reece g L cosé, = = ——4,— + 1 ana fe Taye z (a+ £12)" + (0/2) ) 2, 9.1/2 w+ 24/2 t, cos®,, = ———,= ~40 2 ante ep ((2,/2)" + (0,/2)°) ° 3 Zu ie L fy i. For the bar on the left ge = BMI 5-7 - = uit. axis ° a + Delaaee 6 ° > > > + we have a similar result. Thus B = 2u,M and H = 2M. 9.22 We use the notations of Fig. 9.18. Thus we write NI - Hy2) = Holy = Hyky, F) + F3 or By = By + By. Thus we make the following table using the above H equations and the magnetization curve. 1074000 3125 1250 w7 +76 244 4500 1875 750 +72 54 31 cat 4650 1500 600 +73 +485, +26 --015 4750 1250 500 +73 44 +22 +.07 5000 625 250 +74 +265 ell +2365 Thus it is clear that the relation B, = B) + B, is satisfied for H, = 4675. This can be seen by plotting B) - By - Bz as a function of H, and extrapolating for a zero value. Thus the operating point of the circuit is B, = -73, By = .48, and B, = +257. 9.23 Using the notation of Fig. 9.21 we have 2H 2 £ Tae yas aaa Aye For £, = 1 cm, and since ws >> Ug, then R= fy/iigAgs Thus Plotting this equation on the magnetization curve of Alnico 5 (given in Fig. 9.8c) gives the operating point of the magnet at the intersection of the two curves. 1089.24 We follow the procedure described in Ex. 9.13. Let the fields (H,B) in materials one and two be (tH), B,) and + 9.13 we have (Hy, By). Thus according to E: Hye) + Hof) = NL or Hy = NI/fy — Hy hy /2 ByA, = ByAy or By = ByAy/Ay These two relations have to he consistent with the Hy - By magnetization curve, and H, ~ By magnetization curve of the materials. Thus from Fig. 9.8 for cast steel, we can select the following pairs (Hy, B,): (200, +33), (250, -44), (300, -55), (350, -65), (400, «73), (450, .78), and (500, -83). he substitute these values in the top two equations to generate pairs of (Hy, By)» For example the (200, +33) pair gives (240, +44) as follows: Hy = 40/+1 - 200 x .08/.1 = 400 - 160 = 240 By = 33 x 3x 1074/2.25 x 1074 = Ade Continuing we get the following pairs: (240, -44), (200, .59), (160, +73), (120, +87), (80, -97), (40, 1.04), (0, Lell). We now plot these points on the magnetization curve of material 2 of Fig. 9.8 as shown in the figure. The crossing of the two curves gives the operating point of the system: By = 1.01 7, By = +76T, and flux = BA, = ByAy = 2-28 x 10 Tm. : 9.25 The toroids (circular cross section) have a B field inside : + + + + ($ direction). We want to find an A satisfying ¥ x A= B on the . (z) axis. In Ex. 8-12 and in problem (8.18) we solved the B > > field on the z axis for a current loop, satisfying V x B= uJ 109a Wiel Tor 1-0 B(7) cast steel -6 w“ “4 2 HC Ajm) 100 avo . wu 1R2z and found B(z) = —z°——-z75 where the loop is of radius R, Mee + Rye placed at 2 = 0 with a current I = current density times cross sectional area. For a toroid with B = flux density, giving t)2 ot a) F = flux = 4(5)°B let u,J > B and B + A then : . (5)? e?z Kz) - —§ 242, - > > § Hedg = 2mpH(p) = NjI = 0 or H= 0. But B= uo(H + M), hence + 4 ugM = $B. c) The energy per unit volume is (relevant to Chapter 12) ra a 2n,8, u = §dBeH = $aBae) aD ise., the change in B with H is neglible now, from part (a), Ny = 2mbH,, hence, u = 4H,Byb/p, ise. H varies with p, but B = B, = constant, and b 1 u = fudv = 4H,Byb J p (2mpc)ap = 8TH B bed - a) per cycle. aeCHAPTER 10 10.1 a) m= 1a, 1= = =, d= we, thus a= me = Ps 2 > a b) In the presence of an external B field (B= Bz) we have from 2 : ~ep eB ae @m_>? > > B, therefore mn = 1 e2ncr> ©) X_ = - ge aT (from Eq. 10-13). Thus e -6 xm, ex?> = GEE = 6.1 « 107m? or ce> ~ 768 x 107! me ue N 10.2 a) Since the 2 electrons are moving in the same orbit and speed but in opposite directions, then their angular momenta L will be opposite and will cancel out. Hence it = 0 since it is proportional to L b) For each electron we have, from Eq- 10-9, for the change in 22 : a: Ah = - ¢ ££ %. For two electrons we have twice as much - ey ee moment: Am = — > B. 2m 10.3 The probablity P of being aligned with the field is e™B/kT P = a7 capyer’ If 75% atoms are aligned, then e +e em B/kT P= .75 = We now solve for T emB/KT, .-mB/kT 112B/KT -mB/kT. (e™ +e dd. 4 j-2mB/kT 13 gs Lor gm e + 1 thus 2mB/KT which gives 2BB = an3. Thus T = 3-67°K. 10.4 The results can be obtained from problem 10.3. Note that the previous problem we had 2mB/kT = £n3, then T = 2B which gives 5.6 x 1073°K. 10.5 a) Since n << 1, then from Eq. 10-18 we have 2 Nmou ‘ovo -3 Xpara = —3ppe 7 393 * 107 oS 3 3 b) Since M = xH and H=7-, then M = 23> = (2626) (0-1) = o 0 262.6 A/m m= [Nav = M(volume of bar) = 2.63 * 107? A/n” c) At saturation, M = Nm = (1023atoms/m?) (1-8 x 10723A/m?) = 1.8 x 10 A/am = [Mav = M(vol) = 18 A/m?. 10.6 From the Curie-Weiss Law, we have X = FOP therefore c TH Teg 5282 107 g/ee oe 15% l0ta/ee | eee ce a x which gives T, = 297°K. 113LOL LLL LLL COOOL CT ttt ett: 10.7 The saturation magnetization is M = mN = 8.44 x 10° A/m. 3kT_N From Eq. 10-46 we have y = SS = 2656. u Ms 10.8 From problem 10.7 we have for Ni, N = 9.1 x 1028 atoms/m3, and y = 2656. From Eq. 10.45, yu 82N 2 or an 7 2 ¥B N Y= which gives an = 5.2 x 10729 joules. 114CHAPTER 11 11.1 We use Eq- 11.5, for a loop of radius p < p, and for a loop of radius p > Py. § Beak =F == 4 fieaa or nd 2 . “ 24B 1 E,(2me) je BC ) giving le, | Sy OU ap so p (-1) volts/m, * |B, = -05 for p< o.. For p > p, we have V x E = 0 or 12 const - pap (OE) = 0, thus Ey = . Now at p= Py, E must be continuous, therefore sonst +05 p, giving const = .05 9, + ‘° - 2 Thus le, = 205p¢/p for p > P+ > 11.2. The flux in the coil is F = [B,+d& = uomyInr2Ny as ae = -u nome? 40 2 -7.9 x 1075(3 + dt) volts, and at ere : 2 EL _ (9 x 10~™)U)) -3, t = 2 seconds i = £ = - 2-9* 1 DOD ~ 5.5 x 10a. a > 11.3. a) We use § Acdt = F = [Bed&. For p ¢ R, F = 1B,(t)p”, thus A, = PB,(t)/2. For p > R, we have F = 1B,R?, thus met | = R7B,/2p- + . b) Ef = -0A/3A/at which gives 4B aB, B= Ey = Ye a> for p R + c) The current density in the disk is J = ok = o,$E4- a). The power dissipated is calculated from the power density 2 Jf/o,- Thus B ae)" + 4 tn 3). 6 2 a = fi /o,)dv = oF dv =o oe> : . Sy . 11.4 Since B = Bz and ¥ = %x, then E, = ¢ x B = -v,Boy and E = [0% « Beat = f(-vgB,y¥*(dxx + dyy) = -2v5B,R- 11-5 The wire (long) has current I flowing. Use cylindrical wold coordinates to find B of the wire B(p) = a + We can +> calculate the flux, find the change, or, use E = §(#xB) +d which vu Iz o 2np + is equivalent. Now vy x B= + Since we have only a z-component, only sides parallel to the z axis contribute: E = 2-403 « Beayae + adc x BU + a))az 3 vu tb ie 1 _ vu tb 2 ene oe Qn a2 a he If the loop is stationary, and the current changes, ata i uoIb jap esl 2 oo £ 219 2a dF b a foe an(1 + 3) ub di dF ° aydI 1 ae ce an(1 + Dae and BE, Ya ub Thus M = 52- gn(1 +7) = mutual inductance oo . 11.6 a) We use Eq- 11.21 E -¢ (% x B)edd. Now oop + 7 ¥= 3x 3 = pug, x B= -wpBz, thus + : E = §(% x B)edt = - wz (.03 mx .25T - .05m x .8T) = +85 volts. Thus the current in the loop is E = I = 4.25 amps. 1162 by) E = ’r= 1E = 3.62 watts + + > c) F = force = qv x B= §ide x B, and + + . t- tee 1B x [92 x B) = -1 £2(p)B) — P2By) = IV = 3.62 watts, same as in (b). + 11.7 e Stationary loop with A= 25 cm2. Flux = BeA = BA, aB da 4B thus 4 Te TE AGT BG 7 A Ge (stationary loop) E= aa, AB w cos(wt) = 7-5 cos ut aF 4B b) F = AB, E = Fae ph Use A= &x = fvt, thus 44 ~ ay, and F = tvtBysinot. Thus dt aF E= a” B tv[sin(ut) + wt cos(wt)]. - uw 11.8 a) Using the formula M = 72 u 1 L 1 a ie fat taf at ibs 2)1/2 ~ 7 2, 42, 1/2 o ™L ay +4) (eye 4) 117CT CC CS I Ct CC RT where m refers to the moving wire, and s refers to the stationary wire. ee o i 2 = x Jae log{———_-. x ——_4+—_ an a Z 1 +L -L + Yro+ L Lt vey Expanding as L becomes large we get: M u yu Im _ Yo 270 Pras ie. fat log(r,/r,) = aq log ry/t, b) The induced emf per unit length E_ -IdM. 4 _ x(x + a) + 32) L dt (x +a)? + 2? At the origin we have x = z = 0, and we have ¥ = vgx, thus GUT vy/na. 11.9 Using the dipole approximation where r >> size of loop, loop 1 has area Aj, current Ij, and using Eq. (8.87), we have fi = 114,(% cosut + ¥ sin wt). The field produced by loop 1 at the location of the second loop is calculated from 8.98, and using # = yr 3uj(fem)E = t 5,@) - ++ thus 4ur . 3(yem, )y - ia : B(x) = z = 24 1O¥ sin(we) - ¥ cos(ut) 4ur 4ur 118+ - > > But the area Ay = Ajx, then By(r)*Ay = wl AA. HLA AK cos(ut)) = Bye Thus ane wIAA E, O11 2 y sin(wet) and amr : Ey : wit A Age sin(wt) 2° R Fee) ar > : 11.10 Using Ampere's Law we get B= ujIN$/2mp. The flux ar through N loops is F = N [Beda = (ugIN?/2n)fade/p. Thus F = ugINZa tn(py/p1)/2m, hence the self inductance is L = uyN2a &n(p/04)/20- ° aces 11.11 We know that for a toroidal solenoid: - + B= ujIN$/2nr and F = {Bed&, thus the flux is vin 1 ES ao J {(2zdr) where z = asin®, or uIN 4 Fee he j 2asin® dr. Now r = b -acos®, dr = asin@d@, hence 1192 UgINa ™ sin2oae Fe i b- acosé * From Integral Tables: 1 2 f ein de a2/n2). thus 9 b- acos6 ” 2 - a’) and F = by IN(1 - YI - a/b?) = ByIN(b — 0 vs uN 2 amg To Nb = Yb? - a?) and L/length = 52 (b - A? - a”) 11.12 Let currents 1, and -I, flow along the inner and outer conductors respectively. Take the z axis along current I,. Then Anpere's Law gives = 1,4/2up and B= ull for the inside region and 0 for the outside region. The flux linking the two conductors is that crossing a plane of constant $. Thus ul ule so 0 dpdz °° b F = foeda = f Dae ne Thus L/& = (u/2n) &£n(b/a). 11.13. The difference between this problem and problem 11.12 is that the inner conductor is solid and the current is uniformly distributed. For the region between the conductors the fields and the flux are identical to those in 11.12. An extra flux due to the fields inside the inner conductor has to be accounted for. The B field for the region resulting from J = I,/na” is B= $u1oe/2na2, Thus the extra flux is 120Fe JBeda = Ligly/4n. Thus L/£ = ug/4m + (u/2) &n(b/a)- 11.14 This problem will be solved using the method of images. This procedure is very similar to the electrostatic case discussed in Chapter 3 and in Examples 6.7 and 9.9. We need to find the two currents I and -I that will produce constant potential on both conductors. Referring to Example 6.7 and to its figure 6-4, we will uge the notations in the figure. We find that the currents are located at a and -a where 2 x x a. a = Mitta me 2+ (GP - 1? cease oe The B field in the region between the cylinder and the plate is produced by I i Mot 3g _ Mo 3 2p _ ae 2p, where (4, $4) and (p_, $_) are the distance and unit vector in the $ direction with respect to origins located on the I and -I currents (Example 9.9). Now the flux linking the two conductors can be calculated by the flux penetrating the x-z plane. (atx,-R) wt a uot dp, dz+ J ° 2m, do_dz 2m, 7 ie a-(x,-R) 121ate R yy a 0 0 iia do,dz, or F/2 = >"- tn m a(x -R) 2mp, “P+ 2m Changing to the notation of the present problem, we replace x, by d and R by a and we get: Hw = 2 gnfd dy? _ L/2 = 5 anf + {(Q 1 Now if d >> a, then [(d/a)? - 1]!/2 = d/a, and u L/f = 52 an(2d/a) « 11.15 In this problem there is twice as much flux linking the two conductors. Moreover the distance to the symmetry plane is 4/2. Thus we get from the previous problem 11.16 Consider a loop with a radius a, resistance R, inductance L, in a magnetic field ujHjy, and take 3 = uz. a) Flux = -ra’y,Hjsin(wt). Thus dF 2 E=- qn te oHw cos(wt) or using complex notations 2 jut na*ugHwe*“", Kirchoff's loop law gives E = IR + iwLI which 3 R wy) ae ivt re | a Retin «among Meine R°+ wL > > : : b) t= ii x Band & = IA = 1 (x cos(wt) + y sin(wt))naz 122= Rcos(ut) + wl sin(wt) Rr ta wo? 2 > a I mau Hw. Thus t = IA x B= 00 2 R cos”(wt) ac sin(et) cos(ut) 42,42 42 2 Ro + oe beg Rcos (ut) + wb sin(2ut)/2 ,2,4,2,2,2 = f-5 = c) P= to 2 : we uh . ar aa? 7 _ Sie: @ paeeatiys terete ee o (R= Sg) ‘ + jel Je Pa ute 5 R oF wat nat 2(R° + wh") 2(R? + wi?) “se * R 2.42.22 = Wg ReCIV") = —z—— a Hw in’ : wt’) oo 11.17 a) We use the dipole approximation when the loops are far away from each other. The B field due to loop | at the site of loop 2 is normal to this loop and is equal to 2uom 2 B= —2x! , where m, = 1R,“I,- The flux through loop 2 is then 125 ad 1 Tt Fpy = By2Ay = MUgR,2Ry71)/2 which gives a mutual inductance M = 22 Fo,/Ly = gRy“Ry°/2- b) First by using $)) = 26 - 1, then one can write Eq. 11.37 as cos 26 do . ~HGKYR Ry a/2 1/2 (1 ~ kein’)? Me 7 For Ry = Ry = R and h << R, k becomes nearly one and the integrand diverges when ¢ approaches 1/2. Thus we use a limiting procedure and define 6 = 0/2 - 123LLL LLL LLL LLL CC CC Ct Ct CN CCI | oe cos 26 a6 as ah cos 26 dé : Sd) raaemoeny oly, Tae oo = kcos26y2 0 CL - kee keBe gr SN where we chose 1 - k? << 6'2 << 1. Integrating gives § 4 M = wR [en( ~ an(5) - 2] = uRfen(—t__) - 2] | - v Using (1 ~ k?)/2 = h/2R, then we have M = uyRlen@’y -2]. 11-18 a) The voltage drop across each one of them is V and it is related to L and M as follows V = LydI,/dt + MdI,/dt, V = LodIg/dt + MdI)/dt where I, and I, are the current through them. Eliminating I from the first equation and 1, from the second we get, V(Ly - M) = (LyLy - M2) dI,/dt and V(Ly - M) = (Lyby - M?) dip/dt. Adding the two equations and using I = I) + Ip, then V = b) The effective inductance is defined via V = Logg dI/dt which gives Loge = (LyLy - M*)/(Ly + Ly - 2M). 124CHAPTER 12 12.1 a) From Eq. 12.5, the energy stored is U = lj) LI? 52-1 1E- Mp LCE)” =Mpl=z = 62-5 Joules . 1 us 20 b) From Eq. 12.6, U=',IF, thus F = 5 = 7g = 25 Webers. 12.2 a) Take the z axis along the I current, then from Ampere's law we have B = ujI¢/2np between the two conductors. The magnetic energy density is u = B’/2u, = ug1?/81p2. The total energy per unit length is v j pdodeds Yo b 2 U/2 = —.— r= a b) U='yi12, then L/2 = ugen(b/a)/2m. i ° 12.3 a) Using Anpere's law, inside conductor: we get > wpl a JBeab = uoty= B(2r0) u,me"I/na? therefore B= > ¢ Qua’ Energy stored = Ihe dv, Energy/length = 0 1 2 1 a Qn u 2521? ii Taz { Bedeae = 7,~ fan J ap “= Tea ° oo ° wha b) Now, for inductor U =4L12. Thus energy/length 2 2_ Hot ip: Thus the inductance per unit length L/t is 125+ * 12-4 a) Ampere's law gives B = u,ING/2np. b) The magnetic energy density u = B¢/2u, = uj12n2/8n2p. c) The total magnetic energy is U = fadv = furdedsdz = u1?n?/(8x?)a x 2 se . 242 — lpr? Us (UN a/4n)knp,/p) + But U ='/ LI“, then L = (ugn@a/2m)tn(p9/01)+ 12.5 a) From Eq- 12.37 we have W = § a the hysterisis loop. From the figure we find it is ~ .2 J/m>. But the volume is 20 cm3, then the loss is 4 x 107° Joules. b) Power loss is W/At = 2.4 x 1074 Wate. c) We need to find the current in the loop. Using § Head = NI, and taking 20 A/m for the maximum H, then we get 20 x .2 = 100 Ing, which gives I,,, = 4% 1072 A. Since x L = NF/I where F is the flux through each loop, then L = 100 x .01/(4 x 107%) = 25 He d) If we double the current, H will double, but B and the flux will not because the system is at saturation, therefore L = NF/I will decrease. 12.6 The force can be calculated by finding the change in the magnetic energy of the system when the rod is virtually moved a small distance 4x. Because the flux is constant, then we use U = [(B2/2u)dv. Thus 126Be BZAdx = eh - Lyanx = - Uy - Uy = 7 G eee B= (1+ X_)ilgs thus AU/Ax = F = - BAX, /2Ho(1 + Xm) an °° Qunr (H 7 My)s But 0 attractive force. 12.7 a) Using Ampere's law we get B= u Rr along axis, thus ° 1 2 ee b) When flux remains constant instead of current we have r= -20. to keep the Flux constant, the force mist be in the oe opposite direction —— = —2— 6. turn 2n 12.8 The total energy of the system is the energy of interaction of both magnets with the external B field plus the interaction energy of the magnets with each other: U + Uy + + > Ujp where U) = iy+B = -2mi,Heos®,, Uz = —iiy+B = -3mu,Heosy. Now to find the U,, interaction energy we consider the B field 2 tol tan + aa 2 (3(myor)r ee mj}. At the position of the + due to one By second magnet we have r = x and r = d, thus > u > ‘om * + * By = ay By (eins - cos8,2}. Since Uyy = ~iiy*B, and hy = 3msin®,x + 3mcos®yz, then 127~3u jn? and? Ui = {2sin® sind, - cos®cosé,}, and U = -2gugHcos®, - 3muglicos®, 3y om - —z- {2sin®sin®,- cos® cos8,} and : 1 To show that. 8, = 0) = 0 are equilibrium positions we minimize the energy: 2 au Sum 39,7 2mujHsind, - z{2cos®,sin8, + sin®,cosé,} 1 2nd 2 au 3u.m ° 59,7 3muHsind,- —°;-{2sin@cos®, + cos®,sin®,} = 0 2 2nd These relations are satisfied if @) = 6) = 0, thus proving that such is a position of equilibrium. We can determine the condition that this equilibrium is stable 6; = 6) = 0: 2 3ugm a 3 Fequiring H > 5 2nd 4nd au aug a ae l= 3muH - 7 Tequiring H > 3: Thus for stable 205 and 2nd equilibrium we should have H > 75 : 4nd 12812.9 a) Vector potential of infinite wire is n> + we = ae km - 2 J ide: se a> b) From U =Y/y f JeAdv we get U=l/, f IAcdt = np. > > YW J Jay, ai, where Ijd%, is tne i'th current, and, A; is the i#j vector potential due to the i'th current. ‘he energy of one +> current in one vector potential is U = fIA-d£, for example, > + moving a current fromp ==. So, U= '/y fIyAyedd) + + + > + 2 ut, ino Vp fIgAped&y = fIyAgedhy = ire dz. For a wire L yjtyLyenedz uty Ik wht = Nonled u_Yo12 of length L, U Teese Tr CS ee em ree ae one | eae F_duU_o12_ to 12 . Oe dot «ep Pie 12.11 We use the following expressions for the moments: i, = 1, 1a2Z and ii, = tyna”Z. In the case 2 >> a, £ >> by consider a dipole field due to loop 2. (Fat, yt - a, B®) = Y@ —— « At loop | we ane nave # = 22, hence ,,(2) = 2um2/(4ae) = uit, mb?2/(2m0). : 21 0 2 02 > The magnetic energy of loop 1 in loop 2's field is U = -By)« -u Ltym?b a 3yjml 18° : ott. ms f= 2 -— E and I, are ae ag in the same sense then the force is repulsive, whereas it is 129attractive if they are in the opposite sense. + - 12.10 a) From Ampere's law H = NI,2/L5> and hence > : B= u,NI2/L,- The magnetic energy density is u = B’/2u, = n212/212, and the total energy is U = uV = u x mR? MHgN71? R2/2Lg. b) Since the flux remains constant, then if Lg changes to ly, I, has to change to qj - Iph)/Lo+ ¢) To determine the tension we need to write the energy in terms of B directly since it is kept constant: U = nB?RL,/2u,, thus T = -dU/dL, = -nBR2/2u,. 12-12 Consider moving the rod a distance dx from the field H te Hy. This is equivalent to taking a piece dx from field Hy to field Hj. Before moving piece of rod, the energy stored in the small piece of volume Adx, in the field H, is U, = Y%y wh, 2adx. After moving the piece, there is a vacuum, and the energy Ug =YpugHy2Adx. ‘Thus, by removing the lower piece, the energy 2Adx, Similarly, by placing the charges? MUyottom = V2 (H - Hy)Hy piece in the field Hy the energy changes: Aggy = + Mp (uw - ug)Hy2Adx. Total change Ueotar 72 Cu - tg)(Hy? - Hy2Adx. Now Fe Bale - uray - WA = My Gnd - Hye, Aan 4, db ead 12.13 M= 2 § so UR MII). FB = -F) = vu = cl 2 rel 2 130ie ak -dk, Te "2 § § —=—=— + Now, it is almost obvious that this el c2 |F,- Fy 4 a af, z-F een2 equals af 6.7: =~" 3 then cl c2 Ir,- ry| fe tal oS uy ah ,+ab,(,- F,) Now compare to Biot Savart law ale |F, - tI + > 1 dt) x(r - r)) + But we can show that only a piece of the cross product contributes, that is diy x (dt, * (#) - #,)) can be replaced by dty-dty(F, = #)) as shown below. Use & x (B x é) = (2-2) - (€+B)E so dk, « (ab, « CF, - F)) = (aige(ty - By )aky - Catyraty GE, - F- (ak (E, - £,))ab Deere a cael Now if § § —+———~—j— = 0 then it is right: 3 cl c2 Iep - rl a So (E, - Fab : (r, - fab § at, -[§ 2} = f aa tev x[f —— 2 el nor! s2 el |r, - FI a (ry - Fy) = fdanef vx me 5 ab, = 0. 82 cl lz, - tI 12.14 In Eq. 12.7, AX is the virtual gap, but here AX is a real gap, and we want to see what happens if it changes. (Use X 131here, instead). But we can use the same method. Now we have HL + 2H, X= NI, uA, = WoHee 2u NL Thus H(t + a X) = NI or H= TF OKT,” 2,2 wu NTA =| 2 y oo eee v= fu HL dv + a Sut, dv Tuk + Bak) Now, see what happens when we change X: 2,2 2, 242 au mu NTITA (-2u) Stee F-2 |= MTT 2cuge + 2uxy? Gaye + 2uxy? 132CHAPTER 13 13.1 a) V=RE+LdI/dt. For t <0, V= 0, I = 0 and for o's b) For O T, RI + L dI/dt = 0, I = Ip exp(-R(t - T)/L) 1 L dI/dt = (-RIp)exp(-R(t - T)/L) = -Vo(1 - exp(-RT/L))exp(-R(t - T)/L). 13.2 The transient solution of L di/dt + RI = V is I = 1, exp(-Rt/L) as usual. ‘This is the solution to the homogeneous equation. The inhomogeneous equation L dI/dt + RI = Vgexp(-Rt/L) is a special case. The particular solution must be different from the solution to the homogeneous equation. Using variation of parameters, we write I(t) = u(t) exp(-Rt/L) where u(t) is a function of t, and not a constant. We now get L du/dt - Ru + Ru = V, after factoring exp(-Rt/L). Thus du/dt = V,/L or u(t) = Vot/L + 1, where I, is our transient solution. Thus I = (V,t/L + 1,) exp(-Rt/L). Many degenerate 133differential equations have solutions with an extra power of t multiplying the original solution. 13.3 We write Kirchhoff's loop equations I, + I, = dQ/dt, Q/C + LdIg/dt = 0, RI, - LdI,/dt = 0. Differentiating the second equation with respect to time, and using the resulting equation along with the thjrd equation to eliminate I) and Q dt at from the first equation: was solution of this equation with L = 2R2C is 1 - @& [27 0+ The + I, = e7Kt(A coskt + B sinkt) where k = 1/2 RC. Because at ke t= 0, sin kt. From the 0, then A= 0, and Ip = Be second equation: Q = -kCLe ** B(-sinkt + cos kt). Using Q = Q, at t = 0, then we get the required answer. 13.4 We use results of Ex. 13.2. a) The time constant of the circuit is t= RC= 1074sec. Using Eq. 1317 we get V, = V,/2- b) Using Eq. 13-19, Vp = RC dE/dt = 107*V, where V, is the maximum voltage in one cycle of the source. 13.5 dq/dt = 1 = 86.6 qy exp(-5t) sin(86.6t) is a solution to L d2q/dt2 + R dq/dt + q/e = 0, w = 86.6 s~! with solution a(t) = 4, exp(-at)[eos(ut) + & sin(ut)] 2 2 ae) sin( ut) T(t) = -a, exp(-at) ( a 2ars vedi? #-,- bw Pee + 1) I(t + T) = I(t) exp(-aT) = I(t) exp(-2 na/w) 2 = 1 - exp(-4na/u) ¢) Energy 13.6 I, V,/R, after the switch is opened L di/dt + (Ry + Ry)I = 0 has solution I = T, exp(-(Ry + Ry)t/L)+ Thus V = (Ry + Ro)Tg exp(-(Ry + Ry)t/L) Vax = (Ry + By)Ty = VoCRy + Ry)/ Ky a) Vy = 20 Volts RK, = 100 2 Vyay = 100 Volts Ry + Ry = 5Ry Ry = 4K, = 400 2 b) d1(0)/dt = -(Ry + R)I/L L = 10 Henry, thus d1(0)/dt = -5002 x 200 ma/10 Henry = 10 As?! 13.7 V=IR+L dI/dt, V = IR + iLI where iwL is the impedance of the inductance. Substituting in this equation a. dwt = iwt = T= 1, ltt, and V = YQ ei, gives V, = I,R + iolI,. This has ° the solution 1) = gaa - a) =v < = ih, $= tan !(R/uL). cone tank b) They are in series, the current is the same, A¢ = 0. x c) Vp = gk Vi = tly 46 = F = 90° 135. iwt iwe 13.8 v=vie I= te v I~ il/wC + iwLI + RyT : vy (Ry + RT ~ iol + i/we = =v > 0 (RFR) + tab = Tuc oR +R)? + (wt - Ly? i we 1 2] Vv ° 7 oe a_ |to 1 fto Re &: fel --+ fe 2(wL - 1/we) (L + 1/wC) = 0 ra = (OR, + RD? # Cab = 1/acy?y 1/2 which gives w° = 1/LC or w = infinity. a) maximum for w* = 1/LC and minimum for w= v ee max " Ry + Ry 1 “zl max (Ry + Ry)? + Cul = I/we)? = 4(R, + Ry)? oft? - aufc + 1/w2c? = 3(R, + RA)? 22 4 2 Mo _ 0B, + RY +E # OCR, + RD" + OCR, + RAI! YC) 2 - w 2L’ 13.9 The impedances to be used are iuL, R, l/iwe,, and 1/iwC). The loops on the left and on the right have the following inpedances: z = ial/( - o7Lc)), 1 1° G7iet + Luc 1 2° Tes tea,” R/(1 + duRC,). The total impedance is 136z= 2, + 2+ Taking the imaginary part of z and equating it to zero gives w> = (C) - L/R)/LC)(C, + Cy)+ 13.10 a) += toc +. R + iwL 2 1 € ° vb) b= iwc £ = iwc [1 - —2—] z o€ - + c 0 w - iyo = iw + 7 . Thus by comparison: °° 2. 2. iw/ (Cow) + y/ (Cou) u nae ol » andR = YL = 7 cue cw ‘op oP 13-11 a) Ty 1 =0 2 I)R,R, - IR. » v= Dela2anneng (Ry + R)CRy + RR, - Ry Now for I, = 2A, I) = 1A, Ry = 19, Ry = 2M, Ry = 30 we have V, = 2 volts and V2 = zero volts. ©) 13 = 0) Tgp = IAs and Tay = 2A. 2 sauce +h, thus 2,718 1+ are 2.2, 2 = tol + 2 - Rasulo + tu + tele’ 1+ wre 2 ee tardy - ro? lel 202.2) 2 G+ w RC’) aisa2? For large & we have | a+ wea - py? which 424 wR°C 2 = L peaks at w°CL = 1, with a value of |z| = "ac- wre? 13.13 y, = iwC, yp = W/R- Yeo = iw + 1/R yy, = ~i/oL, y = yy¥Rc/ CL + Yeo) = ee ee ve + ¥_q Gwe + 1/R— tal = ——-k_____ . mus 9 = tan”! (8a : LCR + iwL - R 1+ wLe 13.14 The heat loss is simply Q =Y)ReVI". Since 1 = voetttyz, 2 1 % 2 then Q = M2 Vo" /Z where Z = » or Q=5 E> Cv,". the w + y instantaneous energy stored is the sum of the energy stored in the capacitence and the inductor which is W ='/) Gv? + My Lgt?. ‘The time average gives 2 4 2 ) Cove. we ‘00 = My (OgVg? + Lgl?) = Myc + : iwt, = 13-15 Use E = E,e fF: E, = Rl) + iwhyl, + iwMl, and 0 = RyIy + iwh oT, + iwMI) or E, - (R, + iul,)I) = iuMI,, (Ry + iwhy)1, = -iwM1). Thus E, = (Ry + duly + wM?/(Rtiub))) I) or 138(Ry + wb E, 1 ji 22 (Ry + iwl))(R, + iuL,) + wM = TiN 2° B+ ia, oe ~iuM E De 2 D2 (Ry + dub) (Ry, + iwL,) + wo 2 2442 Py 8 It, | wR, Be O22 1 \t,1 (Ry + a LR, i 2 * eee 7 Qe 2 0 = wM(R, + w LDR, (aR, )(2R,Ry ede aR, i 2.2 i se Pd ee ee why 13.16 Z, = -i/wC. Considering two voltage dividers, Re ~i/ uC, c. = CHEF CTae) * which gives RR Tut) + Tila)” G+ 6, Cc CR oo oe i. [ere ek oe ‘a b way 2 = 2 8 . equate real and imaginary parts R, R, R, 1 34 : wL - ae: = 0 and R, ®, or RLR= RRy+ 13.18 (iwC + 1/R,) Ry = (iwL + RL)/ Ry iwC RyRy, + RyRy = iwLR, + RLR,~ Equating real and imaginary parts give CRjR3 = L and RoRy = RyRy 139WER, R, RORA/R, 4 13-19 a) Kirchhoff's loop laws for phasors give the following three equations for the three small loops (assume three current loops Ij, Ij and 13 flowing clockwise in these three loops from * | left to right) V = iwL,I) [¢- (1, - 14) 27 Hy) Go ter The determinant of this set of equations is -i+t 0 = iwlyly (iat) 4/aC,) i/uc, 0 i/we, ~2i/wC)-i/ we i/wC, = 0 0 i/w0, dub, - i/wC, (iaky - i/wC,){(-21/wC, - 4/w0)ieky - i/ oC) + 1/u%cy?] ~i/(wC,)[i/(wC, )(iwk, ~ i/wC)] = 0. This gives wt = 1/L)C), 0% = W/L, (C; + 2). b) When there is no coupling between the two outer loops (C = 0), then w, = wo = 1/L,C,+ When there is very tight coupling C is large then w = 1/L,C,- 13.20 We can use a similar procedure to the one used in problem i. 13.19 to show that uf = 1/L,C, and wu} = x tor: 1 i 13.21 We write three mesh Kirchhoff's equations using phaser notations: iwl(i, - i) + Ris - ip) =8, 140(iL - t/woyi, - dwLis + Ri, - 72) = 0 R(3 I, - Ty - 13) = 0- Solving these equations we get: “ E (R + 3iwL) E (R + 3iuL) Soa 0 Fo Vy a/c + 2K - 1/wc) = + LC ” E.(R + 2iwL - i/wC) E (R + iuL) ae 0 a2 27 es C+ 2URCOL - 1/uct) + 3 LC Then 1, - 1p = 24 Egett/{w(3 + cR*)] with an amplitude as required. in/2 b) ty) - ty = It) - Iy| e Thus the phase difference is wld. 13.22 a) Series: 2 = R+ iol - i/wC = R+ i (wl-1/wC) 1 _ R= i(wh - 1/0) to ee 2 R° + (wh - 1/wC) 2 a Parallel: y = iwC + Bn iol Rate t ee Ro + wk Ro + WL 222 Rw = =——__R ist __... tl 2 a2 b) v= Iz, |v| = I|z| for constant I. Qe get (wh - 1/ucy? = Re + wt? + 2L/c + /urc?. Series: 141Wl a L ' Wl | ° o ‘ : 20 28202 ©) parallel |z2|? = (RU + wh) R? + (-wt + wer? + wa?) 2 (eee wa - 12 + eR We now find the maximum: factoring twice: So lel? = 2ut2whotn? - au2cn + 1 + utc?r?y = (R? + wt) [2(w2cL - 1)(2weL) + 2u0c2R2) = -2wC(2u2cL? - 21 + cr?) = 0 2 which gives w= Aha Ge . ° 2c. (i) This reduces for R = 0 to w = ole cL x= d/d a= Pg Woe13.23 a) Using the condition for resonance wZC,L = 1, we get c, = 10-6F. b) Q% = woL/R, thus Q = 100. At this frequency the impedance z= (+1) R= 104 (see Eq. 13.97). Since Q, > 10 and % >> 1, then it is a simultaneous phase resonance and antiresonance. ec) i- iwe, + 1/(1/R + iwL). Thus at w/w, = «9, we get z = 25(1 + 18i)Q. This impedance is inductive with a reactance equal to 450 2. 4) To annul the reactance of the parallel circuit with a capacitor we take 1/wC, = 450 which gives C) = .25 x 1076 F. i/iuC, e) Iv /vol = | TF ist, where z is the impedance of the A-B terminal. This gives 4 x 10779 at the resonance frequency 4 aut wy = 10%/s. At w= .9 w, we get 450 2. iwk + R/iwC | R + iw(L - R2c(1 - wb) R+ 1/iwC 22 13.24 a) z= 1+ eRe by z= zel@wnere z= / R21 - 2ueLC) + wk? and 1+ wr2c? tan $= [L—- R@C(1 - w°Lc)]/R- ) Taking = 0 gives the resonance frequencies we = uc - 1/R?c?. d) The average power is

= (Rez)12, where I, = Vj2 is the amplitude of the current through the source. The source voltage ts taken of the form V= Vet. Thus

= RIZ/(1 + u?R?C?). e) We need to calculate the current in the resistor. First let 143us calculate the impedance of the circuit and the current in the inductor and hence in the source for wLC = 1. S 2p2 = vetut a Zo = (R+ t0l)/(1 + w?R2C?), T= voet/z. Vp =v - vy V - iwLT = v1 - iwt/z,) ~iVRwC. Thus Ip = Vp/R = -iVuC. Thus the average power

=1/)R IR Ip = "ye ve wrc?. | i . 1 13.25 2) = tl - 2 2) = goede b) From example 13.7, cosh(ty) = 1 + 2/22 = 1 - (wLc - 1)%/2u2Lc. Then we have the following frequency range: 0 < (wLC - 1)7/wLe < 4. 144CHAPTER 14 14.1 a) Gauss' law between the plates where 2 is normal to q. * qu * - 2 s + aD 0. - he 7 pd a . them gives D = oz = 7~ sin(ut)z, thus J) = 3¢ A c08(wt)Z u The displacement current 1) = J,-A = qyucos(ut)- The conduction current Ig = dq/dt = qq weos( wt) « Observe that the two are equal. b) Since Ip is in the 2 direction then B= Bb e) = ap 1 vx(B, 9) = (uq,w/A) cos (wt)2 > (oB,) (aq, 0/A)eos (ut)2 qu pe, cy) 0 = 24+ tt). Buc B#@ = hi = 0; By wy > x cos(wt) we 8 at p= 0, then C = 0; 9,8 Ta Pcos(ut) d+ therefore B = 14.2 We are given V = 250 sin(377t) volts. (See Ex. 6-5) tee dQ/dt where Q= CV. We know the capacitance of the coaxial cylinders from Eq. 6-36 is gw me. 2 (25m) (6-7_x 8.85 107!28/m) Tn(b/ay ~ 1n(67 +5) : 145LLL LLL LLL LLL LL LO NTC CC CCC TTT Rte: Thus C= 1.0 * 107 Therefore I, = dQ/dt = 9.6 x 1075 : cos(¥77t). In = f Seeda =e f Baad. But inside the capacitor > = O/(2nepL). Since Q/L = (2meV/2n(b/a), then = # V/pln(b/a) )6- Now dE/dt = 250/(pln(b/a))377 cos(377t). Note: 250(377) cos 377t 2 eln(b/a) cos(377t). We see that I) = Tg- fda = 2nLp then t) = P= (9.6 x 107) 14.3, We are given V = 150 sin(500t) volts. (See Ex. 6.6.) I, = dg/dt where Q = 4neV ba/(b ~ a), therefore I, = 7.1 x 10> z)eos(500t). Now the displacement current is if r ec (fb us. we b= 2 then SE ee at 2 at 2 : fuer) 4ner so Ip = I,/(4nr') fda = 1,/(4mr*)4mr* = 1, so Ip = 1 Tp z. z 14.4 From Eq. 14.17 we have v°(') - ue % (5) = 0 B ae B a) E = (Ac) exp(x-ct), B = A exp(x-ct). We use a single notation. E 3 ee 2 : x (8) = (1) A expGr-ety, 5 (B) = (£2) A exp(x-ct) ax” ae Using ve = 1/c* we get 2 E 32 E c c v°(B) - pe =s (8) = (1) 4 exp(x-ct)-(1) A exp(x-ct) = 0. ot 146b) E = (Ac)&n(x + ct), B= -A gn(x + ct) E Ac 1 a2 E e (8) = (-4) sarge and Sy (8) = aC) ax ze | en 1 eo) x Oy ot (x + (x + ct) 14.5 a) We are given E = Ey sin(wt - kz). For free space we have €= €g- Sob = ch = cf = cy sin(ut - kz). We Ore = oro integrate Maxwell's Equation ab * vxi=- 2 where vx 8 = = = KE eos(ue - kz): t B-fouxdae = - KES sim(ut - ke) = - X sin(wt - kz) 2 E t- Be. - 2 sin(wt - kz) = -e E cx sin(ut - kz). we oo Mo b) We are given ii = H fexp(-i(ut + ke))+ Integrating Maxwell's * Equation V x ie Je + a where V x i = iky Hyexp (-iCut + kz)) and 7 O we get: t Be fixate = - « ¥ Hexp(-iCut + kz)) B ks sof "eee? Hyexp(-i(ut + kz). 147