Scribd
Upload
38 views

Problems in Chapter 7

Solution -em

Uploaded by

김익명
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
38 views

Problems in Chapter 7

Solution -em

Uploaded by

김익명
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 14
Problem 7.4 (a) The wave number in free space isk = w4/Ho€o. and that in the nonpermeable medium with a conductivity o is k' = wy/g(€ + 19/w). For normal incidence, the second Fresnel equation in Eq, (7.39) is written as _ Ja-Vetiofo Je+ lento *Yeot Jetiafa Jeo+ fe=iaJa _ €y— Ver +07 /a* + 2iVeylm Je — too eg + le + a* fa + D2feRe Jet aja The relative amplitude of the reflected wave is expressed as Ey feo — fetia/a Jeo ~- fe-isJo Eo Jot Veria/o Jeg+e—is/wo leg + Ve? + 07 /w? — 2f€q Re fe + ia/w &o + Ve? + 57/w? + 2,feoRe Je + ia/w The relative phase of the reflected wave is expressed as a (2) tant ee tant 2 Jeo lm fe — ia/w Eo Re(Ey/Eo) €y— fe? +07/w? (b) For a very poor conductor (o « «e), the Fresnel equation in (a) is approximated as Ey _ feo — Ve(1 + io/(2we)) Ey feo + Ve(1 + ia/(2we)) On the other hand, for a very good conductor (7 > we), VeolEg |? P ofa (2690/0 (revi netti-4) VeolEol? o/w + J2@0/0 v2 v2 1- 2eow/o 205 = RE = 1-2 2690/5 = 1-—— 1+2e,0/0 ow € where 5 = \/2/(fiwo) is the skin depth. Ee Eo Problem 7.5 (a) Let’s place the conducting sheet at 0 < z < D parallel to the x-y plane and suppose that, the incident wave comes from z = —co. Then, we may write the electric field as E,exp[i(kz — wt)] + E, exp[i(—kz — wt)], for z<0 E(z,t)=4 Ey expli(kz — wt)] + E_exp[i(—kz—t)], for 0< zD k = wl Holo * 19/0) This field should satisfy the boundary conditions on both si F, continuous at z=0: £, +E, =F, +E_ Hy continuous at z=0: — fé9(E, — E,) = veo + ia/w (Ey — E_) E, continuous at z=D: — Eye"? + Fe? = B, Hy continuous at z=D: Jeg +ia/w(E,e"? — Ee) = eck, By solving these four coupled equations, we obtain E, = (io /weg)(1 ~ e!*?) EL (2+ ia/wey)(1 — 2") + 2/1 + ta/wey (1 + 2) Ey A/T + to we, ef*? E, (2 + ig /weg)(d — e2*) + 21 + to /eoeg (1 + €2*) For an excellent conductor ( >> €9%). 's of the conducting sheet as below. 1ti 1 k= w\iplio[a) = a Jonea =~ eikD ~ e-C-I/5 = where 5 = /2/apiga is the penetration depth, and A= (1 —i)D/6. Then, —(io/weq)(1 — e~*4) (ig /weo)(1 — e-*4) + 2Yia /we (1 + e-4) ro) (1—e*) 2) + 2Jweg/ia( te) A-e*) ty +e) A fia Tues e* (ia /weo)(1 — e-*4) + 2yfia /wey (1 + e-*4) 4fwe,/io e* _ 2ye? (1—e%4) + 2fweg/io(ate4) (1 a) + y(1 + eA) where 2,/@6,/io = /2we,/o (1-1) (b) For zero thickness (A 0), E,/E; +0 and E/E; 1, which corresponds to the absence of the conducting sheet, On the other hand, for infinite thickness (A > 02), Ey 1 Ei 1+y’ which reproduces the result of Problem 7.4(b). (c) Unless the sheet has a very small thickness, [1 — nators of the result of (b). In this case, VeolEl? _ VeolEvl al > [y(1-te-*)| in the Alylte~2Rea = 2Re(e~24) + e~#Rea 2ye* Here, 2 Iv? = |y2@e07e 2-0 = 1 -DReyl? = 2(Rey)? Re(e~*”) = (e°*4 + e°?4")/2 = e-2R*4 cos(2 Im) Red=D/8, Ima =-D/5 Therefore, a(Rey)e-20/8 e778 cos(2D/6) + e778 For very small thickness, lt-e*[ site| > Wlstyl wd? 5 Dp ¢ Problem 7.6 (a) For normal incidence, the Fresnel equations in Eq. (7.39) are written as 2 T+n(@) where we assumed that the material is nonpermeable (jt = jig). Then, p — Reveoltnl? _ [Bey ponte 7 Re Jeclel? Ei 1+n(w)| Re Je(w) |Er|? =Ren(w) lef _ 4Ren(w) © [1+ n@)P E; Re feo |EiI? (b) Let's set the surface of the semi-it magnetic fields inside the slab as finite slab as the x-y plane, We may write the electric and n(w) E=Eel-0), Ba 2x Beitke-we) where the complex wavenumber is given as. k(w) = wn(w)/c. Then, Re & E-D'-B nw) =Re [= (corns = mor rg) ersinion] = Re |S" ((n-(w))* - n(w)[?) ] Pez ImAone = ew Re[n(w)] Im[n(w)] |Ee|2e~2 MK = Veo/tta Re[n(w)] Im[k(w)] [Ee[2e-2 KO)? ‘The power per unit area transmitted into the slab can be calculated as P/A= [ Re Fen -B Ww) dz = VeoTH Refn(w)]lnfk(a)] [Pf “eM de lo 1 V6 0/ Ho Reln(w)] [Eel On the other hand, the incident power per unit area can be obtained from the Poynting vector = 1 fe s=tExH E;|?2 as 1 P/A = Ve o/to [Bil Therefore, 2 Fe = Reln(w)] el = Re{n(w)] x [1+ no)? which agrees with the relative transmitted power T obtained in (a). (0) Fora conductor with o >> wep, we make the approximation c n= 14 (owe) = (1+ D¥o7 Roe) = 1 +0 where 5 = /2/apig is the penetration depth. Then, the relative reflected and transmitted powers in (a) can be written as 1-n 4-n3P 1-G-dw6/2c| > | = |- 1-(1—Da8/e| = (1- @6/c)? + (w5/c)? = 1-206 /e 4Ren 4c/wd Ac/w5 20d [itak [i+G+dc/osP [A +)c/wsP c Note that R + T = 1 as expected. ‘The power per unit volume in (b) can be written as Re [52 (B- De B-H')] = egw Re{n(w)] Imfn(w)] |E_|Pe-? KO! cy? 1 = cyto (£2) 1p, [2e-22/8 = aerte/s = coun (Zp) lBelem**/# = alba ‘The power per unit area transmitted into the conductor is then 1 1 c P./A = 5 1eo/hig Reln(«)]|Eel? = 5 Veo/ Ha 5 IE? se a F = 24 The ratio of this to the incident power per unit area P,/A = +J€0/tt9 Ei? is c 4 _ 206 @5|A+De/ws? c which agrees with the relative transmitted power T. c El = ae @S IE! ~ @6 [1 +n]? The rate of work done by the wave is 1 1 1 1 freq 4 py = Lo lB,(20-2tmik(wyle = 29-22/8 5Re(I"-B) = SolBl? = 5o1B,Ie open lBel*e Which is the same as Re [2 (E- D* - B- H°)]. The Poynting vector in the conductor is 1 a4 11 sRe(E x H') = 22——Re[n"(w)] [Eq|2e-2IMKI2 = ge E, [2622/8 S=5 ‘e(E x H*) oie e[n*(w)] [Ele rd ele eGo NE, 20~22/8 pgs? Bele Therefore, the following complex form of Poynting's theorem would be violated, if both Re" - B) and Re[2(E-D-B-H°)| are included. ty. (ExH gp —B-H)+4y-E=0 5V (EX) +SEE: D450 B= It is because the work done by the electric current J is double-counted. The contribution of the electric current is included not only in the work +Re(J*-B) but also in the electric permittivity as below. vxn ayo Dest at at n? = €/ey = 1+ i(a/weo) iweoE = -iwek Which isin turn included in Re[(E- D* - B-H")] through D = eB, Therefore, the correct complex form of Poynting's theorem is either 1 3V-(ExH) +2(E-D*—B-H) 0 where the work $J*-E is excluded, and J is implicitly included in “2(E- D* - B- H"), or ty. (Ex Ht) +28. - Be HY) +3" B=0 2 2 ° gv E= included, and the contri where the work Geer is of J is excluded from “2(B- D* — B-H’). In the latter case, Dg = €E, and Re[(B- D" - B- H")| = 0. Problem 7.16 (a) From source-free Maxwell equations, oB VxE+5-=0 > ikxE~ luo =0 VxH->-=0 > ikxH+iwD=0 where for plane waves ~ exp[i(k +x — ct)], V and 0/dt can be replaced by ik and iw, respectively. By taking the vector product of k and the first equation and then combining the ‘two equations to eliminate H, we can obtain k x (Ik x E) + #ow*D = 0. (b) By inserting k= kf into the above equation and applying the BAC-CAB rule, we obt A(R E) — E+ pov€E = 0 Let’s choose the ‘principal’ axes (not ‘principle’ axes as in Jackson) as the coordinate axes. Then, this equation can be expressed in a matrix form as below, ind — 14 v2/v? nn, nny Ey nny ng 14 v2/v} nyM B 0 ngn, Nyt, ng = 1+ v2/v3| Es where v? = 1/p1o¢; is the principal velocity. For this matrix equation to have a nontrivial solution, the determinant of the 33 square matrix should vanish, ie., ind 1+ 02/2 mn mns o=det} nn; nb —14v2/v3 nym Tgny T3Nz n§ 1+ v?/v3. o(mitnetn) — ,(nptnd rete ngent) (nb nb nk Topeptpe po! pth ped epee) oe pee vjvevs EVE Vag Sev EDS VE ve US 2 vi = 2 2. 2) 42 2 2) 4 2 2 2 = Frags int? — vB)? — vf) +n} (v? — vB) (0? — vf) + n3(0 — vO? — v)] iV205 Since v # 0, there are two distinct modes with different phase velocities v corresponding to the two roots of the quadratic equation for v? in the square brackets ({...]). Furthermore, by ing the equation with (v? — v?)(v? — v3)(v? — v3), we obtain (©) Let's express the 3>3 square matrix in (b) as below. in? — 1+ v2/v2 yn yng . many nb 14 v2/v3 nyng =R+ vee Tg, NgNy n3 -— 14+ v?/v3, ni-1 nym, myNs ivi 0 0 A=] nn, n3-1 mong], m=] 0 1/3 0 mgm, any 8-1. 0 0 1/3 Suppose that the two phase velocities of propagation modes are vg and v,. Then, the corresponding eigenvectors Eq and E, should meet the following equations, (A+ vino€)Eq=0 > BpABg + vitoBy@Ba = EaABy + viHoEn€Ep (A+ viucé)E,=0 > EAE, + viuoEc€E, = 0 Here, we used the fact that A and € are symmetric. (e.g, (EpAEg)" = Ey ATE, = E,AE,) ‘These two equations are combined to yield (vi — v5) WoEa€Ey = 0 which indicates EE, = Eq" Dy = 0. Also, E)€Eq = E, ‘Dg = 0. (However, Eq * Ey #0 in general) Then, 1 i 1 Dg: Dy = ExeeE, eons Bab = Bue VEVDHO “a Ho we Here, we used the fact that A? = —A. Problem 7.27 (a) xx (Ex B) = x x [E x (Vx A)]. Using the Einstein summation convention, {xx [Ex (Vx A)]}: = eijeajlE x (VX AD = FijnXjFxim EUW X Am = £17) Ekim ELEmpaIpAg = EnX%jEi( ep Sta ~ kqSo)OpAq = €yn%jEOKAL ~ EyH%jE LAR where éijp_ is the Le symbol in three dimensions, which meets the contracted epsilon identity together with the Kronecker delta 4); as below. EximEmpq = 5kpStq — SkqSip In the above equation, the first term can be expressed as €1j.X;E,0Ay = Ey(x x V)iA). The second term can be written as ijXj EVA = Fije[01(jE AR) — ErAndr% — XA] = ede (Erne) ~ e1jeB Andy ~ eet Andie = €1jn0,(xjE,Ax) — (E x A); — (Kx A),(V- E) where the last term vanishes because V- E = 0. Therefore, fx x [Ex (WV A)]}; = Ex(x x VAL + (EX ADs — F101 (x)BrAn) The volume integral of the last term over all space vanishes because the divergence theorem yields fax £1je01(2)E,An) = f da £:j—m%4XjE AQ = dafi- E(x x A); =0 Is Is where fi is the outward normal unit vector on the surface Sat infinity, and the fields are localized to a finite region of space. Therefore, by = eg | tex Ex WxAND = [ x s|exm+9, acm] wale vfxardnaxvn| (b) In the Coulomb gauge, the electric field is written as, ak i(k-x-t) =t) ame saa laWay(be' -cc] Then, the time average of the first term of L is calculated as E(x,t) = 1 S 3. Lspin = wal dx (EX A) [eq (e) x Gr (Ya (age (i Jel tk ¥en orre'e —[E,(he) x E(k") Ja} (day, (dee # )xei(oto'e dkdk' dx ; ; © oe? tm DD| i +E) x E(k] (Kaj, (JelK) xe-lo-o)t Le C) x & (Ie a (dae ("Je HHI Nello", where w’ = k’c. Since the time averages of the first two terms containing e#!(*") vanish, ae Nygllk-K))xe-(o-a!)t - Spasktasy | AO XG W)aWay ae" e | bspin = Tet tao |! ana x{! [CO x G(R] a3 (aye (k' eK) *eH(o- oe ap gape f [eal x Gy 0k’)Ja, day, (kee ‘| see gt peter dD| 4 kak {* [ed x (kaj (Kay (kJelo-o)" a iw “ae UO, J ak {fexa9 x &ao}a, doa (9 ~ [& 49 Wa; 4y0) 1 If we choose the polarization vectors as @ SG 4%), €,() x Gh. () = Fik, — @.(k) x (= 0 Therefore, bayn = ef Galle, WOH = la CO] ‘The two terms in the integral may be interpreted as the spin angular momentum of the left and right circular polarization. They have positive and negative helicity whose spin angular momentum should be parallel and antiparallel, respectively, to the propagation direction ik, which agrees with the above expression. Similarly, the time-averaged energy of the field can be calculated as Ue ax (E-E) [e0) - Gr (K'JJa, (day (k’)etlktk’)xe-loro’)e Eq ap g8 pry J FIELD - Gr (k')]a (days (e'Je tl) 0) meals Grae ahaa —[E,0) « Gk’) Jay (Ie)a}, (etl # Xe -ilo- wy? ~[E() - Gy (k’]aj (day (ee #)%gi(o-0"F¢ ow? [exlo) Ae YJ, (aj, (Jeon me =a aka?! x es (kK-k’ aod! (aes E(k ay (Way (k'elo-ee a > ow? = amd. J Pk {[e (06) - Gp (I) Jan (Ie) a}y (i) + [2,10 - @(I)] G3 (Kay (1)} oe It we choose the polarization vectors as. @ a ex(k)-@%(K) = 1, @a(k)-€(k) = 0 @, +1). Therefore, ak U = 2eye? [Ellas OOF + Ia-COF Both the two terms in the integral have a positive sign, regardless of the direction of the spin angular momentum (i... circular polarization or helicity), which is the property that the energy should have. Problem 7.28 ce the plane wave is circularly polarized and has a may write the electric field approximately as below. E = [Bo(x,y)(@: + i@2) + Ex(x y@s]e-O Here, the longitudinal field component (x,y) is assumed very small compared to. Eo(x, y) According to Gauss's law, ite extent in the x and y directions, we OE OE i(kz-wt) — VE [ease eecenle ES _ i 0Ey , AE o Ex, y) =e) Therefore, i OE, aed BEY 5 ] pitkz-wt) E [bot y)(@s #18.) +5 ox th a) ele According to Faraday's law, % B=-VxE 5p = ~lwB = -V x __(fi.0 (aE> , OE = i (Eo , E>) , = lia at By) # KE] & + [ikea saa tt le . . OE AE). | i¢kzw ~~ [2ke0@, + thea, + (2152-5) 84] 0 where we neglected the second-order derivatives of E> because Eo(x,y) is assumed slowly varying. Therefore, B A~vxE~ [zke0e + ikE 6: + (iS é. Jee? 3 ‘08s + IKE. A 3 ke i (AE , OE [ote yN(e1 + tea) + ( = +i5e ex] Ooo” = Fie Problem 7.29 Assuming that the longitudinal electric field component is very small compared to Eo(x,y), the time-averaged energy is written as 1 1 U=Je0f ax le? x50 ex IE)? The time-averaged angular momentum component parallel to the propagation direction is calculated as 3 EXA’+ > = 3 of ex rel(exte)-« +) ei(xa, ~yOx) Ti 1 Here, the second term does not make any contri Yin, ~ yar) 1 Ly Zof ex Re cox] 8 ion because itis purely imaginary, as below. [Eo(xdy — v0.) Bo + (tik o)(xd, — yO.) FiEo)] = > rola, ~ y0,)Eo] where we neglected the second-order derivatives of E> because Eo(x,y) is assumed slowly varying. On the other hand, the first term makes a nonzero contribution of 1 £2IEol? 5. f a zoe at ‘Therefore, the ratio of this angular momentum component to the energy is ‘The angular momentum component of a photon parallel to the propagation direction is either +h or —h, depending on whether the photon has positive or negative helicity, whereas the energy of photon is hw, regardless of the helicity. The x-component of angular momentum is written as b= farrr é, 1= 560] dxRe & 3 Ex + )”Bxx W)4j a Here, each of the two terms is calculated as re|(Ex=e")-¢)] Ey OE). 1 (Ey , OEo' Ge) axl ty. ie) 2p OFo Ox Re I> F,(y0, - 20)) = «| i x 2k =Re [Eteo(—iny — 2dy)Eq + (tiEo)(-iky — 20,)(¢1h) = ses For a cylindrically symmetric Eo(x,y), both terms have an odd parity for y and an even parity for x. Therefore, their integrals vanish, which yields L; =0. Similarly, the y- ‘component of angular momentum is written as 1 ly eof #onefea ey a = 500 | aéxRe [(exte)-« +S Hea, Here, each of the two terms is calculated as Re|(Ex" dy ~~ ok? ax 3 Data, - 109%] a = Re [Ete + ikx) Ey + (4iF)(2d, + is) = (xx oi] “& = Re| Re For a cylindrically symmetric Ep(x,y), both terms have an odd parity for x and an even parity for y. Therefore, their integrals vanish, which yields L2 = 0. In conclusion, for a cylindrically symmetric finite plane wave, the transverse components of angular momentum vanish,

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy