Solutions Manual Zill A First Course in Complex Analysis with Applications 2EPreface This student study guide is designed to accompany the text A First Course in Complex Analysis with Appli- cations, Second Edition (Jones and Bartlett Publishers, 2009) by Dennis G. Zill and Patrick D. Shanahan. It consists of seven chapters which correspond to the seven chapters of the text. Each chapter has the following features. R eae ‘Many sections of the study guide are preceded by a review of selected topics from calculus and differential equations that are required for that section. These reviews provide concise summaries of prerequisite no- tation, terminology, and concepts. For additional review, students are encouraged to consult appropriate mathematics texts. Two excellent sources that were used repeatedly for the review topics are Calculus: Early Transcendentals, Fourth Edition (Jones and Bartlett Publishers, 2010) by Dennis G. Zill and Warren §, Wright and Advanced Engineering Mathematics, Third Edition (Jones and Bartlett Publishers, 2006) by Dennis G. Zill and Michael R. Cullen. ‘A summary of every section of the text is provided. The summary reviews all of the key ideas of the section including all terminology, formulas, theorems, and concepts. Figures with two colors are included to aid in geometric understanding, Following the summary, complete solutions are given for every other odd exercise in the section (eg. problems 3, 7, 11, etc.). These are full solutions, supported by figures with two colors, that supply all of the pertinent details of the problem and incorporate the same techniques and writing style used in the text. The solutions also include references to equations, definitions, theorems, and figures in the text. The answer to each problem is typeset in color for easy reference The focus on concepts problems from the text consist of conceptual word, proof, and geometrical problems. Since they are often used as topics for classroom discussion or independent study we have included detailed hints rather than full solutions for these problems. As with the exercises, only every other odd problem is included. Final Note to Students ‘The most effective way to learn mathematics is to work many, many problems. You should not review a solution in this study guide before first working or, at the very least, attempting to work the problem yourself, Learning advanced mathematical topics takes significant time and effort. It may be quicker to look at the solutions, then try to work problems, but ultimately this approach will not lead to an independent understanding of concepts and problem solving strategies that are required for success.Chapte 1.1 Complex Numbers and Their Properties imaginary unit: The imaginary unit i denotes /=T, thus is defined by the property i? 1. complex number: A complex number is any number of the form z = a+ib where a and b are real numbers, The set of all complex numbers is denoted by C. pure imaginary: A complex number of the form z = ib is called pure imaginary. Re(z) and Im(z): If = = a + ib then the real part of 2 is Re(z) = a and the imaginary part of the = ig Im(z) =. Note Im(z) = b NOT ib. equality: Two complex numbers =, = a; + ib and 2» = a2 + iby are equal, = = 22, if a) = a2 and by = by, arithmetic operations: If 2: = a; + ib; and 22 = az + iby then Addition : tee = (a1 +42) + i(br +2), Subtraction : 2122 = (a1 ~ 42) +i(b) ba), Multiplication : 1722 = (a102 ~ bybe) + i(bya2 + ayba), Division : A _ yaa + bibs bia ~ aby vision : m7 ave to asearithmetic laws: The familiar commutative, associative, and distributive laws hold for complex numbers. Commutative Laws : ata ata, nm = 24 Associative Laws at(at+e) = (tz) +23, zi(zaza) = (2122)23 Distributive Law : n(zo+2s) = mz2 +2123 complex conjugate: If z = a+ ib then the complex conjugate of z is .. The following properties are used repeatedly: multiplication: To multiply two complex numbers, use the distributive law and the property i? = ~1. For example, — 8i + 3i — 127? = 2 Si + 12 = 14 — 5. (2+ 3i)(1 - 4i) = division: To divide two complex numbers, multiply the numerator and the denominator of 2;/22 by 73, then simplify. For example, Q4+3i_ Qt3il+4i _ 2481+ 3i+12i? 10 T-% 1-4i 1+ 4i 144i - 41 - 167 3. (5-91) + (2—4i) = (5 +2) + (-9-4)é 7-131 7. (2—3i)(4 + i) = 8 — 125 + 24-31? = 8-101 + 3 = 11 - 108 2-41 2-41 3-5i _ 6101-121 + 2077 1. oo Bee 8453-51 9-15 + 15i— 2577 15. (5-41) — (8+ 7) (5-3) +( (442i) + 2-31) +942 2-1 6-3 2-111 6+i 6-i G+i 12 + 2i — 66%. 36 + Gi 23 — 64 3719. 23, 27. 31. (3+ 6i) + (4—i)(3 +51) + (-2 4 2i)8 - = (3+ Gi) + (12 + 201 — 34 — 517) + 24i au-a-8 5 2+ a (+02 FarHim 2 + 15430 | 854851 | 24: 5 3 5 102 116 | z = (346i) +(17 +17!) + = (-2)8 + 2-242 + -2)9(2i)? + 2 =2)7(2)9 + z FCA aH + 2H 32 + (5)(16)(2i) + (10)(—8)(—4) + (10)(4)(—B4) + (5)(—2)(16) + 325 —32 + 160i + 320 — 320i — 160 + 321, = 128-1281 a.) Re(iz) = Re(i(r +iy)) = Re(iz + iy) = Re(iz -y) -¥ = -Im(z)35. If = —%2 + ¥i, then Ati = ; 7 ‘An additional solution is =2 = 2 - i. 89. Let 2 = a+ ib, then 2? = (a+ ib)? = a? + iab+ tab +707 if and only if Re (z?) = Re(i) and Im (2) = Im(i). This gives the following two equations: at = 0 2ab = 1 From the second equation, b = 3. Substituting this into the first equation we get 1 Poh 0 or t= a Solving for a we find that a = +92. Now since b = 3h, we have two solutions, a = 2, b= 4 and a=—4, §=-4. Therefore, 2 = + (42 + #i) 43. To solve we multiply the first equation by —i then sum the two equations. — =i, 1 + (1~ ia ~i(2 + 101) 4 3-5: = 2 10 = 21 Thus, —izp = 13 — Ti or 22 = 7 + 13i. Now substitute this value of zz into the first equation and solve for 2). iz, -i(7 413i) = 24101 iz, + 13-Ti = 24101 iz +17 a = 1741 AT. (a) Since i4 = , it follows that i#* = (it)* = 1 = 1. So, n = 4k.51. Suppose z122 = r where r is a real number. Then 2, = r/z,. Now multiply the numerator and denominator of r/22 by 2 and explain why the result is of the form kz where k is a real constant. 55. Assume that such a subset P exists and that i is in P. By the second property (applied two times) i3 is in P. Explain why this leads to a contradiction, [1.2 Complex Plane Rev sealar/vector: A scalar is a real number that has only a magnitude. A vector is a quantity that has both a magnitude and a direction. Geometrically, a vector is represented by a directed line segment (an arrow). Two-dimensional vectors are vectors in a coordinate plane. (a, b) : Vectors are usually denoted by boldfaced symbols such as v and a and are described analytically by giving the displacement a of the vector in the x-direction and displacement b of the vector in the y-direction using the notation (a,}). The numbers a and 6 are called the components of the vector (a,b) addition/subtraction/scalar multiplication: If k is a scalar and a = (a),a2) and b = (b1,b2) are vectors, then vector addition, vector subtraction, and scalar multiplication are defined as follows: Addition: a+b = (a, +b,a2 +b) Subtraction: a—b = (a)~b,a2~ba) Scalar Multiplication: ka = (kay, kaa) i, j vectors: The vectors i and j are defined to by i = (1,0) and j = (0,1). Notice that. any vector (a1,a2) can be written as (a1,a2) = a1 i+ a9) addition/subtraction geometrically: If a and b are vectors, then the vector sum a+b and the vector difference a ~ b can be viewed geometrically as shown in the following figures6 Chapter 1 Complex Numbers and the Complex Plane 1 i+ a2j is llall = Yap + a3. dot product: The dot product of vectors a = a +a2j and b = by i + baj is defined to be the scalar length: The length or magnitude of the vector a = ab =ayb; +a2b2 = [al] |||] cos 0 where @ is the angle between the vectors. Sian + iy can be viewed as the point (x,y) in a coordinate ‘The z- and y-axes are points/complex plane: The complex number = = plane, The set of all complex points z is called the complex plane or the 2-plax called the real and imaginary axes, respectively. vectors: The complex number z = z+ iy can also be viewed as two-dimensional position vector whose initial point is (0,0) and whose terminal point is (2, y), ot as the two-dimensional vector zi + yj which has no specified initial point. With this interpretation, complex addition can be viewed geometrically as vector sum and complex subtraction can be viewed as vector difference. modulus: The modulus of z = 2 + iy is |z| = V22 = 2? +9. The modulus is the length of the vector representing z. The following properties are used repeatedly: al_ lal, z2| (al \zazel = lea |z2|, distance: The distance between two points 21 and 2 in the complex plane is given by |22 ~ =1]- triangle inequalities: For any complex numbers 2, 22, and 2s the following inequalities hold: |lenl— zal] < lartal < laud +leal llsul—leal] << le—zel < lal + lea)7. Two adjacent sides of the triangle are (-2-8i)-(-6-5i) = 4~-3i, and (3!) -(-6-5i) = 6+8i. ana an ‘These sides are represented by the two vectors 4i—3j and 61+ 8j. The angle @ between these vectors can be determined using the dot product (see Review Topie: Vectors): (41-35) - (61+ 8j) = 24-24 =0. This implies that cos @ = 0, and so @ = 7/2. Therefore, 21, 22, and 2g are the vertices of a right triangle. 11. 2i 3-4 [ail [3-43] _ VOR © VaR Ea a wal 15. In order to determine which complex number is closest to the origin we compute the distance from =, to 0 and the distance from 22 to 0. lz: - 0) [10 + 8: /C0? + BF vi04 jo) = ju-6| = /@+Co? vis? ‘Therefore, 22 is closest to the origin. On the other hand, the distance from 2: to 1 +i and the distance from z, to 1 +i is: la -(+a)] |9 + Til J OF+ TP v130 eas} ho-7 = /Gor+(-7F = vis9‘Therefore, =: is closest to 1 +i. 19. If |z i] =z — 1), then |z — i? = |2—1]?. Let 2 = 2 + iy, then: \(@+iy)-a? = \@e+ty)-1? je+ity-P = (e-1) +ivl? e(y-1? = @-1P +e Pr Paty = Pode yeh —2y 2 yen The set of points satisfying the equation |z |z—1| lie on the line y = x in the plane. 23, If |z —1| =1, then |z — 1? = 1. Let 2 = + iy, then: \(@+iy)-1P = 1 \(@=1)+iyP = 1 (2-1)? +9? = 1 ‘The set of points satisfying the equation |z—1| = 1 lie on the circle (x ~ 1p +4 and radius 1. with center (1,0) 27. We use the triangle inequalities: | ‘Making the substitutions |2| = 2 and [6 + 8i| = v6" + 8? = 10, we obtain [2-10] <|z+6+8i| $2410. 16 + 8il| <|z+6+8il <|el +|6 +84) Thus, 8 < jz +6+8i] <12. 31. Let 2 = 2 + iy. From Definition 1.1.2, |z| — z= 2+ 1 if and only if Re(\z| ~ 2) = Re(2 + #) Im (|z| + 2) = Re(2 + i) = 1. This gives the following two equations: aes - era VPry-2 Im (V2? +9? - (x + iy) -y Since y = —1 from the second equation, the first equation becomes v2t+1-2 2 Ver+1 = 2+2 wel (242)? P+l = 4440+ 7h Sag c= 3 3 ©, i are:The complex number = ~ i satisfies the given equation. 35. In each case consider the angle between = and iz. 39. Consider |2| 43. Rewrite the equation as (zs — z2) = —k(z1 — 2). Now both zs — zz and 2 — zz can be viewed as vectors with the same initial point zp. Since k is a real number, ~k(2; ~ 22) is just a scalar multi- ple of 21 ~z2. Use the equality above to make a statement: regarding the directions of these two vectors, 47. For both parts, set z = x + iy and use the definition of modulus, 1.3 Polar Form of Complex Numbers yi enone pi polar coordinates: A point P(x,y) in the coordinate plane with Cartesian coordinates (,y) has polar coordinates (r,@) where r is the length of the line segment OP from the origin to P and @ is the angle the line segment OP makes with the positive x-axis. relation to cartesian coordinates: Cartesian and polar coordinates are related by trigonometric proper- ties of a right triangle. 2 = reosd y = rsing Piper L , 4 = tana z using arctangent: Care should be taken in using arctangent to determine @ from x and y. In general, 6 # tan“! ¥. The reason for this is illustrated in the diagram below.Arctangent is defined to return value in the interval (3,4). If(#,y) isin, say, the third quadrant, then tan~!¥ returns the angle ¢ between the positive «- axis and the ray from the origin through (—z. ~y) (which is a point in the first quadrant), instead of the angle @ between the positive z-axis and the ray from the origin through (x,y). Because these rays are in opposite directions, the difference between ¢ and @ is just a multiple of 7. ‘The following table indicates how to correctly use arctangent to find @ when x 4 0. Ifz =0, then 0= 5 when y > 0 and @ = —3 when y <0. Quadrant TI Quadrant T 6 =tan7"(y/x) +7 6 = tan“! (y/x) Quadrant IIT Quadrant IV 6 =tan"(y/z) 6 = tan (y/z) polar form: The polar form of a nonzero complex number = = x + iy is z = r(cos6 + isin @) where (r,6) are the polar coordinates of the point P(z,y). Thus, x = rcos6 and y = rsin9. In order to find r and 6 from x and y we use the formulas r? = x? + y? and tan 6 = y/z. modulus and argument: If z = r(cos@ + isin) is a polar form of a complex number z then |z| =r and an argument of z is the angle 8. Note that there are infinitely many different choices for @ that give the same complex number. For example, —2 = 2(cos + isin 7) = 2(cos 3x + isin 3x) = 2(cos5z + isin 57) and so on. In general, if é is some argument of z, then arg(z) = 60 + 2nk, k = 0,+1,+2,... Notice that both @ and r are real numbers. principal argument: If z is a nonzero complex number, then the argument 6 of z that lies in the interval (—z,x] is called the principal argument and is denoted by Arg(z) = 9. There is only one choice for the prin- cipal argument: to find it use arctangent appropriately (see the table in Review Topic: Polar Coordinates). Don't forget that Arg(z) is a real number. Any other argument for z can be expressed as Arg(z) + 2rk for some integer k.multiplication/division: It is easy to multiply and divide complex numbers in polar form. If 2 = 7,(cos 6; + isin 6) and 22 = r2(cos42 + isind2), then Sey GS ryra(cos(Oy + Oa) + isin(Oy + 62)) al _ aoa a 2 (cos(a, 42) + isin(O, 62)) Integer powers of z are easy to compute in polar form as well. If z = (cos@ + isin), then 2 1" (cosné + isinné). de Moivre’s formula: A special case of the formula for integer powers given above occurs when This formula is called de Moivre’s formula: (cos + isin 6)" = (cosn6 + isin nf). 3. Since —3i = 0 + (~3)i, we identify x = 0 and y = —3. Then r = 0? +(~3)? = 3. Since « y = -3 <0, Are(—3i) . A different argument for —3i is given by @ = Arg(—3i) + 27 Therefore, —3i 3 (cos (- 3) + isin (-3)). using 6 = Arg(—3i) —t% = 3 (os (¥) + isin (2)). using 6 # Arg(—3i). 7. For —V3 +i, we identify x = —V3 and y = 1. Then r = y/(V3)? +1? = 2. Since z = —V3 +i is in the second quadrant, ba 6 ‘Arg(-V3 + i) = tan™? ( tans tans 4) t= A different argument for —V3 + is given by @ = Arg(—V3 + i) + 27 = 4. Therefore, 2 («os () + isin (3)). using @ = Arg(—V3 +i) -V3+i = 2 (cos (272) + isin (72), using 6 # Arg(—V3 +). —V3+i 11. For —V2-+ V7i, we identify x = —V2 and y = V7. Then r = \/(-v2)*+ (v7)? = 3. Since 2 =—v2+ V7i is in the second quadrant, Arg (~v8 + V7) = tan! (-4) + r= 206168A different argument for —V3 + V7i is given by 6 = Arg(—V2 + VTi) + 2x ~ 8.34486. Therefore, -V34+ VTi = 3(cos2.06168 + isin2.06168), using 6 = Arg(—v2 + VTi) -V34+VFi = 3 (cos8.34486 + isin 8.34486), using 0 ¢ Arg(—V2 + V7i). 15. In val 5 (4-4) = 435; 2 2 19. By formula (6) of Section 1.3, us aH 3m . 3t nz = 2 cos + isin) 4 (cos 4 san 52) By formula (7) of Section 1.3 a a 4) + isin (—§)). On the other hand, |1 + i = V2 )). Now by formula (7) of Section 1.327. Since |} + di] = 2 and Arg|} +4: =4, i+ =F (GG) +4n(5)) ‘Therefore, by formula (9) of Section 1.3 (+3)" = [2G rams)" - (3) (mF) om) = FOr) & 32" 31. By formula (9) of Section 1.3 (oe () +n ($)) (6) +150) =) FCG) a) [= (5) (3) 4 By formula (6) of Section 1.3 [==(S) + (F)] fa (= (@) ++)By formula (9) of Section 1.3 (es (§) +i8(5))" a a : o8 Therefore, in order to have (48 + fi)” = —1, we must have &) =n, that is, when n = 6. and sin This will occur when "3 rere 39. Use formula (6) of Section 1.3 and notice how the modulus and argument of z is related to the modulus and argument of 212. Interpret your observation geometrically. 43. Describe the relationship geometrically in terms of a ray emanating from the origin. 4T. Use the fact that if 2) = 2, then |21| = |z2| and arg 21 = arg 22. [1.4 Powers and Roots _ . J nth roots: The n nth roots of a nonzero complex number z = r(cos@ + isin) are given by 2kr 2 we = OF [eos (2 2) + isn (=), where k = 0,1,2 n-1 principal nth root: The principal nth root of a nonzero complex number 2 = r(cos@ + isin 6) is GF [ns (S882) + isin (82)]location of roots: The n nth roots of a nonzero complex number 2 = r(cos@ + isin 9) all lie on a cirde of radius {/F centered at the origin and are spaced at equal angular intervals of 2z/n. 8. Converting to polar form —9 = 9 (cos + isinz). Now identifying r= 9, 9 = Arg(—9) = 7, and n=2 in formula (4) of Section 1.4 with k = 0, 1 gives e293) sfon(§) +) 30+) 3i & principal root sf (8) +s") 3(0-4] —3i. 7. Converting to polar form —1 +i = v2 (cos (3) + isin (3%)) . Now identifying r = V2, 6 = Arg(-1 +i) = 32, andn= in formula (4) of Section 1.4 with k = |. 1, 2 gives Ww [cos (2442008) isin (AO) salon) +050 (5) 2-¥3 42-3; & principal root 1.25992 + 1.259924 V5 [os (242008 ) sis in (SAG )| In 2 #8 [~ z) + (32) = 1.08422 + 0.290521w= (A [o» (#4 +218) en (si z 2)) cioo(2) +0 (3) 0.29052 ~ 1.084224. 11. Converting to polar form 344i = 5 (cos (0)) + isin (0), where 0 = arctan(4/3). Now identifying r = 5 and n = 2 in formula (4) of Section 1.4 with k = 0, 1 gives we = fn) sn (2808) ~ siln(2) (9) w= n(n (22208 ~ sile( See) a ($o>)] - ifm ()-()) Since cos 6 = 3, the half-angle formulas give =3, cos(2) = fLzeost a sn(2) = 3) = Vos and sing) = Therefore, m= if) : - le = 2+i principal root15. (a) (443i)? 16 + 121 +121 + 91? 16+ 241-9 = 7+24i. (b) From formula (4) of Section 1.4, the two square roots have the same modulus but arguments that differ by 27/2 = 7. But from equation (6) of Section 2.3, adding x to the argument of a complex number is equal to multiplying the complex number by 1(cosx + isinz) = —1. Therefore, the two roots of 7 + 24i are 4 + 3i and —4 — 3i. 19. (a) Since 1 = 1(cos0 + isin), formula (4) of Section 1.4 with r= 1 and @ = 0 gives yao [eos (SAZ) + san (2428) m n = cos (2) + vain (2), k=0, 1, 2,..0-1 n n Qk . ( 2kn = cox (22) 40am (2) [=o 3,9} (b) ys ww23. (a) If (z + 2)° + 2° =0, then (2+2)/2, then Since, w® = 1, part (a) of Problem 19 gives = feos (2% thy) pe w {°°8 5 +isin 3 k=0,1,2,3,4. and w + = 2cos(2kr/5). We use these values to determine 2. 2 “w4l 2 (+1) ~ (w+) +0) w+2 “w+ (w+) +1 +2 ~ 1+ 2cos (2kr/5) +1 (2cos (2kx/5) + 2) — i2sin (2k /5 - 2 + 2cos (2k7/5) Sin (2k /5) T+ cos (2k /5)" Notice then that wi z= = -1+ k= 0,1,2,3,4. Therefore, z = —1, —1 £ 3.077681, -1 + 0.726543i. (b) Conjecture: All solutions of (2 + 2)" + 2" = 0 lie on the line 2 = —1. Pere eiee a! 27. Use the fact that the n nth roots all lie on the same circle and are spaced at equal angular intervals of Qn/n. 31. From formula (4) of Section 1.4 any nth root w of z has the form = [om (2) +isin( n If we is real, then Now what does this say about @ (and, consequently, 2)?Sets of Points in the Complex Plane re disks and neighborhoods: The following sets of points in the complex plane are used frequently. lz-2| =p lz—2| Sp lz-2l

0 such that 2, < Rk for every point 2 in S. In other words, a set S is bounded if there exists a neighborhood of 0 that has S completely inside of it. If $ is not bounded, then call $ unbounded.3. By identifying 29 = -3i and p = 2 in (1) we see that this equation defines a circle of radius 2 cen- tered at —3i. 7 If we set 2 = 2+ iy, then Im(Z + 3i) = Im(z — iy + 3i) = 3-y. Thus, Im(z + 3i) = 6 if 3—y = 6, or Therefore, the set of points satisfying Im(z+3i) = 6 is the line y = —3 1. If we set x+iy, then Re (2?) = Re((r+iy)*) = Re((z?—y") + 224i) = gay? Therefore, the set of points satisfying Re (z”) is the hyperbola 2? — y? = 1.15. 19. 23. If we set z = 2 + iy, then Im(z) = Im(z + iy) =u, and so, the set Im(z) > 3 is the half-plane y > 3. (a) Since this set contains none of its boundary points it is open. (b) Since this set does not contain all of its bound- ary points it is not closed. (c) This set is open and connected, so it is a do- main. (d) This set is not bounded because there does not exist a neighborhood of 0 that contains it. (e) This set is connected since any pair of points in it can be joined by a polygonal path completely inside of S. If we set z = x + iy, then Re(z?) = 2? —y?. It follows that the set of points satisfying Re (z”) > 0 is the set x? — y? > 0. That is, (x +y)(t—y) > 0. This inequality holds when x > —y and z > y, or when x < —y and x B is used to denote a function from A to B. values: If the element b € B is assigned to the clement a € A by a function f : A — B, then we call b the image of a under f, or the value of f at a. We write b = f(a) to indicate this assignment. domain: If f: A— B, then the set A is called the domain of f, written as Dom(f). range: If f : A» B, then the set of all images in B is called the range of f, written as Range(f). Using set notation the range is defined by Range() = (/(@) | a € A}. one-to-one If f : A> B, then f is called one-to-one if every element in the range of f corresponds to exactly one element inthe domain of f. In other words, f is one-to-one if whenever a1 # a2, then f(a) # f(a2) onto: If f : A+ B, then f is called onto if the range of f is equal to the entire set B. That is, f is onto if for every b € B there is some a € A so that f(a) =. 29complex function: A complex function, or a complex-valued function of a complex variable, is a function whose domain and range are subsets of the set C. That is, if f : A+ B is a complex function, then we must have AC C and BCC. w = f(z): The notation w = f(z) is used to represent a complex-valued function of a complex variable, while the notation y = f(z) is used to represent a real-valued function of a real variable. Re(f) and Im(f): Given a complex function w = f(z), if we set z = x + iy and express the function in terms of two real functions as: F(z) = u(x,y) + iv(x,y), then the functions u(x,y) and v(c, y) are called the real and imaginary parts of f, respectively. We use the notation Re(f) = u(y) and Im(f) = v(2,y) complex exponential function: The complex exponential function is defined by: * cos y + ie” sin y, polar coordinates: Given a complex function w = f(z), we can replace the complex variable = with its polar form z= r(cosd + isin) and express the function in terms of two real functions as: F(z) = u(r, 8) + iv(r, 6). We still call Re(f) = u(r,@) and Im(f) = v(r,6) the real and imaginary parts of f, but these real functions are different from those obtained using Cartesian form x + iy of the complex variable z. 3. (a) For z= 1 we have |z| = 1 and Arg(z) = 0. Therefore, f(1) = log, 1 + i(0) =0. 4and Arg(2) (b) For z = 4i we have |2| §. Therefore, z 2 = V2 and Arg(z) = f(4i) = log, 4 + i= ~ 1.38629 + 1.570804 (c) For z = 1 +i we have . Therefore, fati= log. V3-+i5 S loge 2+ i> 0.34657 + 0.785403. 7. (a) For z = 3 we have r 3 and @ = Arg(z) = 0. Therefore, I} 3+icos?0 =3+4i(1)?(b) For 2 = —2i we have r = |2| = 2 and 0 = Arg(2) = —§. Therefore, S(-2i) = 2 + icos? ( (c) For z = 2~i we have r = |2| = V5. Moreover, from equating the Cartesian and polar forms of the point z= 2 V3(cos 6 + isin) we obtain cos = 2/5. Therefore, ay 10-1) =v5-+i( V5 + $i x 2.23607 + 0.8%. 11. If we set 2=2 + iy, then J(2) = (2 +iy)?-2c+ iy) +6 = 2° 43eyi + 3ay?i? + y%i? — 22 — yi +6 = 23 4322yi —32y? — yi — 2x — 2yi + 6 = (23 — 3y? — 20 + 6) + (32?y — y* — 2u)i. ? —2n +6 and Im(f) = 31?y — y? — 2y. Therefore, Re(f) = 2° ~ 3 15. If we sot = =x + iy, then 2(z + iy) + i = 2x + (2y + 1)i. So, from (3) of Section 2.1 we have: eet uti e cos(2y + 1) + ie* sin(2y + 1). Therefore, Re( f) = e7* cos(2y + 1) and Im(f) = 22 sin(2y + 1), 19. If we set = = r(cos@ + isin@), then [r(coso + isino)]* = r*(cos40+isin4@) — < see (9) in Section 1.3 f2) = r*cos46 + ir’ sin 49. Therefore, Re(f) = r4 cos 40 and Im(f) = r* sin 40, 23. Since Re(z) and 2? are defined for all complex numbers, the natural domain of f(2) = 2Re(2) ~ i=? is the set C of all complex numbers. 27. (a) Does arg(z) assign one and only one value to 2? (d) See part (a),31. In order to determine the natural domain consider the questions: For which values of zaned ys cos( — y) defined? How about sin(x — y)? In order to determine the range, consider |f(2)| 35. Verify that |e~ +, then use this identity to answer the question. 2.2 Complex Functions as Mappings roe oe eine parametric curve: If f(t) and g(t) are real-valued functions of a real variable t, then the set C of all points (s(t). g(t), where a < t <6, is called a parametric curve. The equations x = f(t) and y = g(t) are called parametric equations for C, and the variable ¢ is called the parameter. eliminating the parameter: Given a set of parametric equations ¢ = f(t), y = g(t) it is sometimes possible to eliminate the parameter t to obtain a single Cartesian equation of the curve. For example, if the parametric equations « = t? — t+ 2, y = —t? +¢ for a curve C are added together we obtain the single equation x +y = 2 (so, the curve C defined by these equations is a Tine). smooth curves: A parametric curve given by x = f(t), y = g(t), a 2 a) After replacing 2 by x + iy in the mapping w = 32, we obtain w = 3(¢ + iy) = 3x + Syi, and so the real and imaginary parts of the mapping w = 3z are (2) u=3r and v=3y.The image S’ of $ under w = 3z is found by using the equations in (2) to transform the bound s on x and y in (1) into bounds on u and v. If we substitute 2 = u/3 and y = v/3 in (1) and simplifs~, then we find that S’ is given by -co6. Therefore, S’ is the half-plane Im(w) > 6. The mapping is depicted below. ue half-plane $ can be described by the two simultaneous inequalities -oo3 and -w 3. The mapping is depicted below, = 11. The line x = 0 can be described by =0 and -w0 oe *" re 2 ; . _ US — - translates the point z along rotates the point 2 through _- magnifies the modulus of the the vector b an angle of @ = Arg(a) point z by a factor of a radians about the origin linear mapping as a composition: If f(z) = az +b is a complex linear mapping and a # 0, then we can express f as: : az) +b. im) (ToM oR) (z) where R(z) = (2 is rotation by Arg(a), z+ bis translation by b. This means that f(z) is the composition f(=) ‘M(z) = |alz is magnification by |a|, and T(z)image of a point under a linear mapping: If /(2) = az +} is a complex linear mapping with a A 0, and if wo = f(20) is plotted in the same copy of the complex plane as zo, then wo is the point obtained by (ji) rotating z» through angle Arg(a) about the origin, (ii) magnifying the result by | (iii) translating the result by b. geometry of linear mappings: Because a non-constant linear mapping is a composition of a rotation, @ magnification, and a translation, the image of a geometric figure can have a different size, but its basic shape is the same. For example, the image of a circle is a circle, the image of a 5-sided polygon is a 5-sided polygon, ete. 3. (a) Using (6) of Section 2.3, we express the linear mapping f(2) Biz as Thus, the image of the closed disk |z| <1 is found by (i) rotating the disk by Arg(i) (id) magnifying the result by |3i| = ™ (iis) translating the result by 0 (no translation). Under the rotation (i), the disk is mapped onto itself and under the magnification (i) the result is mapped onto a disk centered at the origin with radius 3. Therefore, the image is the closed disk |w| < 3. (b) 7. (a) Using (6) of Section 2.3, we express the linear mapping f(z) = z+ 2i as f=) (i) +21 =1(1z) +24. ‘Thus, the image of the triangle is found by2.3 Linear Mappings ov (i) rotating the disk by Arg(1) = 0 (no rotation), (ii) magnifying the result by |1| =1 (no magnification), and (iii) translating the result by 2i. Under the translation (iii), the vertices 0, 1, and i are mapped to 2i, 1 + 2i, and 3i, respectively. Therefore, the image is the triangle with vertices 2i, 1 + 2i, and 3i (b) ‘Thus, the image of the triangle is found by (i) rotating the disk by Arg(—1) (ii) magnifying the result by | —~ 3] = 3, and ™ (iif) translating the result by i. Under the rotation (i), the vertices 0, 1, and i are mapped to 0, ~1, and ~i, respectively. Under the magnification (ii), these vertices are then mapped to 0, ~3, and —3i. Finally, these vertices are mapped to i, 3+ i, and —2i under the translation (ii). Therefore, the image is the triangle with vertices i, —3+ i, and —2i. (b) rotate by = magnify by 3 translate by i15. Using (6) of Section 2.3, we express the linear mapping f(z) = 19. 23. 4z+1-V3i as S(2) Thus, the image of the triangle is found by (i) rotating the disk by Arg(—1) (ii) magnifying the result by |—}) (iii) translating the result by 1 — V3i. =} 3, and There are many ways to map the imaginary axis onto the line containing i and 1 +2i. One ap- proach consists of first rotating the imaginary axis onto the line y = , then translating this line 1 unit in the y-direction. In other words, one solution to the problem is a linear mapping that consists of a rotation clockwise through 7/4 radians followed by a translation by i. This gives f(z) = e~**/42 4 (a) Using 2(t Section 2.2: zo(I— t) + ait and T(z) = = +6, the image of the line segment is given by (11) of w(t) = T(t) z(1—t)+at+b zo(1—t) + ait +B(1 —t) + bt (zo +5)(1 —t) + (zr +d). u ‘Therefore, a parametrization of the image is w(t) = (zo + 8)(1~ 1) + (21 +0), OS 1S 1. From parametrization (7) of Section 2.2, we see that the image is a line segment from =) + to =: +b. (b) Using z(t) = zo(1—t) + zit and R(z) = a2, |a| of Section 2.2: , the image of the line segment is given by (11) w(t) = R(2(t)) a(29(1 -t) + zit) = (az)(1-t) + (ax)t. ‘Therefore, a parametrization of the image is w(t) = (az0)(1~t) + (az)t, 0 0, the image of the line segment is given by (11) of Section 2.2: w(t) M(z(t)) = a(zo(1-t) +211) (az9)(1-1) + (az1)t ‘Therefore, a parametrization of the image is w(t) = (azo)(1~t) +(az)t, 0 < t < 1. From paramettiza- tion (7) of Section 2.2, we see that the image is a line segment from az to az. Since a > 0, the endpoints azp and a2; of the image line segment are obtained by magnifying the points 29 and =, respectively, by a factor of a.27. (a) Given translations T;(2) = 2+; and T;(z) = 2 + be, their composition is Ty 0To(z) = Ty(z + ba) = (z + ba) + br = 2+ (br +a). If we set b = b +ba, then the composition can be written as T; 0T3(z) translation if by +b2 4 0 and the identity if by + bo = 0. Since Ts T(z) composition does not matter. (b) Modify the argument in part (a) (c) Modify the argument in part (a). +b, Therefore, Ty oT2 is a z+ (by + ba), the order of 31. Let f(z) = az + for some complex constants a and b, Consider the identity |z| = |az + b| for the values 2 = 0 and z = 1. What do the two resulting equations tell you about the coefficients a and b, and, consequently, the linear mapping w = f(z)? 35. (a) Solve the system of equations: ant+b = ow a+b = we for a and b. (b) Consider f; a rotation and fz a magnification. What is the image of = 0? 39. (a) Rotate, magnify, and translate the anmulus in order to determine its image, then determine the least and greatest distance a point in the image can be from the origin. (b) Determine the points in the original annulus that map onto the points that are closest and farthest from the origin identified in part (a). (c) Notice that the closer a point f(z) is to the origin, the farther 1/f(z) is from the origin and the farther f(z) is from the origin, the closer 1/(2) is to the origin. [2.4 Special Power Functions one-to-one functions: A function f is one-to-one if whenever f(a1) = f(az), then a) = a2. In other words, f is one-to-one if whenever a1 # az, then f(ai) # f(a2). Put yet another way, f is one-to-one if every clement 6 in the range of f corresponds to exactly one clement @ in the domain of f-verse function: If f is a one-to-one function with domain A and range B, then the inverse function f~! is the function with domain B and range A defined by f(b) =a if f(a) = b. composition properties of inverse functions: If f is a one-to-one function with domain A and range B and if f~! is its inverse, then (fos) @) = S(F7@)) = 6 — forall bin B, and (ef) (a) iO) ee inverse mappings: If a one-to-one function f maps a set S onto a set T, then the inverse function f~! maps the set T onto the set S. finding inverses algebraically: In order to find the inverse of a one-to-one function f algebraically we solve the equation f(a) = b for the independent variable a, then relabel the independent and dependent variables appropriately. For example, the complex funetion f(z) = 5z ~2i is one-to-one. If we solve the equation w = 5z — 2i for the independent variable z we obtain z = (w + 2i)/3. Relabeling the independent and dependent variables gives the inverse function f~1(z) = (z + 2i)/5. restricting domains: If f is a function that is not one-to-one, then it does not have an inverse function. However, it may be possible to restrict the domain of f to a subset on which f is one-to-one and, consequently, use the restricted domain to determine an inverse function. For example, the real function f(x) = x? is not ‘one-to-one on its natural domain (—>0, 00). The function f(z) = x? is one-to-one, however, on the restricted domain (0,00). On this restricted domain we have the inverse function f-*(«) = Vz. complex polynomial function: A complex polynomial function is a function of the form P(z) = an2” + On_12") +... $a1zZ + a9 where n is a positive integer and dp, @y—1, «.. @1, and aq are complex constants. If an # 0, then n is called the degree of the polynomial p(z). complex power functions: A complex power function is a function of the form f(z) = 2* where @ is a complex constant. complex squaring function f(z) = 2?: The complex squaring function f(z) = 2? can be expressed in terms of it real and imaginary parts using cither a Cartesian, polar, or exponential form of the variable =. Each is given below: Cartesian polar ‘exponential ——_. _—»-=+——-——— <> fl) = 2 = (a?-y)+20yi = 1°(cos20+isin2) = rei” the mapping w = 27: As a mapping, w = 2? squares the modulus of a point and doubles its argument. ‘The following summarizes some properties of the mapping w = 2?2.4 Special Power Funetions ray arg(z) = 6 circular segment |z| = 1, oS arg(z) <6 vertical line Re(z) =k circular segment: |10| = 1? 26 < arg(w) < 26 parabola u = k? — v?/(4k?)parabola u = v?/(4k”) horizontal line Im( ‘As a mapping, w = 2" = r"e'”® raises the modulus of a point to the nth power and the mapping w = =": increases its argument by a factor of n. The following summarizes some properties of the mapping w = 2” | circular segment |z| =r, circular segment |w| =r", oS arg(z) $8 ng < arg(w) < nd principal square root function 21/2: The principal square root function is defined by 2? = VfajeAtetey2 vrei? 0 = Arg(z).If the domain of f(z) = 2 principal square root function is the inverse of the function f(2) = is restricted to the set —1/2 < arg(z) < 7/2, then f is one-to-one and the on this restricted domain. principal nth root function 21/”: For integers n > 2, the principal nth root function is defined by an = ile = yre%/n, 6 = Arg(z). If the domain of f(z) = 2” is restricted to the set —7/n < arg(z) < /n, then f is one-to-one and the principal square root function is the inverse of the function f(z) = 2” on this restricted domain. sArg(=)/n multiple-valued functions: A rule F that assigns a set of one or more complex numbers to each complex number in a subset of C is called a multiple-valued function. For example, G(z) = arg(2) is a multiple Valued function that assigns an infinite set of numbers (all differing by a multiple of 2) to a nonzero complex number 2. Multiple-valued functions are not functions (as defined in Section 2.1) pee 3. By identifying k = 3 in equation (3) of Section 2.4.1 we see that the image of the line x = 3 under the mapping w = 2? is the parabola u = 9 ~ v?/36. 7. Notice that the positive imaginary axis is the ray arg(z) = x/2. From the 2.4 Summary, the image of this ray is a ray making an angle of 2(7/2) = x radians with the positive real axis. Therefore, the image is the ray arg(w) = 7, the negative real aixs. 11. The triangle with vertices 0, 1, and 1+é consists of three line segments. We treat each of these segments separately. The first segment from 0 to 1 lies in the ray arg(z) = 0. So, by the discussion on pages 73-74 of Section 2.4, its image under w = 2? lies in the ray arg(w) = 2-0 = 0. Since the endpoints 2 = 0 and z= 1 map to w = 0 and w = 1, respectively, the image is the line segment from 0 to 1. Similarly, the line segment from 0 to 1+ i lies in the ray arg(2) = 3 and so its image lies in the ray arg(w) = 2-£ = &. Since the endpoints = = 0 and z = 1 +i map to w = 0 and w = 24, respectively, the image is the line segment from 0 to 2i. The last line segment from 1 to 1 + lies in the vertical line z = 1, not a ray emanating from the origin, thus its image will lie in a parabola. Identifying k = 1 in equation (3) of Section 2.4.1 we see that the image will lie in the parabola u = 1 ~ v?/4. Since the endpoints 2 = 1 and z = 1 +i map to w = 1 (v = 0) and w = 2i (v = 2), respectively, the image of this last line segment is the parabolic arc u = 1 — v?/4, 0 < v < 2. Therefore, the image consists of the line segment from 0 to 1, line segment from 0 to 2i, and parabolic are u = 1 — v/4,0< 0 <2. ‘The mapping is depicted in the figure below.15. The function f(z) = 227 +1—# can be expressed as a composition f(z) = (ho g)(2) of the linear function h(z) = 2z +1—i and the squaring function g(z) = 2?. Under the squaring function, the ray arg(z) = 7/3 is mapped onto the ray arg(w) = 27/3. Next we consider the action of the linear function on this image. By (6) of Section 2.3, the linear mapping h(z) = 22 + 1 ~ i consists of a rotation by Arg(2/|2|) = 0 (no rotation), followed by a magnification by |2| = 2, and a translation by 1—i. The ray arg(w) = 27/3 is mapped onto itself by the magnification. After the translation by 1 —i, the image is a ray emanating from 1 — i making an angle of 2x/3 with the line y = 1. Since 0 is not in the set arg(z) = 7/3, the point f(0) = 1 — i is not included in the image. Therefore, the im- age can be described as the ray emanating from 1—i and containing (73 ~1)i excluding the point 1— 19. The function f(z) = }e'*/42? can be expressed as a composition f(z) = (hog)(z) of the linear function h(z) = Je'*/4z and the squaring function 9(2) = 2?. Because the squaring function squares moduli and doubles arguments, the circular arc |z| = 2, 0 < arg(2) < 7/2 is mapped onto the circular are |u| = 4, 0 < arg(w) < x. Next we consider the action of the linear function on this image. By (6) of Section 2.3, the linear mapping h(z) = Le'*/4z consists of a rotation by /4, followed by a magnification by 1/4, and a translation by 0 (no translation). Under the rotation, the circular arc |w| = 4, 0 < arg(w) < = maps onto the circular arc |w| = 4, 7/4 < arg(w) < 5/4. Finally, under the magnification, this image is mapped onto the circular arc |w| = 1, /4 < arg(w) < 51/4. Therefore, the image is the circular arc |w| = 1, 7/4 < arg(w) < 51/4. 23. (a) Since the function f(z) = 2? squares the modulus and doubles the argument, the image of the region 1 < |z| <2, 7/4 < arg(z) < 3n/4 is the region 1 < |w| <4, 7/2 < arg(w) < 30/2. (b) Since the function f(z) = ° cubes the modulus and triples the argument, the image of the region 1< |z| <2, 2/4 < arg(z) < 3n/4 is the region 1 < |w| <8, 3x/4 < arg(w) < 9n/4. (c) Since the function f(z) = 2* raises the modulus to the 4th power and quadruples the argument, the image of the region 1 < |z| < 2, 7/4 < arg(z) < 3n/4 is the region 1 < |w| < 16, 1 < arg(w) < 37. That is, the image is the annulus 1 < |w| < 16, 27. From (14) of Section 2.4.2, since | — 1] = 1 and Arg(—1) = 7, we have that (9 = We"s = ; + 4, 31. Since the principal square root function takes the square root of the modulus and halves the argument, the image of the ray arg(2) = 7/4 is the ray arg(w) = 7/8. 35. Since the principal square root function takes the square root of the modulus and halves the ar- gument, the image of the circular are |z| = 9, —1/2 < arg(z) < 7 is the circular arc |w| = 3 —n/4< arg(w) < 7/2.39. We consider the image of the boundary of the region first. The boundary of the region lies in the union of the ray arg(z) = 7/2 and the parabola x = 4— jgy?. Since the principal square root function takes the square root of the modulus and halves the argument, the image of the ray arg(2) = 7/2 is the ray arg(w) — 1/4. In order to determine the image of the parabola we use the fact that f(z) = ='/? is the inverse function of f(z) = 2? defined on the restricted domain —1/2 < arg(z) < 7/2. By (3) of Section 2.4.1 with the identification k? = 4, the image of the lines z = +2 map onto the parabola u = 4— age? under w = z2. Only the line x = 2 is in the restricted domain of {(2) = 2?, and so we conclude that the image of the parabola x = 4 — jgy? under w = z'/? is the vertical line u = 2. Thus, the image is bounded by the ray arg(w) = 7/4 (or, equivalently, the line u = v) and the line u = 2. Since the point 7 + 24i is in the region and (7 + 24i)/2 = 3+ 4i, the image of the region must be the set bounded by the lines u =v and u = 2 containing the point 3+ 4i. Fo! 43. Use the fact that arg(w) = 1/2 and arg(w) = 37/2 are the same ray. 47. Consider different subarcs of the circle |2| = v2. 51. Remember that the range of the principal nth root function is = 0 have images with modulus 1/r > 0, , and the arguments are unchanged. Hence, the image under inversion in the unit circle is the set arg(w) = 7/4. Reflecting this set across the real axis negates arguments and does not change the modulus. Therefore, the image is the ray arg(w) = —7/4. From the summary of Section 2.5 we have that the circle |z + $i = | 4| is mapped onto the horizontal line v = 1/k by the reciprocal mapping w = 1/z on the extended complex plane. By identifying k = 2, we have that the image of the circle |z + i] = 1 is the horizontal line v = 1/2.50 15. 19. Chapter 2 Complex Functions and Mappings From the summary of Section 2.5 we have that the vertical line s = k is mapped onto the circle |w — 4,| =|; | by the reciprocal mapping w = 1/z on the extended complex plane. By making the identifications ky = —2 and ky = —1, we have that the boundary lines x = —2 and x = —1 of the set $ are mapped onto the circles |w +4] = |3| and |w + 3| = |4|, respectively. Since the point z = —3/2 is in the set S, the point w = —2/3 is in the image of the set. Therefore, the image of S under the reciprocal mapping on the extended complex plane is the set bounded by the circles |w + 3] = } and jw+al=a 3 and containing the point w = —2/3. (a) The first function in the composition is f(z) = 1/z. Thus, the mapping w = g(f(z)) first inverts in the unit circle, then reflects across the real axis. From the discussion on page 65 of Section 2.3, the second function in the composition, g(2) = 2iz +1, rotates through an angle of Arg(2i) = x/2 about the origin, magnifies by |2i| = 2, and then translates by 1. Therefore, the mapping w = f(z) inverts in the unit circle, reflects across the real axis, rotates through an angle of Arg(2i) = 7/2 about the origin, magnifies by [2i| = 2, and then translates by 1 (b) Under the reciprocal mapping on the extended complex plane, the vertical line x = 4 maps onto the circle |w — }| = 3. Now under the linear mapping g(z) = 2iz +1, the circle is rotated through an angle of 7/2 about the origin, magnified by a factor of 2, and then translated by 1. The rotation maps the circle onto the circle jw ~ $i] = } with the same radius but whose center has been rotated by 7/2. The magnification maps this image onto the circle |w — }i] = } whose center and radius have been doubled. Finally, the translation maps this image onto the circle with the same radius, but whose center has been transformed to w = 1+ di. Therefore, the image of the line « = 4 is the circle |w-1- 43 =4 (c) Under the reciprocal mapping on the extended complex plane, the circle |2 + 2| = 2 maps onto the vertical line u = —4. Now under the linear mapping 9(2) = 2iz +1, the line is rotated through an angle of %/2 about the origin, magnified by a factor of 2, and then translated by 1. The rotation maps the line onto the line v = —4. The magnification maps this image onto the line v = 2. Finally, since the translation is along a vector (b = 1+0i) in the same direction of this line, the line maps onto itself. Therefore, the image of the circle |z + 2| = 2 is the line v = 4 23. 27. Modify the procedure used in Example 2 of Section 2 (a) Notice that L is a generalized circle with A = 0 (see (7) in Problem 25 of Section 2.5). By Problem 26, the image of L is the generalized circle given by (8). What must be true about the coefficients in (8) in order for the image to be a line? (b) Use part (a) and consider the coefficients of the image line (c) As in part (a), use (8) to determine an equation of the image circle. Complete the square in order to determine the center and radius.[2.6 Limits and Continuity 7 reas limit of a real function: Informally, the limit of f as x tends to zo exists and is equal to L, denoted lim f(z) = L, means that values of the real function f(z) can be made arbitrarily close to the real number L if values of x are chosen sufficiently close to, but not equal to, ». The precise definition is: lim f(z) = L if for every € > 0 there exists a 6 > 0 such that |f(x) — L| <¢ whenever 0 < |x — ao| < 6. left and right-hand limits: The limit of f as x tends to 9 from the left, denoted by lim f(z), and the limit from the right, denoted lim, f(c), are defined by: . lim f(z) = L if for every ¢ > 0 there exists a 6 > 0 such that |f(x) — L| <¢ whenever x96 < x < xo; for lim, f(x) = L, we require x9 <2 < 29 +6. ‘A real limit exists when the left and right-hand limits are equal. That is, lim f(r) =L if andonly if lim f(z) = lim f(x) =L. properties of real limits: The following properties allow many real limits to be evaluated in a somewhat mechanical fashion. Suppose that f and g are real functions. If lim f(x) = L and Jim g(x) = M, then (i) lim ¢=e, where cis a real constant, (i) lim x= 20, (iii) lim ef (x) = cL, where ¢ is a real constant, (iv) Jim (/(@) + o(@)) = LM, (v) im S(@)- oa) = LM, (09) tim £2) > provided M 4 0. at (2) continuity of a real function: A real function f is continuous at a point 29 if lim f(x) = f(to)- If follows from the properties of real limits that real polynomial and rational functions are continuous at all points in their domains.limit of a real multivariable function: The definition of limit for a real multivariable funetion F(z, y) is similar to that for a real function f. (ea B yay FO 9) = E if for every = > 0 there evists a 6 > 0 such that |f(z.y) ~ L| < e whenever 2.4) (4 0< Ve = a0) + my <5. continuity of a real multivariable function: A real multivariable function F is continuous at a point (20, yo) ier pli F(z,y) = F(xo,yo). If follows from properties of real multivariable limits that two- (2.9) (20.0 variable polynomial and rational functions are continuous at all points in their domains. limit of a complex function: Informally, the limit of f as z tends to zo exists and is equal to L, denoted lim f(z) = L, means that values of the complex function f(z) can be made arbitrarily close to the complex number L if values of 2 are chosen sufficiently close to, but not equal to, 29. The precise definition is similar to that of a real function: lim f(z) = L if for every € >0 there exists a 6 > 0 such that |f(2) — L| <¢ whenever 0 <|z — z9| <6. In this definition, both ¢ and 6 are real numbers, while 2, 29, {(2), and L are complex numbers. criterion for the nonexistence of a limit: In the limit of a complex function, z is allowed to approach zp from any direction in the complex plane. This provides a means of showing that certain complex limits do not exist: if f(z) approaches two complex numbers L; 4 Lo as z approaches 29 along two different. curves, then lim f(2) does not exist. real and imaginary parts of a limit: One method that is used to evaluate complex limits is to consider the real multivariable limits of the real and imaginary parts of f. If f(z) = u(x,y) + iv(z,y), 20 = to + imo, and L = uo + ivo, then Jim f(2)=L ifandonly if | lim | u(a.y) =o and lim | v(2,y) = vp. (2) (eo.w) (2,0) +o.) properties of complex limits: The following properties allow many complex limits to be evaluated in a somewhat mechanical fashion. Suppose that f and g are complex functions. If lim f(z) = L and lim g(z) = M, then (i) lim c= where c is a complex constant, (i) Jim 2 = 20, (iii) lim of (2) = cL, where c is a complex constant,(in) Jim (f(2) + 9(2)) = LM, (®) Bim f(2)-9(2) = LM, (vi) im & = provided M #0. continuity of a complex function: A complex function f is continuous at a point zo if lim f(2) = f(o) criteria for continuity at a point: A complex function f is continuous at a point 29 if each of the following three conditions hold: () lim f(z) exists, (ii) f is defined at 2, and (ii) Tim f(z) = F(20)- real and imaginary parts of a continuous function: If f(2) = u(x,y) + iv(x,y) and 29 = to + iyo, then the complex function f is continuous at the point 29 if and only if both real multivariable functions u and v are continuous at the point (to, yo)- properties of continuous functions: If f and g are contimious at the point zo, then ef (where ¢ is a complex constant), f +9, and f +g are continuous functions at the point z. If g() # 0, then f/g is also continuous at the point zo. From this it follows that all complex polynomial functions and all complex rational functions are continuous on their domains. a bounding property for continuous functions: If f is a continuous function defined on a closed and bounded region R, then f is bounded on R. That is, there is a real constant M > 0 such that |f(z)| yrs reas - rit velocity field force field complex representation of a vector field: The complex function f(z) = P(,y) + iQ(z,y) is called the complex representation of the vector field F(z,y) = P(x,y)i + Q(x,y)j- In general, both F(z,y) = P(z,y)i + Q(z,y)j and its complex representation f(z) = P(r,y) + iQ(#,y) are called vector ficlds. graphical representation: A graphical representation of a vector field f(z) = P(2,y)+iQ(c,y) is obtained by plotting the complex number f(z) as a vector based at the initial point = in the complex plane. fluid flow: A planar flow is the flow of a fluid in which the motion and physical traits of the fluid are the same in all planes parallel to the zy—plane. If f(z) represents the velocity of a particle of the fluid located at the point z in the complex plane, then f(z) is called a velocity field.streamlines: If z(t) = 2(¢) + iy(t) is a parametrization of a path that a particle follows in a planar fluid flow, then the derivative 2'(t) = 2'(t) + iy (t) represents the velocity of the particle at point z(t). Therefore, if f(z) = P(a,y) + iQ(z,y) is the velocity field of the fluid, then the following equations must hold: de G = Pew) di = Mey). The family of solutions to this system of differential equations is called the streamlines of the planar fluid flow with velocity field f(z). finding streamlines: One method for solving the system of differential equations dx G = Pew dy H = A,v). is to convert it to an ordinary differential equation using the chain rule dy dy dz 4, dy _ dy/dt dt dx dt dx dxjdt This gives the ordinary differential equation dy _ Q(z,y) dx Plry) Separation of variables and exact equations are two methods that can sometimes be used to solve this ‘equation. separation of variables: An ordinary differential equation of the form dy g(z)h(y) is called separable and can be solved by integration as follows: ® = gery) If H(y) and G(zr) are antiderivatives of 1/h(y) and g(z), respectively, then the equation Hy) = G(x) +e " gives solutions of the differential equation. exact equations: An ordinary differential equation of the form M(z,y)dr + N(x,y)dy = 060 Chapter 2 Complex Functions and Mappings is called exact if am _ on Oy oe If M, N, and their first partials are continuous, then this condition ensures that there is a function F(z, y) such that 2F/9z = M and OF/dy = N. We find F using partial integration: or M(z,y) F(z,y) = [Morapac +r) where h is found by differentiating Hy) = oF z [[ mcemar] S New) ~ 5 [f sornad]. 3. By evaluating f we find that FQ) f+i) f-i) fi) = (a) plotted as position vectors (b) plotted as vectors in the vector field7. By evaluating f we find that (a) plotted as position vectors (b) plotted as vectors in the vector field 11. (a) Letting 2 = e+iy we have f(z) = i(e+iy) = -y-+ix. By identifying P(x, y) = —y and Q(z, y) = in (3) of Section 2.7, we see that the streamlines are solutions to the system of differential equations dr dt dy dt = 2 ‘Phis system can be transformed into a single ordinary differential equation by using the chain rule dy _ dy dz dy _ dy/dt dt dx dt dx deja” Substituting dy/dt = x and dx/dt = —y into the second equation yields oe dirsChapter 2. Complex Functions and Mappings 62 This differential equation can be solved by separation of variables: Cea dry (b) Streamlines for Problem 11 Focus on Concepts 15. (a) Let c =a + bi, then solve the system a a dy Bao which can be converted into the ordinary differential equation: dy _b g¢- (b) Consider the magnitude and direction of the velocity ficld at an arbitrary point z. Do these quantities vary across the complex plane?Preface ‘This student study guide is designed to accompany the text A First Course in Compler Analysis with Appli- cations, Second Edition (Jones and Bartlett Publishers, 2009) by Dennis G. Zill and Patrick D. Shanahan. It consists of seven chapters which correspond to the seven chapters of the text. Each chapter has the following features. Rev nt Many sections of the study guide are preceded by a review of sclected topics from calculus and differential equations that are required for that section. These reviews provide concise summaries of prerequisite no- tation, terminology, and concepts. For additional review, students are encouraged to consult appropriate mathematics texts. Two excellent sources that were used repeatedly for the review topics are Calculus Early Transcendentals, Fourth Edition (Jones and Bartlett Publishers, 2010) by Dennis G. Zill and Warren S. Wright and Advanced Engineering Mathematics, Third Edition (Jones and Bartlett Publishers, 2006) by Dennis G. Zill and Michael R. Cullen. A summary of every section: of the text is provided. The summary reviews all of the key ideas of the section including all terminology, formulas, theorems, and concepts. Figures with two colors are included to aid in geometric understanding. Look ociad Following the summary, complete solutions are given for every other odd exercise in the section (eg. problems 3, 7, 11, ete,). These are full solutions, supported by figures with two colors, that supply all of the pertinent details of the problem and incorporate the same techniques and writing style used in the text. The solutions ‘also include references to equations, definitions, theorems, and figures in the text. The answer to each problem is typeset in color for easy reference. The focus on concepts problems from the text consist of conceptual word, proof, and geometrical problems, Since they are often used as topics for classroom discussion or independent study we have included detailed hints rather than full solutions for these problems. As with the exercises, only every other odd problem is included. Final Note to Students The most effective way to learn mathematics is to work many, many problems. You should not review a solution in this study guide before first working or, at the very least, attempting to work the problem yourself, Learning advanced mathematical topies takes significant time and effort. It may be quicker to look at the solutions, then try to work problems, but ultimately this approach will not Jead to an independent understanding of concepts and problem solving strategies that are required for success.Chapter 3 3.1 Differentiability and Analyticity derivative: If a complex function f is defined in a neighborhood of a point 29, then the derivative of f at zis '(%) = Jim, Leo A) = fea) aim, provided this limit exists. If f’(zo) exists, then f is said to be differentiable at. zo. differentiation rules: The familiar rules of differentiation in calculus also apply to complex functions. Constant Rules: de= 0 and fete) =ef'(2) Sum Rate: Hie) ta = se) 49) Product Rall trove = se) +/O012) f)] _ oS") ~ Se) Guotient ale abel: WFChain Rule: é S(g(2)) = S'(g(z))9'(2) d n= Power Rule: n2"-1, nan integer de analytic: A complex function f is said to be analytic at a point 2 if f is differentiable at 25 and at every point in some neighborhood of zo. analytic is differentiable: If f is analytic at zo, then f is also differentiable at zo, but the converse need not be true. That is, there do exist complex functions f such that f is differentiable at zo but not analytic at 2. entire: A function f is analytic in a domain D if f is analytic at every point in D. If f is analytic at every point in the complex plane, then f is called an entire function. polynomial and rational functions: Complex polynomial functions are entire, complex rational functions are analytic on their domains. analyticity rules: If f and g are analytic in a domain D, then the sum f(2)+9(z), the difference f(2)—9(=), and the product f(z)g(z) are analytic in D. The quotient f(z)/g(z) is analytic at all points where g(z) # 0. differentiable is continuous: If f is differentiable at zo, then f is also continuous at 29, but the converse need not be true. That is, there do exist complex functions f such that f is continuous at zo but not differentiable at =o L’Héspital’s rule: If f and g are analytic at zo, f(20) = 0, g(z0) = 0, and g'(zo) # 0, then £0) _ yy, LO) 3, gle) 2% ae)” 3. To find f"(z), we replace 2p with z in (1) of Section 3.1. Since f(z) = iz* f(2+Az)— f(z) = ile + Az)’ — 72+ Az)? - (iz? - 72) i(z3 + 327Az + 32z(Az)? + (Az)%) — 7(2? + 22Az + (Az)?) — iz9 + 72? = if +3iz2Az + 3iz(Az)? + i(A2)> — 7f — MdzAz - (Az)? iP +7? 3iz?Az + 3iz(Az)? + i(Az)? — 142Az — 7(Az)?. So, (1) gives , Biz*Az + Bi2(Az)? + i(Az)® — 14zAz — (Az)? f@ = bm, ‘Az nh Az(3iz? + 3izAz + i(Az)? - 14z — 7Az) = aro AzTherefore, ’(2) = 7. Since f(z) = 5z? — 102+ 8, So, (12) gives, (2) ~ f(20) = lm lim (3iz? + 3izAz + i(Az)? — 14z — 7Az) aro Biz? — 142. = 527-102 +8 — (523 — 10z9 + 8) 52? — 102+ B— 52 + 10z- 8 52? — 102 — 528 + 10z0. 52? — 102 — 522 + 1029 eae 70 = tim 32 = 28) = 10(2 ~ 20) Eto 2— 20 = jim = 20)(5( + 20) ~ 10) ro zz = im (6(2+ 20) ~ 10) = 5(z0 +20) -10 = 102-10. ‘Therefore, after replacing 2 with the symbol 2, f’(2) = 102 ~ 10. 11. If f(2) = (2— a) + izt — 32? +48, then f(z) ds, @-gFe +i (2-1) -524+i-423-3-22 Liq-o2]+4 fe] 41+ 20) — summa a 24-327 +0 + constant rules de + power rule (10 — 5i)z4 + 4iz* — 62. Therefore, f'(z) = (10 — 5i)24 + 4iz* — 62.15. If (2) = then 3241 — i) [iz? - 22] - (G2? - 22)¢ = fe) = (32+1-i)# [ie as 12) # [32 +1 - i] — quotient rule (32 +1 — i)(2iz — 2) — (42? — 22)(3) constant, sum, power rules @e+1 1? 2 tas Giz? + (—4 + 2é)z — 2+ 24 — Biz? + 62 (Bz+1-i)? Biz? + (2+ 2i)z-— 2+ 21 @zt+1-.? 2+ (24 2i)z2-2+2i @+1-) Therefore, f(z 19. (a) Identifying f(2) = and 29 = 0 in (1) of Section 3.1 gives $0) = Yn, Az! = lo? imo Az im 2? ao Az m 2 Be aro Az Sine f/(0) = 0, f is differentiable at the origin. (b)In order to show that f is not differentiable at any point z # 0 we use (1) of Section 3.1 and the critrion for the nonexistence of a limit from Section 2.6.1. Let 20 = 0 + iyo with zo # 0 or yo # 0. By/1) of Section 3.1 we have f'(20) = jim, Be T£ 4: approaches 0 along the real axis, then Az = Ax + 0i where the real number Az is approaching O. for this approach we have 0 S'(20) tm 20+ ito + Aa? — [oo + ivol? aro ‘Be23. jim 0+ Az)*+ Wh = 23~ Az0 Ar it fh + 2xoAx + Ax? #3 aro az = tim Azz + As) Ar0 Ar = Jm,@r0+a2 = 219. On the other hand, if If Az approaches 0 along the imaginary axis, then Az = 0+ iAy where the real number Ay is approaching 0. For this approach we have + Az)? —|zol? tim 20+ zl? = leo! Jo) = jim, Bz = lim |zco + iyo + iAy|? — |x0 + iyal? ayno iAy 2 2 yg 2 = tim H+ lve dy? B~WB Ar iAy = i + Av 8 azo iAy A = tim S¥@y+ Ay) azo iAy = tim, 2v0+ Ay aro i = —2iyo. Since both zo and yo are real numbers, 29 = —2iyo if and only if 9 = yo = 0. However, we started with the assumption that either x9 or yo is nonzero. Therefore, the limits along the two curves above are not equal, and consequently, f’(zo) does not exist. If we identify f(z) = 27 +i and g(z) = 24 +1, then f= +i i+i=0 and g(i)=i441=-14 7 So the limit lim ay has indeterminate form 0/0. Since f and g are polynomial functions, both are necessarily analytic at 29 = i. Using S'(2) = 728, g(z) = 1428, f'(i) = TH = g'(i) = 1413 = 143,wwe see that (13) of Section 3.1 gives lim Z+i tim ww _-7_1 PU ZM yp] ~ et da! ~ Tay 3" i2-2 | Be+1-i analytic when 32 + 1 numbers z such that 24 —3 + 3% it is analytic on its domain. So f fails to be Therefore, f is analytic for all complex 27. Since f(z) = hi. Now evaluate this limit using Theorem 2.6.2(i) and Definition 3.1.1 35. (a) If Az approaches 0, then r+ 0, but @ need not approach a limit. (b) Use the fact that 2 Dz (c) Consider different limiting values of 0 and use the criterion for the nonexistence of a limit. + Az to help simplify the limit. partial derivatives: If F(z,y) is @ function of two real variables, then the partial derivative of F with respect to x is oF = dim Oz Aro Ar Similarly, the partial derivative of F with respect to y is OF ig Pleeu + Av) ~ Fle.) Oy 4y-0 Ay 5 The partial derivatives OF/dx and OF/Oy are often represented by the alternative notation F, and Fy, respectively.computing partial derivatives: To compute 0F/dz, use the laws of ordinary differentiation while treating yas a constant. To compute OF /dy, use the laws of ordinary differentiation while treating « as a constant. second-order partial derivatives: Since 0F/dz and OF /dy are functions of two variables, we can compute partial derivatives of these functions. This gives four, second-order partial derivatives defined by oF _ oe Or _ 8 (oF Oz? ~ dz \ Or Ordy ~ Ox \ By @F _ 90 (aF or. 2a) ioe (3 a > &(% mixed second-order partial derivatives: If a function F(z, y) has continuous second-order partial deriva- tives, then the mixed second-order partial derivatives are equal, that is, @F OF Bady Dye" chain rule: If F(u,») is differentiable and if u = g(x,y) and v = f(z. y) have continuous first-order partial derivatives, then dF @F du AF dv dz dudz Ov Oz ees Cauchy-Riemann equations: Suppose that f(z) = u(x, y) +iv(z, y) is differentiable at a point z = x + iy. Then at (x,y) the first-order partial derivatives of u and v exist and satisfy the Cauchy-Riemann equations ou ang SU dr By dy Ox criterion for non-analyticity: If the Cauchy-Riemann equations fail be satisfied at one or more points z in a domain D, then f is not analytic in D. criteria for analyticity: If u(r,y) and v(z,y) are continuous and have continuous first-order partial derivatives in a domain D, and if u and v satisfy the Cauchy-Riemann at all points of D, then f(z) = u(t, y) + iv(x,y) is analytic in D. Notice that this statement tells us that the Cauchy-Riemann equations alone are not sufficient to ensure analyticity. formula for f’(z): If u(x, y) and v(z,y) are continuous and have continuous first-order partial derivatives in some neighborhood of a point z, and if u and v satisfy the Cauchy-Riemann at z, then f(z) = u(z,y)+iv(z,y) is differentiable at z andconstant functions: The Cauchy-Riemann equations can be used to establish the following results regarding constant functions. Suppose f(z) is analytic in a domain D. (i) If |{(2)| is constant in D, then f(z) =e in D, where c is a complex constant (ii) If f'(2) =O in D, then f(z) = c in D, where c is a complex constant. polar form: Suppose f(z) = u(r,@) + iv(r,6) is a complex function expressed in terms of a polar form z= r(cos0 + isin8) of the variable z. The polar form of the Cauchy-Riemann equations is 1dv av __1du and 5 = 1g ou or ‘A polar form of the derivative of f is 3. Ife set z= r+iy, then f(2) = Re(z) = Re(x +iy) = 2. We now identify u(x, y) = x and (x,y) ‘Conputing the first-order partial derivatives gives ou av a 7 1 a 7 0 au av ao 20 oy oy Sive the first Cauchy-Riemann equation du av aro Be 1 F9-5, satisfied for no points in the plane, we conclude that f(2) = Re(z) is nowhere analytic. 7. Fm f(z) =2? + y? we identify u(z,y) = 2? + y? and v(z,y) = 0. Computing the first-order partial Givatives gives ou Ov az -* 27° ou ov zs = Yy =~ = 0. oy ¥ oy “p first Cauchy-Riemann equation av be 2-9 By jatisfied if and only if « = 0. On the other hand, the second Cauchy-Riemann equation ou ov oy Onis satisfied if and only if y = 0. Therefore, the Cauchy-Riemann equations are satisfied only at the point 2 =0. Consequently, in any domain D there exists points in the domain where the Cauchy-Riemann equations fail to be satisfied. By the criterion for non-analyticity, this implies that f(z) nowhere analytic. 11. (a) From f(z) = e* —¥ cos 2ry + ie* ¥ sin Qey we identify u(x, y) = ce -” sin2ry. Computing the first-order partial derivatives gives 2e?*-¥" (x cos 2ry — ysin 2ry) ou aay? Le oy — ysi Be = 2et¥ (wcos dy — ysin2zy) ou ay = -267*-¥* (ycos ry + sin: ay ( y ry) ‘The Cauchy-Riemann equations oa Or ou ay —2e*-¥" (y cos 2ry + xsin 2ry) av oe ov oy 2 + y? is —¥* cos 2ry and v(x, y) = = 2c" -¥*(ycos 2zy + rsin 2ry) = 2e**-¥ (xcos 2ry — ysin 2ey). 0 a 2 oe are satisfied at every point in the complex plane. Since u, v, and their first-order partial derivatives are continuous at all points in the plane, it follows that f(=) = e*~»” cos2ry + ie?” sin 2ry is an entire function, (b) By formula (9) in Section 3.2, Fe for all z in the complex plane. 15. (a) From wwe identify 08 6 u(z,y) = Computing the first-order partial derivatives gives a 30 ou or ey or and sind v(z.y) = av _ sind or” ov = _ 6086 00 r a ~ 788 eos ~ 76 = 2e**-¥" (2 cos Ary — ysin 2ry) + i267 (y cos 2ry + xsin Zry)19, 23. ate satisfied at all points = r(cos6 + ising) such that r #0. Since w, v, and their first-order partial 6 o derivatives are continuous at all points z such that r # 0, it follows that f(z) = = = is analytic in the punctured complex plane, 2 # 0. (b) By formula (11) in Section 3.2, for all z in the complex plane. (:) From f(2) = 2? + y? + 2izy we identify u(x, y) fist-order partial derivatives gives a? +y? and v(z,y) = 2zy. Computing the ou av cue RS a oe Ou ov oy = wy oy Qn. “he first Cauchy-Riemann equation GS poof or Oy iisatisfied at all points in the complex plane, On the other hand, the second Cauchy-Riemann equation au av Oe Loy = ay = -S 3g ae satisfied if and only if y = 0. Therefore, the Cauchy-Riemann equations are satisfied at all points ach that y = 0, that is, at all points on the taxis. Since there is no neighborhood in which the (uchy-Riemann equations are satisfied, {(2) = 2? +y? + 2izy is nowhere analytic. However, since the v, and their first-order partial derivatives are continuous at all points in the complex plane, and since ‘and v satisfy the Cauchy-Riemann equations at every point on the z—axis, f(z) =2* +y? + 2iry is ifferentiable on the 2—axis. ¥) By formula (9) in Section 3.2, on the x—axis we have f(z) 2a + iy 2n. —y=0 on the r—axis 4) From (3) in Section 2.1, f(2) = e* = e* cosy + ie* siny, and so we identify u(x, y) = e* cosy and ‘r, y) =e sin y. Computing the first-order partial derivatives gives ou Ov Mey Bs esiny ae e* siny ou e* cos, ee = wu by Oy‘The Cauchy-Riemann equations E av ome = cos y a Bu eersiny as = ay US Oe are satisfied at every point in the complex plane. Since u, v, and their first-order partial derivatives are continuous at all points in the plane, it follows that f(z) = e* is an entire function. (b) By formula (9) in Section 3.2, f"(z) = e cosy + ie* siny, therefore, f’(2) = f(2). on Concepts 27. Let f(z) = u(z,y) + iv(z,y), then g(z) = f(2) = u(z,y) — iv(x,y). Since f is analytic, u and v satisfy the Cauchy-Riemann equations. Can U(z,y) = u(z,y) and V(z,y) = —v(c. y) also satisfy the Cauchy-Riemann equations? 31. If f(z) = u(e,y) + iv(e,y) is analytic and f'(2) = a+ ib where a and b are real constants, then by formula (9) of Section 3.2 we have S'(2) =a+ib= Partially integrate the equations du and b= oe ~ oe and then use the Cauchy-Riemann equations to determine formulas for u(r,y) and (2, y) a 35. To solve for uz in (12), multiply the first equation by cos 8, multiply the second equation by —} sin 8, and then sum the resulting equations. Modify this procedure to solve for vz in (13). Finally, substitute these expressions in for uz and vz in formula (9) in Section 3.2 3.3 Harmonic Functions Laplace’s equation: Let ¢(z,y) be a function of two real variables. The second-order partial differential equation &o #is oles Laplace’s equation. This equation appears in many applications in science and engineering. The o e# sum 55+ as js also denoted by the symbol V2¢. harmonic function: A function (cy) is called harmonic in a domain D if it has continuous first and second-order partial derivatives and satisfies Laplace's equation at all points in D. real and imaginary parts of an analytic function: If f(2) = u(x, ) + iv(2,y) is analytic in a domain D, then the functions u(z,y) and v(z,y) are both harmonic functions in D. harmonic conjugate: If two functions u(z,y) and v(z,y) are both harmonic in a domain D. end ifu and 1 satigy the Cauchy-Riemann equations in D, then v(1,y) is called a harmonic conjugate of u(x.) in D. Noticethat order is important; v is called a harmonic conjugate of 1, Put u is not called a harmonic conjugate if v. corresponding analytic function: If u(r,y) is a harmonic function in a domain D and if v(x,y) is a harmonic onjugate of u(z,y) in D, then the corresponding complex function f(z) = u(a,v) + iv(a,y) is analytic inD. . (a) tom the partial derivatives eu 2, 2 ts Bye 2\P wes that u satisfies Laplace's equation at every point in the complex plane Pu Bp Pu Ox? ‘Thefore, u is harmonic in C. hjt-01(2,) is @ harmonic conjugate, then u and v must satisfy the Cauchy-Riemann equations, and Sor xmust have Ou ov = 7 oy Ou av Fon eT Tee Pec integration of the first equation with respect: to the variable y gives v(ay) = [= dy = 2zy + h(a). ‘Tieartial derivative with respect to © of this last equation is & 2y + h'(2).‘When this result is substituted into the second Cauchy-Riemann equation we obtain —2y = —2y-h(2) (2) = 0 Az) = C where Cis a real constant. Therefore, v(x, y) = 2ry + C is a harmonic conjugate of u(z,y)- (c) The corresponding analytic function is f(z) = 2? — y? + i(2ry + C). (a) From the partial derivatives ou Pu = eF cosy + e*(zcosy — ysiny), = 2c* cosy + e*(ecosy — ysiny), or oe G2 e(—ycosy — rsiny — siny). os 2(—2 2cosy + ysiny) B= &(-yeosy - -siny), SE = e(-wcosy—2cos 2y ‘ycosy y v) ae cosy y+ysiny) we see that 1 satisfies Laplace's equation at every point in the complex plane ya oS + ot 2c cosy + €*(x cosy — ysiny) + e*(—xcosy — 2cosy + ysiny) = 2c" cosy +.xe* cosy — ye* sin y — xe” cosy — 2e” cosy + ye* siny 0. ‘Therefore, u is harmonic in C. (b) If v(z,y) is a harmonic conjugate, then u and v must satisfy the Cauchy-Riemann equations, and so we must have Cums oa aeee ny) = 2 Fe 7 cosy + e*(ceosy—ysiny) = ou 8 (-yeosy—zsiny—siny) = —% ay hie vsiean on" Partial integration of the first equation with respect to the variable y gives v(a,y) = [teeous e*(wcosy — ysiny)) dy =e [oosuay set f coswdy Jvsoudy — and e* treated as constants = esiny+ze*siny — (-reosy + / cosy a) ‘integration by parts for 3rd integral e siny + ze" siny + ye* cosy e siny + h(e) = ze siny + ye cosy + h(z).uu. The partial derivative with respect to z of this last equation is a = ze siny + e*siny + ye* cosy +h'(x). When this result is substituted into the second Cauchy-Riemann equation we obtain e*(—yoosy—rsiny—siny) = —(ve*siny +e siny + ye* cosy + h'(2)) —ye® cosy — ze" siny—e"siny = —ze" siny — e* siny — ye* cosy ~ h'(z) h(z) = 0 A(z) = C where C'is a real constant. Therefore, v(z,y) = ce" siny + ye* cosy + C is a harmonic conjugate of u(z,y). (c) The corresponding analytic function is {(2) = ¢*(x cosy — ysiny) + i(e*(xsiny + yeosy) + C). From the partial derivatives au Pu au eu uth ao LF , aa =0 os yt ae 0, ay r+2, ay? wwe see that 1 satisfies Laplace's equation at every point in the complex plane Ou | Pu 0+0=0. oy On? ‘Therefore, u is harmonic in C. If v(z, y) is a harmonic conjugate, then u and v must satisfy the Cauchy-Riemann equations, and so we must have mw oy 2 ba y “By ou Ov oy 74? = ae Partial integration of the first. cquation with respect to the variable y gives 1 v(z,y) = for 1)dy = av +y+ h(a). ‘The partial derivative with respect to x of this last equation is By = yn m7) ‘When this result is substituted into the second Canchy-Riemann equation we obtain z+2 = -hi(2) N(q) = -2-2 A(z) = f 2 —2)dx lia ha) = -p2?-22+0where C is a real constant. Therefore, v(z, y) = u(e,y). (c) The corresponding analytic function is f(z) = ry +2 + 2y—5+i(3y?+y—}327—2c+C). If dy? +y — 42? — 2 + C is a harmonic conjugate of, f(2i) = -1+ 54, then (02) +0420) - 544 (4a? +2- $10)? 20) +) = -1+5i since z= 2i,2=0 and y=2 -¥+i(4+C) = -Y+5i 4+C 5 Cc 1 Therefore, the analytic function satisfying the condition f(2i) = La =aytet2y—54i( 597+ 15, From the partial derivatives du _ ao au Ou ua 0, <%= Greos3e, St = Fe = 3T? cos3, Fz = Greos30, 55 we see that 1 satisfies the Laplace's equation in polar coordinates at every point in the complex plane excluding the origin peu, du, Pu or "Or OR = 1? (6rcos36) + r (3r? cos 36) — 9r° cos 38 = 6r° cos36 + 3r* cos 36 — 9r° cos 34 0. Therefore, u is harmonic in the punctured complex plane z # 0. Preece a 19. (a) By the chain rule we have ae a a(x? + y? + 27)? Next, by the product and chain rules % ot 3a? 1 eryt2 32? — 2? -y-Now repeat for 0°¢/dy? and 6¢/0z? and sum the resulting second-order partial derivatives. (b) Modify the procedure in part (a) [3.4 Applications gradient: If f(z, y) is a differentiable function of two real variables, then the vector-valued function is called the gradient of f. The symbols i and j here represent the standard basis vectors (1,0) and (0, 1), respectively. del differential operator: The gradient vector can also be interpreted as the result of applying the del (short for “delta”) differential operator to the function f(z, y)- level curves: For a function f(x,y), the curves defined by f(x,y) = c, where c is a constant, are called the level curves of f. orthogonality of the gradient: The gradient vector Vf(zo, yo) at a point (9.yo) is perpendicular (or, orthogonal) to the level curve of f passing through (z0,yo) divergence: If F(x, P(x, y)i + Q(a,y)j is a vector field, then the divergence of F is the scalar aP 4Q div F = — += NE Oe By ‘This quantity can also be represented as the dot product of the del operator with the vector field 8,8 aP , aQ div F=V-P = (214 23). (Pi+qj) = 2+ = iw F=V (2+) (Pi+ Qi) = 5 +5 If the vector field F is the velocity field of a fiuid, then the divergence represents the net volume of fluid flowing through an element of surface area per unit time. Therefore, div F(xo, yo) > 0 indicates the presence of a source of fuid flow at the point (zo, yo), whereas, div F(r9,4o) <0 indicates the presence of a sink of fluid flow at the point (0, 0). curl: If F(x,y) = P(,y)i + Q(a,y)j is a vector field, then the curl of F is the vector curl FThis vector can also be represented as cross product of the three-dimensional del operator with the three- dimensional vector field F = Pi + Qj + 0k as follows culF = VxF © Blom If the vector field F is the velocity field of a fluid, then the curl represents the tendency of the fluid to turn a paddlewheel inserted into the flow perpendicular to the zy-plane. gradient field and potential function: A vector field F(x, y) = P(x,y)i + Q(2,y)j is called a gradient field if there exists a function ¢(x,y) such that F = V¢. If F = V¢ is a gradient field, then $(z,y) is called a potential function for F. criterion for a gradient field: Let F = P(x, y)i+ Q(x, y)j be a vector field defined in a simply connected domain D. If P and Q have continuous first-order partial derivatives in D then F is a gradient field if and oP only if 2 = Fy throughout D (this is equivalent to V x F = 0in D). orthogonal families: If f(z) = u(z,y) + iv(2,y) is analytic in a domain D, then level curves u(z,y) = ¢, and v(z, y) = ¢2 are orthogonal families in D. This means that at any point 2 = 9 + iyo in D, the tangent line L; to the level curve u(z,y) = uo = u(zo, Yo) is perpendicular to the tangent line Lo to the level curve v(2,y) = vo = v(t0; Yo). gradient fields and Laplace’s equation: Laplace’s equation occurs naturally in several applications involving gradient fields. For example, if F represents the electric field intensity in a suitable domain D due to a collection of charges on the boundary of D, then Faraday’s law implies that F = —V¢ is a gradient field. (With the hypotheses of the criterion for a gradient field, this is equivalent to curl F = 0 in D.) If there are no charges in the domain, then Gauss’ law asserts that div F = 0. Thus, we have ee PLV.Pe __ (84a 22), (86, , 8.) _ 8 _ FH 0=div F=V-F=-V-Vo= (Sai gel) (314 38s) oer ‘This means that the electric field potential ¢ satisfies Laplace's equation V7¢ = 0 in D. For a gravitational force field F the situation is similar except it is the law of conservation of mechanical energy that ensures F = V@. If F represents the velocity field of a planar fluid flow then the flow is called incompressible ifdiv F: 0 and irrotational if curl F = 0. As above this implies that there is a velocity potential ¢ which satisti Laplace’s equation. Finally, if ¢ represents the temperature at point (x,y) at time ¢, then the heat, equatn states KV6+VK-V6= ope wher is the thermal conductivity, o is specific heat, and p is the density of the medium. When K is a constit and the temperature is independent of time (called the steady-state temperature), then VK = 0 and #/4t =0, and so the the heat equation reduces to Laplace’s equation V2¢ = 0. comjex potential: If a real potential function 9(c, y) satisfies Laplace's equation V7¢ = 0 in a domain D aniif $(z,y) is a harmonic conjugate of #(z,y) in D, then the function Q(z) = o(z,y) + (x,y) is amijtic in D and is called the complex potential function corresponding to . equiptential curves: If (2) = ¢(x,y) + iv(e,y) is a complex potential function in a domain D, then the leel curves (2, y) = ¢1 are called equipotential curves. The level curves $(z,y) = cr and (z,y) = c2 formin eethogonal family in D. applations of complex potentials: The following table summarizes some common applications of comix otential functions (2) = (z,y) + #¥(z,y). [pplication | Potential Function ¢ | Level Curves ¢ =; | Level Curves = ca cictrostatics | electrostatic potential | equipotential curves __| lines of force avitation | gravitational potential | equipotential curves _| lines of force Wat flow steady-state temperature | isotherms lines of heat flux fid flow | velocity potential equipotential curves | streamlines of flow Dirichit Problem: If D is a domain in the plane and g is a function defined on the boundary C of D, then thproblem of finding a function (zx, y) which (j) sisfies Laplace's equation V2¢ = 0 in D, and (ii) exals g on the boundary C of D is calla Di ichlet_ problem. Diriclit problem in an infinite strip: The Dirichlet problem illustrated below6 , Solve: Batt Bat -l with radius | 3 : Ber Ber If 1 =0, then the level curve consists of the single point > = 0. For the second family, if cp #0, then we complete the square in the variable y. _¥ _ x+y?= tty Lo 24 yf? = Payt to tout ag (4). 24(v+ 2) 2e Ye) * ‘Therefore, the level curves —y/(x? + y?) = c2, co # 0 are circles centered at BI- ole = gi with radius Ps If cy = 0, then the level curve consists of the single point z = 0. Since f(z) = 1/z is analytic for all # 0, these families of circles are orthogonal at all points z # 0. In the following figure, the level curves xr/(x? + y2) = ¢1 are shown in color and the level curves —y/(2? + y) = ¢2 are shown in black. Orthogonal families for f(z) = 1/2. 7. We identi u(z,y) = e-* cosy and v(z,y) = —e~* siny. First we use implicit differentiation to find 0 < product rule 0 «chain rule.1. a siny+e* 7 siny = 0 + product rule is -2 dy - siny+e-*cosy 2 = 0 «chain rule. az Solving for the derivative in the last equation yields dy _ siny dr cosy The product of the two slopes is cosy siny si Therefore, the level curves are orthogonal. cosy (a) We repeat the procedure from Example 2 in Section 3.4 with appropriate modifications. The shape of the domain D along with the fact that the two boundary conditions are constant suggest that a solution ¢ is independent of y, that is, ¢ is a function of x alone. With this assumption Laplace’s equation becomes &o ae We integrate this equation twice with respect to the variable x. &o tar = f odz 4 2.0 + after first integration do | rae = / adr o(z) = az+b. «after second integration The boundary conditions 4(0, y) = 50 and (1, y) = 0 allow us to solve for the coefficients a and b. In particular (0,y)=a(0)+b=50 =>» b=50. Given b = 50 we then have (1, ) = a(1) +50 = Therefore, (2, y) = 50x + 50. (b) If U(z,y) is a harmonic conjugate, then 4 and 1 must satisfy the Cauchy-Riemann equations, and so we must have Oo ONL a Oy a6 _ ob wy o> ae84 Chapter 3. Analytic Functions Partial integration of the first equation with respect to the variable y gives ~ [ s0ay = ‘The partial derivative with respect to x of this last equation is 2% ni) u(z,y) —50y + h(z). When this result is substituted into the second Cauchy-Riemann equation we obtain 0 h(x) —hi(2) c where C is a real constant. Therefore, (x.y) = —50y + C is a harmonic conjugate of o(r,y). A function is then given by Q(z) = 50x +50 + i(—50y + C) = —50(x + iy) +50 + iC. Replacing x + iy with the symbol 2 gives (z) = 502 + 50 + iC. 15. Set z =x + iy in f(z) and simplify in order to determine v(x, y), then factor the equation v(z,y) = 0 in order to determine the level curve.