Exercise Set 4.

5 103

45. The period of f (x) is 2π, so check f (0) = 3, f (2π) = 3 and the critical points. f  (x) = −2 sin x − 2 sin 2x =
−2 sin x(1+2 cos x) = 0 on [0, 2π] at x = 0, π, 2π and x = 2π/3, 4π/3. Check f (π) = −1, f (2π/3) = −3/2, f (4π/3) =
−3/2. Thus f has an absolute maximum on (−∞, +∞) of 3 at x = 2kπ, k = 0, ±1, ±2, . . . and an absolute mini-
mum of −3/2 at x = 2kπ ± 2π/3, k = 0, ±1, ±2, . . ..

47. Let f (x) = x − sin x, then f  (x) = 1 − cos x and so f  (x) = 0 when cos x = 1 which has no solution for 0 < x < 2π
thus the minimum value of f must occur at 0 or 2π. f (0) = 0, f (2π) = 2π so 0 is the minimum value on [0, 2π]
thus x − sin x ≥ 0, sin x ≤ x for all x in [0, 2π].

49. Let m = slope at x, then m = f  (x) = 3x2 − 6x + 5, dm/dx = 6x − 6; critical point for m is x = 1, minimum value
of m is f  (1) = 2.

51. lim f (x) = +∞, lim f (x) = +∞, so there is no absolute maximum value of f for x > 8. By Table 4.4.3 there
x→+∞ + x→8
2x(−520 + 192x − 24x2 + x3 )
must be a minimum. Since f  (x) = , we must solve a quartic equation to find the
(x − 8)3
critical points. But it is easy to see that x = 0 and x = 10 are real roots, and the other two are complex. Since
x = 0 is not in the interval in question, we must have an absolute minimum of f on (8, +∞) of 125 at x = 10.

53. The absolute extrema of y(t) can occur at the endpoints t = 0, 12 or when dy/dt = 2 sin t = 0, i.e. t = 0, 12, kπ,
k = 1, 2, 3; the absolute maximum is y = 4 at t = π, 3π; the absolute minimum is y = 0 at t = 0, 2π.

 b  b
55. f (x) = 2ax + b; critical point is x = − . f (x) = 2a > 0 so f − is the minimum value of f , but
2a 2a
  2  
b b b −b2 + 4ac b −b2 + 4ac
f − =a − +b − +c = thus f (x) ≥ 0 if and only if f − ≥ 0, ≥ 0,
2a 2a 2a 4a 2a 4a
−b2 + 4ac ≥ 0, b2 − 4ac ≤ 0.

57. If f has an absolute minimum, say at x = a, then, for all x, f (x) ≥ f (a) > 0. But since lim f (x) = 0, there is
x→+∞
some x such that f (x) < f (a). This contradiction shows that f cannot have an absolute minimum. On the other
1
hand, let f (x) = 2 . Then f (x) > 0 for all x. Also, lim f (x) = 0 so the x-axis is an asymptote, both
(x − 1)2 + 1 x→+∞
as x → −∞ and as x → +∞. But since f (0) = 12 < 1 = f (1) = f (−1), the absolute minimum of f on [−1, 1] does
not occur at x = 1 or x = −1, so it is a relative minimum. (In fact it occurs at x = 0.)
y
1

x
-3 -2 -1 1 2 3

Exercise Set 4.5

(b) y is as large as possible when x = 1/2.

3. A = xy where x + 2y = 1000 so y = 500 − x/2 and A = 500x − x2 /2 for x in [0, 1000]; dA/dx = 500 − x, dA/dx = 0
when x = 500. If x = 0 or 1000 then A = 0, if x = 500 then A = 125, 000 so the area is maximum when x = 500
ft and y = 500 − 500/2 = 250 ft.
104 Chapter 4

Stream

5. Let x and y be the dimensions shown in the figure and A the area, then A = xy subject to the cost condition
3(2x) + 2(2y) = 6000, or y = 1500 − 3x/2. Thus A = x(1500 − 3x/2) = 1500x − 3x2 /2 for x in [0, 1000].
dA/dx = 1500 − 3x, dA/dx = 0 when x = 500. If x = 0 or 1000 then A = 0, if x = 500 then A = 375, 000 so the
area is greatest when x = 500 ft and (from y = 1500 − 3x/2) when y = 750 ft.
Heavy-duty

Standard y

7. Let x, y, and z be as shown in the figure and A the area of the rectangle, then A = xy and, by similar triangles,
z/10 = y/6, z = 5y/3; also x/10 = (8 − z)/8 = (8 − 5y/3)/8 thus y = 24/5 − 12x/25 so A = x(24/5 − 12x/25) =
24x/5 − 12x2 /25 for x in [0, 10]. dA/dx = 24/5 − 24x/25, dA/dx = 0 when x = 5. If x = 0, 5, 10 then A = 0, 12, 0
so the area is greatest when x = 5 in and y = 12/5 in.

z
y
10
8

√ √
9. A = xy where √ x2 + y 2 = 202 = 400 so y =√ 400 − x√2 and A = x √400 − x2 for 0 ≤ x ≤ 20; dA/dx =
2(200 − x2 )/ 400 − x2 ,√dA/dx = 0 √
when x = 200 √
= 10 2. If x = 0, 10 2, 20 then A = 0, 200, 0 so the area is
maximum when x = 10 2 and y = 400 − 200 = 10 2.

10 y

11. Let x = length of each side that uses the $1 per foot fencing, y = length of each side that uses the $2 per foot
fencing. The cost is C = (1)(2x) + (2)(2y) = 2x + 4y, but A = xy = 3200 thus y = 3200/x so C = 2x + 12800/x
for x > 0, dC/dx = 2 − 12800/x2 , dC/dx = 0 when x = 80, d2 C/dx2 > 0 so C is least when x = 80, y = 40.

13. Let x and y be the dimensions of a rectangle; the perimeter is p = 2x + 2y. But√ A = xy thus y = A/x so
p = 2x + 2A/x for x > 0, dp/dx = 2√− 2A/x2 = 2(x 2 2 2 2 3
√ − A)/x , dp/dx = 0 when x = A, d p/dx = 4A/x > 0 if
x > 0 so p is a minimum when x = A and y = A and thus the rectangle is a square.

15. Suppose that the lower left corner of S is at (x, −3x). From the figure it’s clear that the maximum area of the
intersection of R and S occurs for some x in [−4, 4], and the area is A(x) = (8 − x)(12 + 3x) = 96 + 12x − 3x2 .
Exercise Set 4.5 105

Since A (x) = 12 − 6x = 6(2 − x) is positive for x < 2 and negative for x > 2, A(x) is increasing for x in [−4, 2]
and decreasing for x in [2, 4]. So the maximum area is A(2) = 108.

12

4
-8 -4 x 8
-3 x

-12

17. Suppose that the lower left corner of S is at (x, −6x). From the figure it’s clear that the maximum area of the
intersection of R and S occurs for some x in [−2, 2], and the area is A(x) = (8 − x)(12 + 6x) = 96 + 36x − 6x2 .
Since A (x) = 36 − 12x = 12(3 − x) is positive for x < 2, A(x) is increasing for x in [−2, 2]. So the maximum area
is A(2) = 144.

12

2
-8 -2 x 8
-6 x

-12

19. Let the box have dimensions x, x, y, with y ≥ x. The constraint is 4x + y ≤ 108, and the volume V = x2 y. If we
take y = 108 − 4x then V = x2 (108 − 4x) and dV /dx = 12x(−x + 18) with roots x = 0, 18. The maximum value
of V occurs at x = 18, y = 36 with V = 11, 664 in3 . The First Derivative Test shows this is indeed a maximum.

21. Let x be the length of each side of a square, then V = x(3 − 2x)(8 − 2x) = 4x3 − 22x2 + 24x for 0 ≤ x ≤ 3/2;
dV /dx = 12x2 − 44x + 24 = 4(3x − 2)(x − 3), dV /dx = 0 when x = 2/3 for 0 < x < 3/2. If x = 0, 2/3, 3/2 then
V = 0, 200/27, 0 so the maximum volume is 200/27 ft3 .

23. Let x = length of each edge of base, y = height, k = $/cm2 for the sides. The cost is C = (2k)(2x2 ) + (k)(4xy) =
4k(x2 + xy), but V = x2 y = 2000√ thus y = 2000/x2 so C = 4k(x2 + 2000/x) for x > 0, dC/dx = 4k(2x −
2000/x ), dC/dx = 0 when x = 1000 = 10, d2 C/dx2 > 0 so C is least when x = 10, y = 20.
2 3

25. Let x = height and width, y = length. The surface area is S = 2x2 + 
3xy where x2 y = V , so y = V /x2 and
2
S = 2x + 3V /x for x >0; dS/dx = 4x − 3V /x , dS/dx = 0 when x = 3 3V /4, d2 S/dx2 > 0 so S is minimum
2

3 3V 4 3 3V
when x = ,y= .
4 3 4

27. Let r and h be the dimensions shown in the figure, then the volume of the inscribed cylinder is V = πr 2 h. But
 2  
2 h 2 2 2 h2 2 h2 h3 dV
r + = R so r = R − . Hence V = π R − h = π R2 h − for 0 ≤ h ≤ 2R. =
 2 4 4 4 dh
3 dV √ √ 4π
π R 2 − h2 , = 0 when h = 2R/ 3. If h = 0, 2R/ 3, 2R then V = 0, √ R3 , 0 so the volume is largest
4 √ dh  3 3
when h = 2R/ 3 and r = 2/3R.
106 Chapter 4

h
h R
2

29. From (13), S = 2πr2 + 2πrh.  But V = πr2 h thus h = V /(πr2 ) and so S = 2πr2 + 2V /r for r > 0. dS/dr =
4πr − 2V /r, dS/dr = 0 if r = V /(2π). Since d2 S/dr 2 = 4π + 4V /r3 > 0, the minimum surface area is achieved
2 3

when r = 3 V /2π and so h = V /(πr2 ) = [V /(πr3 )]r = 2r.

31. The surface area is S = πr2 + 2πrh where V = πr2 h = 500 so h= 500/(πr2 ) and S = πr2 + 1000/r for r > 0;
dS/dr = 2πr − 1000/r2 = (2πr3 − 1000)/r2 , dS/dr = 0 when r = 3 500/π, d2 S/dr 2 > 0 for r > 0 so S is minimum
 500 500  π 2/3 
when r = 3 500/π cm and h = 2 = = 3 500/π cm.
πr π 500
r

33. Let x be the length of each side of the squares and y the height of the frame, then the volume is V = x2 y. The total
length of the wire is L thus 8x + 4y = L, y = (L − 8x)/4 so V = x2 (L − 8x)/4 = (Lx2 − 8x3 )/4 for 0 ≤ x ≤ L/8.
dV /dx = (2Lx − 24x2 )/4, dV /dx = 0 for 0 < x < L/8 when x = L/12. If x = 0, L/12, L/8 then V = 0, L3 /1728, 0
so the volume is greatest when x = L/12 and y = L/12.

1
35. Let h and r be the dimensions shown in the figure, then the volume is V = πr2 h. But r2 + h2 = L2 thus
3
1 1 dV 1 dV √
r2 = L2 − h2 so V = π(L2 − h2 )h = π(L2 h − h3 ) for 0 ≤ h ≤ L. = π(L2 − 3h2 ). = 0 when h = L/ 3.
3 3 dh 3 dh
√ 2π 3 √ 
If h = 0, L/ 3, 0 then V = 0, √ L , 0 so the volume is as large as possible when h = L/ 3 and r = 2/3L.
9 3

L
h

√ 1
37. The area of the paper is A = πrL = πr r2 + h2 , but V = πr2 h = 100 so h = 300/(πr2 ) and A =
3 
 90000 90000
πr r + 90000/(π r ). To simplify the computations let S = A , S = π r r2 + 2 4
2 2 4 2 2 2
= π2 r4 +
π r r2
dS 180000 4(π 2 6
r − 45000) 
for r > 0, = 4π 2 r3 − = , dS/dr = 0 when r = 6 45000/π 2 , d2 S/dr 2 > 0, so S and
dr r3 r3
 √  300  
hence A is least when r = 6 45000/π 2 = 2 3 75/π cm, h = 3
π 2 /45000 = 2 3 75/π cm.
π
Exercise Set 4.5 107

h L

1 2 r R Rh
39. The volume of the cone is V = πr h. By similar triangles (see figure) = √ , r = √ so
3 h 2
h − 2Rh 2
h − 2Rh
1 h3 1 h2 dV 1 h(h − 4R) dV
V = πR2 2 = πR2 for h > 2R, = πR2 , = 0 for h > 2R when h = 4R, by
3 h − 2Rh 3 h − 2R dh 3 (h − 2R)2√ dh
the first derivative test V is minimum when h = 4R. If h = 4R then r = 2R.

h−R
h2 − 2Rh

h
R

1
41. The revenue is R(x) = x(225 − 0.25x) = 225x − 0.25x2 . The marginal revenue is R (x) = 225 − 0.5x =
(450 − x).
 
2
Since R (x) > 0 for x < 450 and R (x) < 0 for x > 450, the maximum revenue occurs when the company mines
450 tons of ore.

43. (a) The daily profit is P = (revenue) − (production cost) = 100x − (100, 000 + 50x + 0.0025x2 ) = −100, 000 +
50x − 0.0025x2 for 0 ≤ x ≤ 7000, so dP/dx = 50 − 0.005x and dP/dx = 0 when x = 10, 000. Because 10,000 is
not in the interval [0, 7000], the maximum profit must occur at an endpoint. When x = 0, P = −100, 000; when
x = 7000, P = 127, 500 so 7000 units should be manufactured and sold daily.

(b) Yes, because dP/dx > 0 when x = 7000 so profit is increasing at this production level.

(c) dP/dx = 15 when x = 7000, so P (7001) − P (7000) ≈ 15, and the marginal profit is $15.

45. The profit is P = (profit on nondefective) − (loss on defective) = 100(x − y) − 20y = 100x − 120y but y =
0.01x + 0.00003x2 , so P = 100x − 120(0.01x + 0.00003x2 ) = 98.8x − 0.0036x2 for x > 0, dP/dx = 98.8 − 0.0072x,
dP/dx = 0 when x = 98.8/0.0072 ≈ 13, 722, d2 P/dx2 < 0 so the profit is maximum at a production level of about
13,722 pounds.

1
47. The area is (see figure) A = (2 sin θ)(4 + 4 cos θ) = 4(sin θ + sin θ cos θ) for 0 ≤ θ ≤ π/2; dA/dθ = 4(cos θ −
2
2
sin θ + cos θ) = 4(cos θ − [1 − cos2 θ] + cos2 θ) = 4(2 cos2 θ +√cos θ − 1) = 4(2 cos θ − 1)(cos θ√+ 1). dA/dθ = 0
2

when θ = π/3 for 0 < θ < π/2. If θ = 0, π/3, π/2 then A = 0, 3 3, 4 so the maximum area is 3 3.
4 cos θ

2 sin θ
2
θ

2 cos θ
4

cos φ
49. I = k 2
, k the constant of proportionality. If h is the height of the lamp above the table then cos φ = h/ and
108 Chapter 4

√ h h dI r2 − 2h2 dI √
= h2 + r2 so I = k 3 = k for h > 0, =k 2 , = 0 when h = r/ 2, by the first
(h2 2
+r ) √3/2 dh 2
(h + r ) 5/2 dh
derivative test I is maximum when h = r/ 2.
 
51. The distance between the particles is D = (1 − t − t)2 + (t − 2t)2 = 5t2 − 4t + 1 for t ≥ 0. For convenience,
we minimize D2 instead, so D2 = 5t2 − 4t + 1, dD2 /dt = 10t − 4, which 2 2 2
√ is 0 when t = 2/5. d D /dt > 0 so D
2

hence D is minimum when t = 2/5. The minimum distance is D = 1/ 5.

53. If P (x0 , y0 ) is on the curve y = 1/x2 , then y0 = 1/x20 . At P the slope of the tangent line is −2/x30 so its equation
1 2 2 3 3 3
is y − 2 = − 3 (x − x0 ), or y = − 3 x + 2 . The tangent line crosses the y-axis at 2 , and the x-axis at x0 .
x0 x0 x0 x0 x0 2
9 9
The length of the segment then is L = 4 + x20 for x0 > 0. For convenience, we minimize L2 instead, so
x0 4
9 9 dL 2
36 9 9(x6
0 − 8) √ d2 L2
L2 = 4 + x20 , = − 5 + x0 = , which is 0 when x6
0 = 8, x0 = 2. > 0 so L2 and hence L
x0 4 dx0 √ x0 2 2x50 dx20
is minimum when x0 = 2, y0 = 1/2.

dy 2x dm 2(3x2 − 1)
55. At each point (x, y) on the curve the slope of the tangent line is m = =− for any x, = ,
dx (1 + x2 )2 dx (1 + x2 )3
dm √ √
= 0 when x = ±1/ 3, by the first derivative test the only relative maximum occurs at x = −1/ 3, which is
dx √
the absolute maximum because lim m = 0. The tangent line has greatest slope at the point (−1/ 3, 3/4).
x→±∞

57. Let C be the center of the circle and let θ be the angle  P W E. Then  P CE = 2θ, so the distance along the shore
2θ 2 cos θ θ
from E to P is 2θ miles. Also, the distance from P to W is 2 cos θ miles. So Nancy takes t(θ) = + = +
8 2 4
π 1
cos θ hours for her training routine; we wish to find the extrema of this for θ in [0, ]. We have t (θ) = − sin θ, so
2 √ 4
π −1 1 −1 1 1 −1 1 15
the only critical point in [0, ] is θ = sin ( ). So we compute t(0) = 1, t(sin ( )) = sin ( )+ ≈ 1.0314,
2 4 4 4 4 4
π π
and t( ) = ≈ 0.3927.
2 8
π π
(a) The minimum is t( ) = ≈ 0.3927. To minimize the time, Nancy should choose P = W ; i.e. she should jog
2 8
all the way from E to W , π miles.
√
−1 1 1 1 15
(b) The maximum is t(sin ( )) = sin−1 ( ) + ≈ 1.0314. To maximize the time, she should jog
4 4 4 4
1
2 sin−1 ( ) ≈ 0.5054 miles.
4

P

2

W E
C

59. With x and y as shown in the figure, the maximum length of pipe will be the smallest value of L = x + y.
y x 8x 8x dL 128
By similar triangles = √ , y = √ so L = x + √ for x > 4, = 1− 2 ,
8 2
x − 16 2
x − 16 2
x − 16 dx (x − 16)3/2
Exercise Set 4.6 109

dL
= 0 when (x2 − 16)3/2 = 128, x2 − 16 = 1282/3 = 16(22/3 ), x2 = 16(1 + 22/3 ), x = 4(1 + 22/3 )1/2 , d2 L/dx2 =
dx
384x/(x2 − 16)5/2 > 0 if x > 4 so L is smallest when x = 4(1 + 22/3 )1/2 . For this value of x, L = 4(1 + 22/3 )3/2 ft.

y
8

x
x2 – 16

61. Let x = distance from the weaker light source, I = the intensity at that point, and k the constant of pro-
kS 8kS dI 2kS 16kS 2kS[8x3 − (90 − x)3 ]
portionality. Then I = 2 + if 0 < x < 90; = − + = =
x (90 − x)2 dx x3 (90 − x)3 x3 (90 − x)3
2
kS(x − 30)(x + 2700) dI dI
18 3 3
, which is 0 when x = 30; < 0 if x < 30, and > 0 if x > 30, so the intensity is
x (x − 90) dx dx
minimum at a distance of 30 cm from the weaker source.

dθ 12 2 10(24 − x2 )
63. θ = α − β = cot−1 (x/12) − cot−1 (x/2), = − 2
+ 2
= , dθ/dx = 0 when
√ √ dx 144 + x 4+x (144 + x2 )(4 + x2 )
x = 24 = 2 6 feet, by the first derivative test θ is maximum there.

10

␣
␪
␤ 2

65. The
√ total time required for the light to travel from A to P to B is t = (time from A to P )+ (time from P to B) =
x2 + a2 (c − x)2 + b2 dt x c−x √
+ , = √ −  but x/ x2 + a2 = sin θ1 and
v1 v2 dx 2
v1 x + a 2 2
v2 (c − x) + b 2

 dt sin θ1 sin θ2 dt sin θ1 sin θ2

67. s = (x1 − x̄)2 + (x2 − x̄)2 + · · · + (xn − x̄)2 , ds/dx̄ = −2(x1 − x̄) − 2(x2 − x̄) − · · · − 2(xn − x̄), ds/dx̄ = 0
1
when (x1 − x̄) + (x2 − x̄) + · · · + (xn − x̄) = 0, (x1 + x2 + · · · + xn ) − nx̄ = 0, x̄ = (x1 + x2 + · · · + xn ),
n
1
d2 s/dx̄2 = 2 + 2 + · · · + 2 = 2n > 0, so s is minimum when x̄ = (x1 + x2 + · · · + xn ).
n

69. If we ignored the interval of possible values of the variables, we might find an extremum that is not physically
meaningful, or conclude that there is no extremum. For instance, in Example 2, if we didn’t restrict x to the
interval [0, 8], there would be no maximum value of V , since lim (480x − 92x2 + 4x3 ) = +∞.
x→+∞

Exercise Set 4.6

(b) Positive, positive, speeding up.