Midterm 2 Solutions Version B

(2 points) 1. Carefully read the instructions on the front page of this exam.
(10 points) 2. Compute the antiderivative (Hint: if you use complex exponentials, make sure you end
up with real functions in your final answer)
Z
sin2 (x) cos(πx)dx

Solution: Method 1 (Complex exponentials)

As always, we start by writing sine and cosine in terms of complex exponentials and
expand:
e − e−ix 2 eπix + e−πix
Z Z ix
2
sin (x) cos(πx)dx = ( ) dx
2i 2
(eix − e−ix )2 eπix + e−πix
Z
= dx
−4 2
Z
1
=− (eix − e−ix )2 (eπix + e−πix )dx
8
Z
1
=− (e2ix + e−2ix − 2)(eπix + e−πix )dx
8
Z
1
=− e(2+π)ix + e(π−2)ix + e(2−π)ix + e(−2−π)ix − 2(eπix + e−πix )dx
8
Z
1
=− e(2+π)ix + e−(2+π)ix + e(π−2)ix + e−(π−2)ix − 2(eπix + e−πix )dx
8
Now we have two options: either rewrite these complex exponentials as sine and cosine,
then integrate, or integrate, then rewrite as sine and cosine afterward. To save space,
we’ll try the former:

e(2+π)ix +e−(2+π)ix +e(π−2)ix +e−(π−2)ix −2(eπix +e−πix ) = 2 cos((2+π)x)+2 cos((π−2)x)−4 cos(πx),

meaning
Z Z
1
2
sin (x) cos(πx)dx = − e(2+π)ix + e−(2+π)ix + e(π−2)ix + e−(π−2)ix − 2(eπix + e−πix )dx
8
Z
1
=− cos((2 + π)x) + cos((π − 2)x) − 2 cos(πx)dx
4
1 sin((2 + π)x) sin((π − 2)x) 2
=− [ + − sin(πx)] + C.
4 2+π π−2 π

Method 2 (Integration by parts)

Let u = sin2 (x), dv = cos(πx)dx. By the chain rule, du = 2 sin(x) cos(x)dx, which we
know by a trig identity (one we could remember using complex exponentials, for example,
or just memorize) is sin(2x). Since v = π1 sin(πx), we know that
Z Z
2 1 2 1
sin (x) cos(πx)dx = sin (x) sin(πx) − sin(2x) sin(πx)dx.
π π
Name and PID: Version B

To find this second integral, we have to integrate by parts twice (or use the angle addition
formulas, but that is essentially method 1). First let
1
u = sin(2x), dv = sin(πx) ⇒ du = 2 cos(2x)dx, v = − cos(πx).
π
Then
Z Z
1 2
sin(2x) sin(πx)dx = − sin(2x) cos(πx) + cos(2x) cos(πx)dx.
π π
Finally, we set
1
u = cos(2x), dv = cos(πx)dx ⇒ du = −2 sin(2x), v = sin(πx),
π
resulting in
Z Z
1 2
cos(2x) cos(πx)dx = sin(πx) cos(2x) + sin(2x) sin(πx).
π π
Plugging this into our earlier calculation
Z Z
1 1 2 1 2
sin(2x) sin(πx)dx = − sin(2x) cos(πx)+ [ sin(πx) cos(2x)+ sin(2x) sin(πx)dx],
π π π π π
hence
Z
4 1 2
(1 − ( 2 ) sin(2x) sin(πx)dx = ( sin(πx) cos(2x) − sin(2x) cos(πx)),
π π π
which implies
Z
4 1 2
sin(2x) sin(πx)dx = (1 − 2 )−1 ( sin(πx) cos(2x) − sin(2x) cos(πx))
π π π
2
π − 4 −1 1 2
=( ) ( sin(πx) cos(2x) − sin(2x) cos(πx))
π2 π π
π2 1 2
= 2 ( sin(πx) cos(2x) − sin(2x) cos(πx))
π −4π π
π 2
= 2 ( sin(πx) cos(2x) − sin(2x) cos(πx))
π −4 π
Finally, we can plug this back into our original integration by parts equation to get that
Z Z
2 1 2 1
sin (x) cos(πx)dx = sin (x) sin(πx) − sin(2x) sin(πx)dx
π π
1 1 2
= sin2 (x) sin(πx) − 2 ( sin(πx) cos(2x) − sin(2x) cos(πx))
π π −4 π
Name and PID: Version B

1 1 1 2 1
= · (1 − cos(2x)) sin(πx) − 2 · sin(πx) cos(2x) + 2 sin(2x) cos(πx)
π 2 π −4 π π −4
1 1 1 2 1
= sin(πx) − ( + 2 · ) sin(πx) cos(2x) + 2 sin(2x) cos(πx).
2π 2π π − 4 π π −4

To make this look more like the answer from the first method, we recall that
1 1 1 1
= ( − )
π2 −4 4 π−2 π+2
, and that finding a common denominator shows that
1 1 2 π
+ 2 · =
2π π − 4 π 2(π 2 − 4)
1 1 1
= ( + ).
4 π−2 π−2
Thus we can rewrite
1 1 2 1
−( + 2 · ) sin(πx) cos(2x) + 2 sin(2x) cos(πx)
2π π − 4 π π −4
as
1 1 1
[ (sin(2x) cos(πx) − cos(2x) sin(πx)) − (sin(2x) cos(πx) + cos(2x) sin(πx))]
4 π−2 π+2
1 sin((2 − π)x) sin((π + 2)x)
= ( − )
4 π−2 π+2
1 sin((π − 2)x) sin((π + 2)x)
=− ( − )
4 π−2 π+2
where the second line comes from the angle addition formulas. Finally, we get that

1 sin((π − 2)x) sin((π + 2)x)

Z
1
sin2 (x) cos(πx)dx = sin(πx) + ( − )+C
2π 4 π−2 π+2
1 sin((2 + π)x) sin((π − 2)x) 2
=− [ + − sin(πx)] + C
4 2+π π−2 π
as before.
 Name and PID: Version B

(10 points) 3. Compute the antiderivative

6x2 + 4x + 34
Z
dx
(x + 7)(x2 − 2x + 12)

Solution: Call the integrand f (x). Then f (x) is proper, since its numerator has degree
2 and its denominator has degree 3. Also, x2 − 2x + 12 is irreducible (we can see this
by completing the square or using the quadratic formula to check that its roots are not
real), so we know its partial fraction decomposition has the form

A Bx + C
f (x) = + 2 .
x + 7 x − 2x + 12
Finding a common denominator, this means that

6x2 + 4x + 34 = A(x2 − 2x + 12) + (Bx + C)(x + 7).

There are many ways to solve for A, B, and C from here.

Method 1 (Plug in values):

Set x = −7. Then

6 · 49 − 4 · 7 + 34 = A(72 + 2 · 7 + 12) + (C − 7B)(0),

so
300 = 75A;
ie,
A = 4.
Next, we can plug in x = 0 to find

34 = 4 · 12 + 7C
⇒ −14 = 7C
⇒ −2 = C.

Lastly, (or by comparing the x2 coefficients), we could set x = 1 to find that

44 = 44 + (B − 2) · 8
⇒ 0 = (B − 2) · 8
⇒ B = 2.

Method 2 (Compare coefficients):

Rewrite the first equation as

6x2 + 4x + 34 = A(x2 − 2x + 12) + (Bx + C)(x + 7)

= (A + B)x2 + (−2A + 7B + C)x + (12A + 7C),
Name and PID: Version B

which means A, B, C satisfy the linear equations

A + B = 6, −2A + 7B + C = 4, 12A + 7C = 34.

We can solve these to get

A = 4, B = 2, C = −2
again.

Now we can integrate:

6x2 + 4x + 34 2x − 2
Z Z
4
2
dx = + 2 dx
(x + 7)(x − 2x + 12) x + 7 x − 2x + 12
= 4 log |x + 7| + log |x2 − 2x + 12| + C,

where for the last equality we used the substitution u = x2 − 2x + 12 to transform

2x − 2
Z
dx
x2 − 2x + 12
into Z
1
du.
u
 Name and PID: Version B

(2 points) 4. (a) Find the general form of the partial fraction decomposition for the rational function

x3
(x + 11)4 (x2 + 29)2

but do not solve for the actual constants involved.

Solution:
x3 A B C D F x + G Hx + I
= + + + + +
(x + 11)4 (x2 + 29)2 (x + 11)4 (x + 11)3 (x + 11)2 x + 11 (x2 + 29)2 x2 + 29
 Name and PID: Version B

(8 points) (b) Compute the integral

Z∞
2
dx
(x2 + 1)2
0
(Hint: this integral converges; use trig substitution)

Solution: Finding an antiderivative of this was essentially an example in class. We

make the substitution
x = tan2 (θ)
to exploit the identity
tan2 (θ) + 1 = sec2 (θ).
As dx = sec2 (θ)dθ, the substitution becomes
sec2 (θ)
Z Z
2
dx = 2 dθ
(x2 + 1)2 (tan2 (θ) + 1)2
sec2 (θ)
Z
=2 dθ
(sec4 (θ)
Z
=2 cos2 (θ)
Z
1
=2 (1 + cos(2θ))dθ
2
1
= θ + sin(2θ) + C
2
= θ + cos(θ) sin(θ) + C.
If we wanted this as a function of x (not strictly necessary in this case because we could
just figure out what the bounds are for θ), we would draw a right triangle

√
x2 + 1
x

θ
1

which shows that

x 1
sin(θ) = √ , cos(θ) = √ .
2
x +1 2
x +1
Thus we have that
Z
2 x
dx = arctan(x) + 2 + C.
(x2 + 1) 2 x +1
Name and PID: Version B

Now we can calculate the definite integral:

Z∞ ZR
2 2
dx = lim dx
(x + 1)2
2 R→∞ (x2 + 1)2
0 0
 R
x
= lim arctan(x) + 2
R→∞ x +1 0
 
R
= lim arctan(R) + 2 − arctan(0)
R→∞ R +1
R
= lim arctan(R) + 2
R→∞ R +1
π
= +0
2
π
= .
2
 Name and PID: Version B

(6 points) 5. (a) Show that

Z∞
1
√ dx
x x2 − 81
9
converges.
Solution: Again, we can use trig substitution to find the antiderivative. If we let
x = 9 sec(θ), then
Z Z
1 9 sec(θ) tan(θ)
√ dx = p dθ
2
x x − 81 9 sec(θ) 81 tan2 (θ)
1 sec(θ) tan(θ)
= dθ
9 sec(θ) tan(θ)
1
= arcsec(θ) + C
9
1 x
= arcsec( ) + C.
9 9
It turns out that this improper integral is improper on both sides, because x2 −81 =
0 when x = 9. Thus, we need to take limits on both sides:
Z∞
1 1 R S
√ dx = ( lim arcsec( ) − lim+ arcsec( ))
x x2 − 81 9 R→∞ 9 S→9 9
9
1 π
= ( − 0)
9 2
π
= .
18
Therefore, the integral converges.
 Name and PID: Version B

(4 points) (b) Does

Z∞
1
p dx
x log(x)x2 − 81
9

converge? (Hint: 9 > e)

Solution: Notice that log(x) > 1, because (for example) x > 9 > e ∼ 2.7 and
taking log preserves inequalities. This means that

log(x)x2 − 81 > x2 − 81,

and the same is true when we take square roots:

p √
log(x)x2 − 81 > x2 − 81.

When we divide by both sides, the inequality flips, so

1 1
p <√ ,
2
log(x)x − 81 2
x − 81

meaning
1 1
p < √ .
x log(x)x2 − 81 x x2 − 81
Thus, by comparison to the integral from part a, which we just showed converges,
Z∞
1
p dx
x log(x)x2 − 81
9

converges as well.
Name and PID: Version B

Scratch paper