Math 135-2, Homework 3

Solutions

Problem 53.4
Use the methods of both Examples 1 and 2 to solve each of the following differential equations:

(a) y ′′ + 5y ′ + 6y = 5e3t , y(0) = y ′ (0) = 0.

(a) First, lets use (13).

1
L[A(t)] =
p(p2 + 5p + 6)
1
=
p(p + 3)(p + 2)
B C D
= + +
p p+2 p+3
1 = B(p + 2)(p + 3) + Cp(p + 3) + Dp(p + 2)
1 = 6B p=0
1 = −2C p = −2
1 = 3D p = −3
1 1 1
L[A(t)] = − +
6p 2(p + 2) 3(p + 3)
1 1 1
A(t) = − e−2t + e−3t
6 2 3
f (t) = 5e3t
f ′ (t) = 15e3t
Z t
y(t) = A(t − τ )f ′ (τ ) dτ + f (0)A(t)
0
Z t   
1 1 −2(t−τ ) 1 −3(t−τ ) 3τ 1 1 −2t 1 −3t
= − e + e (15e ) dτ + 5 − e + e
0 6 2 3 6 2 3
Z t
5 3τ 15 −2t+5τ 5 5 5
= e − e + 5e−3t+6τ dτ + − e−2t + e−3t
0 2 2 6 2 3
 t
5 3τ 3 −2t+5τ 5 −3t+6τ 5 5 5
= e − e + e + − e−2t + e−3t
6 2 6 0 6 2 3
5 3t 3 3t 5 3t 5 5 5
= (e − 1) − (e − e ) + (e − e−3t ) + − e−2t + e−3t
−2t
6 2 6 6 2 3
1 3t −2t 5 −3t
= e −e + e
6 6

1
Next, lets repeat with (12).
1
L[h(t)] =
p2 + 5p + 6
1
=
(p + 3)(p + 2)
C D
= +
p+2 p+3
1 = C(p + 3) + D(p + 2)
1=C p = −2
1 = −D p = −3
1 1
L[h(t)] = −
p+2 p+3
h(t) = e−2t − e−3t
f (t) = 5e3t
Z t
y(t) = h(t − τ )f (τ ) dτ
0
Z t 
= e−2(t−τ ) − e−3(t−τ ) (5e3τ ) dτ
0
Z t
= 5e−2t+5τ − 5e−3t+6τ dτ
0
 t
5
= e−2t+5τ − e−3t+6τ
6 0
3t −2t 5 3t 5 −3t
=e −e − e + e
6 6
1 3t −2t 5 −3t
= e −e + e
6 6

Problem 53.8
The current I(t) in an electric circuit with inductance L and resistance R is given by the
equation (4) in Section 13:

dI
L + RI = E(t),
dt
where E(t) is the impressed electromotive force. If I(0) = 0, use the methods of this section
to find I(t) in each of the following cases:
(a) E(t) = E0 u(t)
(b) E(t) = E0 δ(t)
(c) E(t) = E0 sin ωt

2
 1
L[h(t)] =
Lp + R
1 1
=
L p + R/L
1
h(t) = e−Rt/L
L
Z t
I(t) = h(t − τ )E(τ ) dτ
0
Z t
1 −R(t−τ )/L
= e E(τ ) dτ
0 L
Z t
1
= e−Rt/L eRτ /L E(τ ) dτ
L 0
Z t
1
I(t) = e−Rt/L eRτ /L E0 u(τ ) dτ part (a)
L 0
E0 −Rt/L t Rτ /L
Z
= e e dτ
L 0
 t
E0 −Rt/L L Rτ /L
= e e
L R 0
 
E0 −Rt/L L Rt/L L
= e e −
L R R
E0  
= 1 − e−Rt/L
R
1 −Rt/L t Rτ /L
Z
I(t) = e e E0 δ(τ ) dτ part (b)
L 0
E0 −Rt/L
= e
L
Z t
1
I(t) = e−Rt/L eRτ /L E0 sin ωτ dτ part (c)
L 0
E0 −Rt/L t Rτ /L
Z
= e e sin ωτ dτ
L 0
Z t 
E0 −Rt/L
= e Im e(R/L+iω)τ dτ
L 0
 t !
E0 −Rt/L 1 (R/L+iω)τ
= e Im e
L R/L + iω 0
 
E0 −Rt/L R/L − iω 
(R/L+iω)t
= e Im e −1
L (R/L)2 + ω 2
E0   
= 2 2
e−Rt/L Im (R/L − iω) eRt/L cos ωt + ieRt/L sin ωt − 1
L((R/L) + ω )
 
E0 −Rt/L R Rt/L

Rt/L
= e e sin ωt − ω e cos ωt − 1
L((R/L)2 + ω 2 ) L
E0 
−Rt/L

= 2 R sin ωt − Lω cos ωt + Lωe
R + L2 ω 2

3
Problem 69.2
Show that f (x, y) = y 1/2
(a) does not satisfy a Lipschitz condition on the rectangle |x| ≤ 1 and 0 ≤ y ≤ 1.
(b) does satisfy a Lipschitz condition on the rectangle |x| ≤ 1 and c ≤ y ≤ d where 0 < c < d.

(a) Let y1 = 0 and y2 = ε.

√ √
f (x, y1 ) − f (x, y2 ) y1 − y2
=
y1 − y2 y1 − y2
1
=√ √
y1 + y2
1
=√
ǫ
which is unbounded.
(b) Noting y1 , y2 ≥ c > 0,
√ √
f (x, y1 ) − f (x, y2 ) y1 − y2
=
y1 − y2 y1 − y2
1
=√ √
y1 + y2
1
≤ √
2 c
=R

provides a bound.

Problem 69.4
Show that f (x, y) = xy 2
(a) satisfies a Lipschitz condition on the rectangle a ≤ x ≤ b and c ≤ y ≤ d.
(b) does not satisfy a Lipschitz condition on any strip a ≤ x ≤ b and −∞ ≤ y ≤ ∞.

(a) Note that |x| ≤ max(|a|, |b|) = A and |y| ≤ max(|c|, |d|) = C.

f (x, y1 ) − f (x, y2 ) xy12 − xy22

=
y1 − y2 y1 − y2
= |x(y1 + y2 )|
≤ |x|(|y1 | + |y2 |)
≤ 2AC

is a bound.
(b) Choose any x 6= 0 (possible unless a = b), y1 = 0, and y2 → ∞.

f (x, y1 ) − f (x, y2 ) xy12 − xy22

=
y1 − y2 y1 − y2
= |x(y1 + y2 )|
= |x|y2
→∞

4
is unbounded.

Problem A
The problem yy ′ = 1, y(0) = 0 seems like it should have no solution. Show that it actually has
two solutions. How is this possible? This demonstrates that plugging the initial conditions
into an ODE and producing a contradiction does not suffice to show that there is no solution.

yy ′ = 1
1 2
y =x+c
2 p
y = ± 2(x + c)
√
0 = y(0) = ± 2c
√
y = ± 2x
√ √
The two solutions are y = 2x and y = − 2x. Plugging y = 0 into yy ′ = 1 and deriving a contradiction
implicitly assumes that y ′ is finite, which it is not. In actuality, plugging in y = 0 produces 0 · ∞, which is
indeterminate.

Problem B
Consider the ODE x3 y ′ = 2y.
(a) Find all solutions if y(0) = 0.
(b) Find all solutions if y(0) = 1.

(a) First, lets find the general solution. The equation is separable.

x3 y ′ = 2y
y −1 y ′ = 2x−3
ln |y| = −x−2 + c0
y = c1 e−x
−2

This satisfies is a solution for any c1 . We have not lost any solutions by dividing by zero, since y = 0 is
captured by c1 = 0.
(b) We also know that y(0) = 1 is not possible since we have already worked out the general solution. Note
that it is not sufficient to plug x = 0 and y = 1 into the ODE to derive a contradiction, as demonstrated by
Problem A.

Problem C
Find the Lipschitz constant (or show that it does not have one) for each of the following
functions on the indicated interval. (The Lipschitz constant is a tight bound for the Lipschitz
condition.)
(a) cos x sin x, (−∞, ∞)
(b) | sin x|, (−∞, ∞)

Note that if f (x) is differentiable on some interval [a, b], then the Lipschitz constant L for that interval

5
is obtained by looking at its derivative.

f (a) − f (b)
= f ′ (c) for some c ∈ [a, b]
a−b
Thus, any value of this fraction that can be obtained is also obtained by the derivative somewhere in the
interval. What is more, the derivative is obtained in the limit a → c and b → c, so

f (x) − f (y)
L= sup
a≤x<y≤b x−y
= sup |f ′ (z)|
a≤z≤b

(a) This one is differentiable. L = max |f ′ (x)| = max | cos 2x| = 1.

x x
(b) From |x| = |y + (x − y)| ≤ |y| + |x − y| and |y| = |x + (y − x)| ≤ |x| + |y − x| we deduce ||x| − |y|| ≤ |x − y|.
Then,

f (x) − f (y) | sin x| − | sin y|

=
x−y x−y
sin x − sin y
≤
x−y
≤ sup | cos z|
z
=1

Now, I need to show that L = 1 is tight. This follows from

f (ǫ) − f (0) | sin ǫ| − | sin 0|

lim = lim
ǫ→0+ ǫ−0 ǫ→0 + ǫ−0
sin ǫ
= lim
ǫ→0+ ǫ
=1

Problem D
Derive the time delay rule

L[u(x − a)f (x − a)] = e−ap F (p).

For which choices a is this rule valid?

6
 Z ∞
L[u(x − a)f (x − a)] = e−px u(x − a)f (x − a) dx
Z0 ∞
= e−p(z+a) u(z)f (z) dz x=z+a
−a
Z ∞ Z 0
= e−p(z+a) u(z)f (z) dz + e−p(z+a) u(z)f (z) dz
0 −a
Z ∞ Z 0
= e−p(z+a) f (z) dz + e−p(z+a) u(z)f (z) dz
0 −a
Z ∞ Z 0
= e−ap e−pz f (z) dz + e−p(z+a) u(z)f (z) dz
0 −a
Z 0
−ap
=e F (p) + e−p(z+a) u(z)f (z) dz
−a

If a ≥ 0, then the remaining integral is over negative values of z, for which u(z) = 0. Thus, we will have the
desired identity. If a < 0, then the remaining integral will in general be nonzero, since all three factors will
generally be nonzero. Thus, the identity is true only for a ≥ 0.