Authors: Joe Benson, Denis Bashkirov, Minsu Kim, Helen Li, Alex Csar

Evans PDE Solutions, Chapter 2

Joe: 1, 2,11; Denis: 4, 6, 14, 18; Minsu: 2,3, 15; Helen: 5,8,13,17. Alex:10, 16
Problem 1. Write down an explicit formula for a function u solving the initial-value problem
!
ut + b · Du + cu = 0 on Rn × (0, ∞)
u = g on Rn × {t = 0}

Here c ∈ R and b ∈ Rn are constants.

Sol: Fix x and t, and consider z(s) := u(x + bs, t + s)

Then
ż(s) = b · Du + ut
= −cu(x + bs, t + s)
= −cz(s)

Therefore, z(s) = De−cs , for some constant D. We can solve for D by letting s = −t. Then,
z(−t) = u(x − bt, 0)
= g(x − bt)
= Dect
i.e. D = g(x − bt)e−ct
Thus, u(x + bs, t + s) = g(x − bt)e−c(t+s)
and so when s = 0, we get u(x, t) = g(x − bt)e−ct . !
Problem 2. Prove that Laplace’s equation ∆u = 0 is rotation invariant; that is, if O is an orthogonal
n × n matrix and we define
v(x) := u(Ox) (x ∈ R)
then ∆v = 0.
Solution:
Let y := Ox, and write O = (ai j ). Thus,
v(x) = u(Ox)
= u(y)

"n
where y j = i=1 a ji xi . This then gives that
n
∂v # ∂u ∂y j
=
∂xi j=1
∂y j ∂xi
#n
∂u
= a ji
j=1
∂y j
1
2

Thus,
 ∂v     ∂u 
 ∂x1  a11 . . . an1   ∂y 

 ..   .. .   . 1 
 .  =  .  
..   .. 
 ∂v     
∂xn
a1n . . . a nn
∂u
∂yn
 ∂u 
 ∂y1 
T 
 
= O  ... 
 ∂u 
∂yn
T
D x · v = O Dy · u

Now,
∆v = D x v · D x v
= (OT Dy u) · (OT Dy u)
= (OT Dy u)T OT Dy u
= (Dy u)T (OT )T OT Dy u
= (Dy u)T OOT Dy u
= (Dy u)T Dy u because O is orthogonal
= (Dy u) · (Dy u)
= ∆u(y)
=0

Problem 3. Modify the proof of the mean value formulas to show for n ≥ 3 that
* * + 1
1 1 1 ,
u(0) = gdS + − f dx,
nα(n)rn−1 ∂B(0,r) n(n − 2)α(n) B(0,r) |x|n−2 rn−2
provided


−∆u = f
 in B0 (0, r)


 u=g on ∂B(0, r).

Solution: Set *
1
φ(t) = u(y)dS (y), 0 ≤ t < r,
nα(n)tn−1 ∂B(0,t)
and * *
1 1
φ(r) = u(y)dS (y) = gdS .
nα(n)rn−1 ∂B(0,r) nα(n)rn−1 ∂B(0,r)
Then,
* , t+ 1 * *
t+ 1 , −1
φ'(t) = ∆u(y)dy = − f dy = f dy.
n α(n)tn B(0,t) n α(n)tn B(0,t) α(n)tn−1 B(0,t)
(See the proof of Thm2)
 3

Let $ > 0 be given.

* r * * r
1
(1) φ($) = φ(r) − φ'(t)dt = gdS − φ'(t)dt.
$ nα(n)rn−1 ∂B(0,r) $
Using integration by parts, we compute
* r * r *
1
− φ'(t)dt = f dydt
$ nα(n)t
n−1
$ B(0,t)
* r *
1 1
= f dydt
nα(n) $ tn−1 B(0,t)
* 2r * r 1 1 *
1 +1 1 1 ,
= f dy − f dS dt
nα(n) 2 − n tn−2 B(0,t) $ $ 2−nt
n−2
∂B(0,t)
*
+ r 1 * * * ,
1 1 1
= f dS dt − f dy + f dy
n(n − 2)α(n) $ tn−2 ∂B(0,t) rn−2 B(0,r) $ n−2 B(0,$)
+ * ,
1 1
=: I − n−2 f dy + J .
n(n − 2)α(n) r B(0,r)

Observe that *
1
J: f dy ≤ C · $ 2 , for some constant C > 0
$ n−2 B(0,$)
and * * r *
1 1
f (x)dx = dt f dS .
B(0,$) |x|n−2 0 ∂B(0,t) tn−2
3
As $ → 0, I + J → B(0,$) |x|1n−2 f (x)dx. Thus,
* r +* * ,
1 1 1
lim − φ'(t)dt = f (x)dx − f dy
$→0 $ n(n − 2)α(n) B(0,r) |x|n−2 rn−2 B(0,r)
* + 1
1 1 ,
= − f dx.
n(n − 2)α(n) B(0,r) |x|n−2 rn−2
Therefore, letting $ → 0, we have from (1)
* * + 1
1 1 1 ,
u(0) = φ(0) = gdS + − f dx.
nα(n)rn−1 ∂B(0,r) n(n − 2)α(n) B(0,r) |x|n−2 rn−2
!

Problem 4. We say v ∈ C 2 (Ū) is subharmonic if

−∆v ≤ 0 in U.
(a) Prove for subharmonic v that
!
v(x) ≤ v dy for all B(x, r) ⊂ U.
B(x,r)
(b) Prove that therefore maxŪ v = max∂U v.
(c) Let φ : R → R be smooth and convex. Assume u is harmonic and v := φ(u). Prove v is
subharmonic.
4

(d) Prove v := |Du|2 is subharmonic, whenever u is harmonic.

Solution.
"
(a) As in the proof of Theorem 2, set φ(r) := ∂B(x,r) v dS (y) and obtain
r
!
'
φ (r) = ∆v(y)dy ≥ 0.
n B(x,r)
For 0 < $ < r,
* r
φ' (s)ds = φ(r) − φ($) ≥ 0.
$
Hence, φ(r) ≥ lim φ($) = v(x). Therefore,
$→0
* * r 4* 5
1 1
!
v dy = v dy = v(z) dS (z) ds
B(x,r) α(n)rn B(x,r) α(n)rn 0 ∂B(x,s)
* r *
1 n−1 1 r n−1
= nα(n)s φ(s) ds ≥ n ns v(x) ds = v(x)
α(n)rn 0 r 0

(b) We assume that U ⊂ Rn is open and bounded. For a moment, we assume also that U is
connected. Suppose that x0 ∈ U is such a point that v(x0 ) = M := maxŪ v. Then for
0 < r < dist(x0 , ∂U),
!
M = v(x0 ) ≤ v dy ≤ M.
B(x0 ,r)
Due to continuity of v, an equality holds only if v ≡ M within B(x0 , r). Therefore, the set
u−1 ({M}) ∩ U = {x ∈ U|u(x) = M} is both open and relatively closed in U. By the connect-
edness of U, v is constant within the set U. Hence, it is constant within Ū and we conclude
that maxŪ v = max∂U v.

Now let {Ui |i ∈ I} be the connected components of U. Pick any x ∈ U and find j ∈ I
such that x ∈ U j . We obtain
v(x) ≤ max v = max v ≤ max v
Ū j ∂U j ∂U

and conclude that maxŪ v = max∂U v.

(c) For x = (x1 , ..., xn ) ∈ U and 1 ≤ i, j ≤ n,
∂2 v ∂2 ∂u ∂u ∂2 u
(x) = φ(u(x)) = φ'' (u(x)) · (x) · (x) + φ' (u(x)) · (x).
∂xi ∂x j ∂xi ∂x j ∂xi ∂x j ∂xi ∂x j
Since φ is convex, then φ'' (x) ≥ 0 for any x ∈ R. Recall that u is harmonic and obtain
# n 4 52 # n 4 52
'' ∂u '' ∂u
∆v = φ (u) · + ∆u = φ (u) · ≥ 0.
i=1
∂xi i=1
∂xi
n +
" ,
∂u 2
(d) We set v := |Du|2 = ∂xk
. For x = (x1 , ..., xn ) ∈ U and 1 ≤ i, j ≤ n,
k=1
#n 6 7
∂2 v ∂2 u ∂2 u ∂u ∂3 u
(x) = 2 (x) · (x) + (x) · (x) .
∂xi ∂x j k=1
∂x i ∂xk ∂x i ∂x j ∂xk ∂xi ∂x j ∂xk
 5

Therefore,
4 5 4 5
∂2 v #n
 ∂2 u 2 ∂u ∂ ∂2 u 
=2  + ·  ,
∂xi 2 ∂x ∂x ∂x ∂x ∂x 2 
k=1 i k k k i

# 4 ∂2 u 52 # n
∂u ∂ + , # 4 ∂2 u 52
∆v = 2 + · ∆u = 2 ≥ 0.
1≤i,k≤n
∂x i ∂x k k=1
∂x k ∂x k 1≤i,k≤n
∂xi ∂xk

!
Problem 5: Prove that there exists a constant C, depending only on n, such that
4 5
max |u| ≤ C max |g| + max | f |
B(0,1) ∂B(0,1) B(0,1)

whenever u is a smooth solution of



− , u = f
 in B0 (0, 1)


u = g on ∂B(0, 1).
M M
Proof: Let M := maxB(0,1) | f |, then we define v(x) = u(x) + 2n
|x|2 and w(x) = −u(x) + 2n
|x|2 . We
first consider v(x) . Note that
− , v = − , u − M = f − M ≤ 0.
So, v(x) is a subharmonic funcion.
From Problem 4 (b), we have
M
max v(x) = max v(x) ≤ max |g| + .
B(0,1) ∂B(0,1) ∂B(0,1) 2n
That is
1
max u(x) ≤ max v(x) ≤ max |g| + max | f |.
B(0,1) B(0,1) ∂B(0,1) 2n B(0,1)
Then, for w(x), we have
− , w = ,u − M = − f − M ≤ 0.
Again, we can get
M
max w(x) = max w(x) ≤ max |g| + .
B(0,1) ∂B(0,1) ∂B(0,1) 2n
i.e.
1
max −u(x) ≤ max w(x) ≤ max |g| + max | f |.
B(0,1) B(0,1) ∂B(0,1) 2n B(0,1)
Combining these two together, we finally proved the problem. !

Problem 6. Use Poisson’s formula for the ball to prove

r − |x| r + |x|
rn−2 u(0) ≤ u(x) ≤ rn−2 u(0)
(r + |x|)n−1 (r − |x|)n−1
whenever u is positive and harmonic in B0 (0, r). This is an explicit form of Harnack’s inequality.
6

Solution.
Since y ∈ ∂B(0, r), then |x − y| ≤| x| + r. Therefore,
*
r2 − |x|2 g(y)
u(x) = dS (y)
nα(n)r ∂B(0,r) |x − y|n
* *
r2 − |x|2 g(y) n−2 r − |x| 1
≥ dS (y) = r · g(y)dS (y)
nα(n)r ∂B(0,r) (r + |x|)n (r + |x|)n−1 nα(n)rn−1 ∂B(0,r)
r − |x| r − |x|
!
n−2
=r g(y)dS (y) = rn−2 u(0)
(r + |x|)n−1
∂B(0,r) (r + |x|)n−1
r+|x|
The inequality u(x) ≤ rn−2 (r−|x|)n−1 u(0) can be proven in a similar way. !
Problem 7. Prove Poisson’s formula for a ball: Assume g ∈ C(∂B(0, r)) and let
*
r 2 − x2 g(y)
u(x) = dS (y) for x ∈ B0 (0, r).
nα(n)r ∂B(0,r) |x − y|n

Show that
Proof.
Problem 8.
Let u be the solution of


,u = 0
 in Rn+


u = g on ∂Rn+

given by Poisson’s formula for the half-space. Assume g is bounded and g(x) = |x| for x ∈ ∂Rn+ ,
|x| le1. Show Du is not bounded near x = 0. (Hint: Estimate u(λenλ)−u(0) .)

Proof: From formula (33) on page 37, we have

*
2xn g(y)
u(x) = dy,
nα(n) ∂Rn+ |x − y|n

and u(0) = g(0) = 0. Thus, using hint, we get

*
u(λen ) − u(0) 2 g(y)
= dy
λ nα(n) ∂R+ |λen − y|n
n
* *
2 g(y) 2 g(y)
= dy + dy
nα(n) |y|≤1 ∂Rn+ |λen − y|
8 n nα(n) |y|>1 ∂Rn+ |λen − y|n
8

Taking absolute value on both sides, we have

99 u(λen ) − u(0) 99 99 2 * g(y) 99 2
*
|g(y)|
9 9≥ 9 dy 9 − dy
λ nα(n) |y|≤1 8 ∂Rn+ |λen − y|n nα(n) |y|>1 8 ∂Rn+ |λen − y|n
=I1 − I2 .
 7

Since g is bounded, so it is obvious that I2 is bounded and independent of λ. For I1 , in this case,
g(y) = |y|, so
*
2 |y|
I1 = dy
nα(n) |y|≤1 8 ∂Rn+ |λen − y|n
*
2 |y|
≥ dy
nα(n) |y|≤1 ∂Rn+ (λ + |y|)n
8

|y|
Note that for fixed y, (λ+|y|)n
is increasing when λ is decreasing to 0, so by Monotone Convergence
theorem, we have
*
2 |y|
lim dy
λ→0 nα(n) |y|≤1 8 ∂Rn (λ + |y|)n
* +

|y|
= n
dy
|y|≤1 ∂Rn+ |y|
8
*
|y|
= n
dy
Bn−1 (0,1) |y|
* 1 * * 1
1 1 n−2
= dr dS (y) = C r dr = ∞.
0 r
n−1 n−1
0 ∂Bn−1 (0,r) |y|

So, Du is unbounded near x = 0. !

Problem 10.
Suppose u is smootha nd solves ut − ∆u = 0 in Rn × (0, ∞).
(i) Show uλ (x, t) := u(λx, λ2 t) also solves the heat equation for each λ ∈ R.
(ii) Use (i) to show v(x, t) := x · Du(x, t) + 2tut (x, t) solves the heat equation as well.

(i) uλt (x, t) = λ2 ut (λx, λ2 t) and uλxi (x, t) = λu(λx, λ2 t) for each i. Then uλxi xi (x, t) = λ2 u xi (λx, λ2 t).
Consequently, ∆uλ = λ2 ∆u and uλt − ∆uλ = λ2 (ut − ∆u), so uλ solves the heat equation for
all λ ∈ R.
(ii) We differentiate u(λx, λ2 t) = u(λx1 , . . . , λxn , λ2 t) with respect to λ we get
#
xk u xk (λx1 , . . . , λxk , λ2 t) + 2λtut (λx1 , . . . , λxn , λ2 t) = x · D(λx, λ2 t) + 2tut (λx, λ2 t).
k

Taking λ = 1, we then have that v(x, t) = x · Du(x, t) + 2tut (x, t). u is smooth, so the second
derivatives of u(λx, λ2 t) are continuous, meaning the mixed partials are equal. Therefore,
vt − ∆v = ∂t∂λ
∂
u(λx, λ2 t) − ∆ ∂λ
∂
u(λx, λ2 t) = ∂λ∂t
∂
u(λx, λ2 t) − ∂λ
∂
∆u(λx, λ2 t) = ∂λ
∂
(uλt − ∆uλ ) = 0,
since uλ satisfies the heat equation for all λ. Thus v does as well.
2
Problem 11: Assume n = 1 and u(x, t) = v( xt ).
a) Show
ut = u xx
if and only if

(2) 4zv”(z) + (2 + z)v' (z) = 0 (z > 0)

8

b) Show that the general solution of (1) is

* z
v(z) = c e−s/4 s−1/2 ds + d
0

2
c) Differentiate v( xt ) with respect to x and select the constant c properly, so as to obtain the funda-
mental solution Φ for n = 1.

Solution:
a) Assume that ut = u xx . Then
4 5
x2 ' x2
ut = − 2 v
t t
and 5 4 4 25
x2
' 2 '' x
u xx = 2v + 4x v
t t
So ut = u xx implies that
4 5 4 25 4 25
x2 ' x2 ' x 2 '' x
− 2v = 2v + 4x v
t t t t
or 4 5 4 5 4 5
4x2 '' x2 2 x2 ' x2
v + + 2 v =0
t2 t t t t
x2
If we let z = t
, we get
4 5
4z '' 2 z '
v (z) + + v (z) = 0
t t t
Multiplying this equation by t gives the desired equality.
For the other direction, reverse the steps, and hence our proof is done.

b)
4zv'' + (2 + z)v' = 0
=⇒
v'' 11 1
= − −
v' 2z 4
=⇒
√ z
(by integrating) log(v' ) = − log z− +c
4
=⇒
v' = Cz−1/2 e−z/4
=⇒ * z
v=C e−s/4 s−1/2 ds + d
0
 9

as is desired.

c) * z
v(z) = c e−s/4 s−1/2 ds + d
0
=⇒
4 5 * xt2
x2
v =c e−s/4 s−1/2 ds + d
t 0
=⇒ 5 4 4 5−1/2
x2' 2x − x4t2 x2
v =c e
t t t
or 4 25
' x 2c x2
v = √ e− 4t
t t
Now we want to integrate over R and set the integral equal to 1. Thus we get
*
2c ∞ − x4t2
1= √ e dx
t ∞
Letting y = √x ,
we get dy = (4t)−1/2 dx and substituting, we get
4t
*
2c ∞ √ −y2
1= √ 4te dy
t ∞
or * ∞
2
1 = 4c e−y dy
∞
3∞ 2 √
−y
Employing the identity ∞ e dy = π and solving for c, we get
1
c= √
4 π
Thus,
5 4
x2
'
Φ(x, t) : = v
t
2c x2
= √ e− 4t
t
1 −x4t2
= √ e
2 πt
is easily shown to solve the equation
Φt = Φ xx
!
Problem 12. Write down an explicit formula for a solution of


ut − ∆u + cu = f
 in Rn x(0, ∞)


 u = g on Rn x{t = 0},
where c∈ R.
10

Solution: Set v(x, t) = u(x, t)eCt . Then, vt = ut eCt + CeCt u and v xi xi = u xi xi eCt .
⇒
vt − ∆v = ut eCt + CeCt u − eCt ∆u
= eCt (ut − ∆u + Cu)
= eCt f.
So, v is a solution of


vt − ∆v = eCt f
 in Rn x(0, ∞)


 v=g on Rn x{t = 0},

By (17) (p.51),
* * t*
v(x, t) = Φ(x − y, t)g(y)dy + Φ(x − y, t − s)eCs f (y, s)dyds
Rn 0 Rn

where Φ is the fundamental solution of the hear equation. Since v(x, t) = u(x, t)eCt , we have
+* * t* ,
Ct
u(x, t) = e Φ(x − y, t)g(y)dy + Φ(x − y, t − s)eCs f (y, s)dyds .
Rn 0 Rn

Problem 13: Given g : [0, ∞] → R, with g(0) = 0, derive the formula

* t
x 1 −x2
u(x, t) = √ e 4(t−s) g(s)ds, x > 0

4π 0 (t − s)3/2
for a solution of the initial/boundary-value problem



 ut − u xx = 0 inR+ × (0, ∞)



 u = 0 onR+ × {t = 0},


 u = g on{x = 0} × [0, ∞).
Proof. We define


u(x, t) − g(t)
 x > 0,
v(x, t) = 

−u(−x, t) + g(t) x ≤ 0.

So, we have


ut (x, t) − g' (t)
 x > 0,
vt (x, t) = 

−ut (−x, t) + g' (t) x ≤ 0,

and


u xx (x, t)
 x > 0,
v xx (x, t) = 

−u xx (−x, t) x ≤ 0.
 11

Hence,
 


 
−g' (t)
 x > 0,


 vt (x, t) − v xx (x, t) = 


 
g' (t) x ≤ 0.





 v(x, 0) = 0,


 v(0, t) = 0.
By formula (13) on page 49, we get
* t !* 0 * ∞ :
1 −(y−x)2
'
−(y−x)2
'
v(x, t) = √ e 4(t−s) g (s)dyds − e 4(t−s) g (s)dyds
0 4π(t − s) −∞ 0

Note that(page 46 Lemma) * ∞

1 −(y−x)2
√ e 4(t−s) dy = 1,
−∞ 4π(t − s)
so when x > 0, we let y − x = −z and obtain
u(x, t) = v(x, t) + g(t)
* t * ∞
' 1 −(y−x)2
= v(x, t) + g (s)ds √ e 4(t−s) dy
0 −∞ 4π(t − s)
* t * 0
1 1 −(y−x)2
=2 √ (t − s)− 2 e 4(t−s) dy g' (s)ds
0 4π −∞
* t * ∞ 2
1 1 −z
= √ (t − s)− 2 e 4(t−s) dz dg(s)
0 π x

Integrating by parts, we get

* ∞ 2
1 −z
u(x, t) = √ (t − s) −1/2
e 4(t−s) dz g(s)| s=t
s=0
π x
* t * ∞ 2
1 1 −z
− g(s) √ (t − s)−3/2 ds e 4(t−s) dz
0 π2 x
* t * ∞ 2
1 −z −z2
− g(s) √ (t − s)−1/2 ds e 4(t−s) dz
0 π x 4(t − s)2
* t * ∞ 2
1 1 −3/2 −z
= I1 − g(s) √ (t − s) ds e 4(t−s) dz
0 π2 x
* t * ∞
1 −z −z2
+ g(s) √ (t − s)−1/2 ds de 4(t−s)
0 π x 2(t − s)
* t * ∞ 2
1 1 −3/2 −z
= I1 − g(s) √ (t − s) ds e 4(t−s) dz
0 π2 x
* t
1 −z2
+ g(s) √ (t − s)−3/2 ds (−z) e 4(t−s) |z=∞ z=x
0 4π
* t * ∞ 2
1 1 −z
+ g(s) √ (t − s)−3/2 ds e 4(t−s) dz
0 π2 x
* t
x 1 −x 2
= I1 + √ e 4(t−s) g(s)ds.
4π 0 (t − s) 3/2
12

z2
Now, we focus on I1 and define w2 to be 4$
,
*
1 −1/2 ∞ −z4$2
I1 = lim+ √ $ e dz g(t − $)
$→0 π x
* ∞
1 2
= g(t) lim+ √ 2e−w dw = 0.
$→0 π x2 /4$
Thus, we proved
* t
x 1 −x2
u(x, t) = √ e 4(t−s) g(s)ds, x > 0.

4π 0 (t − s)3/2

Next, we need to show that

lim u(x, t) = g(t).
x→0+
Note that for any fixed δ > 0.
* t
x 1 −x2
lim+ u(x, t) = lim+ √ e 4(t−s) g(s)ds
x→0 x→0 4π t−δ (t − s)3/2
* t−δ
x 1 −x2
+ lim+ √ e 4(t−s) g(s)ds
x→0 4π 0 (t − s) 3/2
* t
x 1 −x2
= g(t) lim+ √ e 4(t−s) ds
x→0 4π t−δ (t − s)3/2
* δ
x 1 −x4s2
= g(t) lim+ √ e ds
x→0 4π 0 s3/2

For fixed x, we let s = x2 /w2 and get

* x2 /δ
x w3 −w4 2 −2x2
lim+ u(x, t) = g(t) lim+ √ e dw
x→0 x→0 2 π ∞ x3 w3
* ∞
1 −w2
= g(t) lim+ √ e 4 dw
x→0 π x2 /δ
* ∞
1 −w2
= g(t) √ e 4 dw = g(t).
π 0
Hence, we are done. !
Problem 14. We say v ∈ C12 (UT ) is a subsolution of the heat equation if
vt − ∆v ≤ 0 in UT .

(a) Prove for a subsolution v that

* *
1 |x − y|2
v(x, t) ≤ n v(y, s) dyds
4r E(x,t;r) (t − s)2
for all E(x, t; r) ⊂ UT .
(b) Prove that therefore maxŪT v = maxΓT v
 13

Solution.

(a) We may well assume upon translating the space and time coordinates that x = 0 and t = 0.
As in the proof of Theorem 3, set
* *
1 |y|2
φ(r) := n v(y, s) 2 dyds,
r E(r) s
n |y|2
ψ(y, s) := − log(−4πs) + + n log r
2 4s
and derive
* * n
' 1 2n #
φ (r) ≥ n+1 −4n∆vψ − vy yi dyds
r E(r) s i=1 i
# n * *
1 2n
= 4nvyi ψyi − vyi yi dyds = 0.
i=1
r n+1
E(r) s
For 0 < $ < r,
*r
φ' (z)dz = φ(r) − φ($) ≥ 0.
$
3 3 |y|2
Hence, φ(r) ≥ lim φ($) = v(0, 0) · lim $1n E($) s2
dyds = 4v(0, 0), and the statement follows.
$→0 $→0
(b) Suppose there exists a point (x0 , t0 ) ∈ UT with u(x0 , t0 ) = M := maxŪT u. Then for all
sufficiently small r > 0, E(x0 , t0 ; r) ⊂ UT . Using the result proved above, we deduce
* *
1 |x − y|2
M = v(x0 , t0 ) ≤ n v(y, s) dyds ≤ M,
4r E(x0 ,t0 ;r) (t − s)2
since
* *
1 |x0 − y|2
1= n dyds.
4r E(x0 ,t0 ;r) (t0 − s)2
Conclude that u|E(x0 ,t0 ;r) = M. The argument used in the proof of Theorem 4 will finish the
proof.

!
Problem 15.
(a) Show the general solution of the PDE u xy = 0 is
u(x, y) = F(x) + G(y)
for arbitrary functions F,G.
(b) Using the change of variables ξ = x + t, η = x − t, show utt − u xx = 0 if and only if uξη = 0.
(c) Use (a),(b) to rederive d’Alembert’s formula.
Solution:
(a) 3
u xy = 0 ⇒ u x = f (x) ⇒ u(x, y) = 3 f (x)dx + G(y)
uyx = 0 ⇒ uy = g(y) ⇒ u(x, y) = g(y)dy + F(x)
14

This implies u(x, y) = F(x) + G(y).

(b)
x = ξ+η
2
, y = ξ−η
;2 <
Define ũ := u ξ+η2
, ξ−η
2
1 1 1 1 1 1 1
ũξ = u x + ut and ũξη = u xx − u xt + utx − utt = (u xx − utt )
2 2 4 4 4 4 4
Hence, ũξη = 0 ⇔ utt − u xx = 0.

(c)
By (b), utt − u xx = 0 ⇒ uξη = 0, and u(ξ, η) = F(ξ) + G(η) by (a)
,i.e, u(x, y) = F(x + t) + G(x − t).
Since u(x, 0) = g, ut (x, 0) = h,
(3) u(x, 0) = F(x) + G(x) = g(x),
ut (x, 0) = F ' (x) − G' (x) = h(x)
Integration ⇒
* x
(4) F(x) − G(x) = h(y)dy + C, C:constant.
0
* x
1; <
(2) + (3); F(x) = g(x) + h(y)dy + C
2
*0 x
1; <
(2) − (3); G(x) = g(x) − h(y)dy − C
2 0
Thus,
* x+t * x−t
1; < 1; <
u(x, y) = F(x + t) + G(x − t) = g(x + t) + h(y)dy + C + g(x − t) − h(y)dy − C
2 0 2 0
* x+t * 0
1; <
= g(x + t) + h(y)dy + C + g(x − t) + h(y)dy − C
2 0 x−t
* x+t
1= > 1
= g(x + t) + g(x − t) + h(y)dy (x ∈ R, t ≥ 0).
2 2 x−t
!

Problem 16.
Assume E = (E 1 , E 2 , E 3 ) and B = (B1 , B2 , B3 ) solve Maxwell’s equations:
Et = curl B
Bt = − curl E
div B = div E = 0
Show that utt − ∆u = 0 where u = Bi or E i for i = 1, 2, 3.
Solution.
 15

curl(curl E) = curl(−Bt )
4 2 3 5
∂ B ∂2 B2 ∂2 B3 ∂2 B1 ∂2 B2 ∂B1
= − + ,− + ,− +
∂y∂t ∂z∂t ∂x∂t ∂z∂t ∂x∂t ∂y∂t
∂
= − curl B
∂t
∂
= − Et
∂t
∂2 E
=− 2
∂t
However, we also know that curl(curl E) = ∇(div E) − ∇2 E = −∇2 E. Then E i satisfies utt − ∆u = 0
for i = 1, 2, 3.
2
Similarly, curl(curl B) = curl Et = − ∂∂tB2 , and curl(curl B) = ∇(div B) − ∇2 B = −∇2 B, so Bi satisfies
utt − ∆u = 0 for i = 1, 2, 3.
Problem 17.(Equipartition of energy) Let u ∈ C 2 (R × [0, ∞)) solve the initial value problem for
the wave equation in one dimension:


utt − u xx = 0
 in R × (0, ∞)


u = g; ut = h on R × {t = 0}.
3∞
Suppose g, h have compact support. The kinetic energy is k(t) := 21 −∞ u2t (x, t)dx and the potential
3∞
energy is p(t) := 12 −∞ u2x (x, t)dx. Prove
(i) k(t) + p(t) is constant in t.
(ii) k(t) = p(t) for all large enough times t.
3∞ + ,
Proof. (i.) We define e(t) = k(t) + p(t) = 12 −∞ u2t + u2x dx. Since g, h have compact support, so
we have
*
d e(t) 1 ∞
= 2ut utt + 2u x u xt dx
dt 2 −∞
* ∞ * ∞
ut utt dx − u xx ut dx
−∞ −∞
* ∞
= ut (utt − u xx ) dx = 0.
−∞
Hence, e(t) ≡ e(0).
(ii.)By d’Alembert’s formula on page 68, we have
* x+t
1= > 1
u(x, t) = g(x + t) + g(x − t) + h(y)dy.
2 2 x−t
So,
1= ' > 1
ut = g (x + t) − g' (x − t) + [h(x + t) + h(x − t)] ,
2 2
and
1= ' > 1
ux = g (x + t) + g' (x − t) + [h(x + t) − h(x − t)] .
2 2
16

We assume that there exists a positive constant M so that [−M, M] ⊇ supp(g' ) and [−M, M] ⊇
supp(h).
Note that for a fixed t > M, −M ≤ x − t ≤ M ⇔ 0 < t − M ≤ x ≤ t + M and −M ≤ x + t ≤ M ⇔
−t − M ≤ x ≤ −t + M < 0.
Thus, when t > M :

(a) 0 < t − M ≤ x ≤ t + M.
Then we have
h(x + t) = g(x + t) = 0.
So,
1 1 1
u2t = g' (x − t)2 + h(x − t)2 − g' (x − t)h(x − t) = u2x .
4 4 2
(b) −t − M ≤ x ≤ −t + M < 0.
Then,
1 1 1
u2t = g' (x + t)2 + h(x + t)2 + g' (x + t)h(x + t) = u2x .
4 4 2
(c) Otherwise
g' (x + t) = g' (x − t) = h(x + t) = h(x − t) = 0.

So, combining all the cases, it is obvious that when t > M, k(t) = p(t). !
Problem 18. Let u solve
!
utt − ∆u = 0 in R3 × (0, ∞)
u = g, ut = h on R3 × {t = 0},
where g, h are smooth and have compact support. Show there exists a constant C such that
|u(x, t)| ≤ C/t (x ∈ R3 , t > 0).

Solution.

From the conditions it follows that there exist R, M > 0 such that spt g, spt h ⊂ B(0, R) and
g(y) ≤ M, |Dg(y)| ≤ M, h(y) ≤ M for any y ∈ R3 . Kirchhoff’s formula gives the solution of the
initial-value problem:
!
u(x, t) = th(y) + g(y) + Dg(y) · (y − x) dS (y).
∂B(x,t)

Denote by Σ the intersection ∂B(x, t) ∩ B(0, R). Observe that the area of Σ is not greater than the
area of the sphere ∂B(0, R). Then, for t > 0, we obtain
99! 99 9* 99
99 99 1 999 9
99 th(y) + Dg(y) · (y − x) dS (y)9 = 9 th(y) + Dg(y) · (y − x) dS (y)99
∂B(x,t) 9 4πt 9 ∂B(x,t)∩B(0,R)
2 9
*
1
≤ t · |h(y)| + |Dg(y)| · |y − x| dS (y)
4πt2 ∂B(x,t)∩B(0,R)
1 2 2R2 M
≤ · 4πR · (tM + tM) = .
4πt2 t
 17

For t > 1, using the same argument, we get

99! 99 9* 99
99 99 1 999 99 1 2 R2 M R2 M
99 g(y) dS (y) 99 = 9 g(y) dS (y) 99 ≤ · 4πR · M = ≤ .
∂B(x,t) 4πt2 9 ∂B(x,t)∩B(0,R) 4πt2 t2 t
Notice now that the area Σ is not greater than the area of the sphere ∂B(x, t). Then for 0 < t ≤ 1,
99! 99 9* 99
99 99 1 999 99 1 M
99 g(y) dS (y)9 = 9 g(y) dS (y) 99 ≤ · 4πt2 · M ≤ .
∂B(x,t) 9 4πt 2 9 ∂B(x,t)∩B(0,R) 4πt 2 t
Without loss of generality, we can take R > 1. Then, combining the estimates obtained above, we
2
conclude |u(x, t)| ≤ 3Rt M . !

Evans PDE Solutions, Chapter 5

Alex: 4, Helen: 5, Rob H.: 1

Problem 1.
Suppose k ∈ {0, 1, . . .}, 0 < γ < 1. Prove C k,γ (Ū) is a Banach space.
Solution:
1. First we show that || · ||C k,γ (Ū) is a norm, where we recall that
# #
||u||C k,γ (Ū) = ||Dα u||C(Ū) + [Dα u]C 0,γ (Ū) ,
|α|≤k |α|=k
and ! :
|u(x) − u(y)|
[u]C 0,γ (Ū) = sup .
x!y∈U |x − y|γ
For the sake of opaqueness we now omit subscripts on all norms unless it is unclear from context.
2. For any λ ∈ R we have first
|λu(x) − λu(y)| |u(x) − u(y)|
[λu] = sup = |λ| sup = |λ| [u] ,
x,y∈U |x − y|γ
x,y∈U |x − y|γ
and certainly
||Dα (λu)||C(Ū) = ||λDα u|| = |λ| · ||Dα u|| .
So
# #
||λu|| = ||Dα (λu)|| + [Dα (λu)]
|α|≤k |α|=k
# #
= |λ| ||Dα u|| + |λ| [Dα u]
|α|≤k |α|=k
= |λ| · ||u|| .

3. If u = 0 it is obvious that ||u|| = 0. On the other hand, ||u|| = 0 implies that

||Dα u||C(Ū) = 0
18

for every |α| ≤ k. In particular this is true for α = 0 so that the supremum of D0 u = u on U is 0, i.e.
u ≡ 0.
4. Finally we must prove the triangle inequality. We know the triangle inequality is true for the sup
norm || · ||C(Ū) . We can also see that for any α which makes sense
[Dα (u + v)] = [Dα u + Dα v] ≤ [Dα u] + [Dα v] .
Therefore we can easily conclude
# #
||u + v|| = ||Dα (u + v)|| + [Dα (u + v)]
|α|≤k |α|=k
# #
≤ (||Dα u|| + ||Dα v||) + ([Dα u] + [Dα v])
|α|≤k |α|=k
= ||u|| + ||v||.

5. We need only show that C k,γ (U) is complete. So let {um } be a Cauchy sequence. Then {um (x){
is a Cauchy sequence for every x, so define u to be the pointwise limit of the um . Now if V is any
bounded subset of U, then V̄ is compact, so that um ⇒ u uniformly on any V. Since the um are
uniformly continuous on V̄ by assumption, this implies that u is uniformly continuous on V̄ as well
(and so, a fortiori u ∈ C(U)). Therefore u ∈ C(Ū).
What we would really like would be to have u ∈ C k (Ū). But similar arguments show that u has
derivatives Dα u for all |α| ≤ k on U by restricting first to bounded subsets of U to find the derivatives
and then using uniform convergence on these subsets to show the derivatives must also be uniformly
continuous on bounded subsets since the Dα um were.
This leaves us with only showing that the norm of u is finite, so that in fact u ∈ C k,γ (U). But for
every n we have
# # |Dα un (x) − Dα un (y) − Dα u(x) + Dα u(y)|
||un − u|| = sup |Dα un (x) − Dα u(x)| + sup
|α|≤k x∈U |α|=k x,y∈U
|x − y|γ
 
 # # |D α
u n (x) − D α
u n (y) − D α
u m (x) + Dα
um (y)| 
= lim  sup |Dα un (x) − Dα um (x)| + sup 

m⇒∞
|α|≤k x∈U |α|=k x,y∈U |x − y|γ

= lim ||un − um ||.

m⇒∞

In particular, since {um } is Cauchy there is some N so that n, m ≥ N implies ||un − um || ≤ 1. Letting
m approach ∞, this implies that ||uN − u|| < 1. Now the triangle inequality applies to give
||u|| ≤ ||uN − u|| + ||uN || < 1 + ||uN || < ∞.
!
Problem 4.
EN
Assume U is bounded and U ⊂⊂ i=1 Vi . Show there exist C ∞ functions ζi (i = 1, . . . , N) such that


0 ≤ ζ1 ≤ 1, supp ζi ⊂ Vi i = 1, . . . , N

" N ζ = 1

 on U.
i=1 i

The functions {ζi }1N for a partition of unity.

 19
EN
Solution. Assume U is bounded and U ⊂⊂ i=1 Vi . Without loss of generality, we may assume
that the Vi are open, for if they are not, we can replace Vi by its interior. We note that, since U is
bounded, U is compact. Each x ∈ U has a compact neighbourhood N x contained in Vi for some i.
Then {N x◦ } is an open cover of U, which then has a finite subcover N x◦1 , . . . , N x◦n . We now let Fi be
the union of the N xk contained in Vi . Fi is the compact since it is the finite union of compact sets.
The C ∞ version of Urysohn’s Lemma (Folland, p.245) allows us to find smooth "n functions ξ1 , . . . , ξN
such that ξi = 1 on Fi and supp(ξi ) ⊂ Vi . Since the Fi cover U, U ⊂ {x : 1 ξi (x) > 0} and we can
"
use Urysohn again to find ζ ∈ C ∞ with ζ = 1 on U and supp(ζ) ⊂ {x : n1 ξi (x) > 0}. Now, we let
"N+1
ξN1 = 1 − ζ, so 1 ξi > 0 everywhere. We then take
ξi
ζi = "N+1
1 ξj
as our partition of unity.
Problem 5 (Helen) Prove that if n = 1 and u ∈ W 1,p (0, 1) for some 1 ≤ p < ∞, then u is equal a.e.
to an absolutely continuous function, and u' which exists a.e. belongs to L p (0, 1).
Proof. Since u ∈ W 1,p (0, 1), so by definition on page 242 and 244, we have some function v ∈
L p (0, 1) such that
* *
u Dφdx = − vφdx, ∀φ ∈ Cc∞ ((0, 1)) .
(0,1) (0,1)
Note that v ∈ L p (0, 1), so by Hölder’s inequality, we have 4v4L1 ≤ 4v4L p 414Lq < ∞, which means
v ∈ L1 (0, 1). Thus, we can define function f (x) on (0, 1) by the following formula
* x
1
f (x) = u( ) + v(t)dt, ∀x ∈ (0, 1).
2 1
2

According to the Fundamental Theorem of Calcalus, f is absolutely continuous. Now we will

prove u = f a.e.
By the definition of f , we have f ' = v a.e. So for any φ ∈ Cc∞ ((0, 1)) we get
* * *
'
f Dφdx = − f φdx = − vφdx.
(0,1) (0,1) (0,1)
Therefore,
*
( f − u) Dφdx = 0 ∀φ ∈ Cc∞ ((0, 1)) ,
(0,1)

which means u = f + const. And note that u( 21 ) = f ( 12 ), hence u = f a.e. So u' exists a.e. and
satisfy u' = v a.e., so u' ∈ L p (0, 1). !