and Solutions
Igor Yanovsky
Contents
1 Trigonometric Identities
3 Separation of Variables:
Quick Guide
Quick Guide
5 FirstOrder Equations
5.1 Quasilinear Equations . . . . . . . . . . . . . . .
5.2 Weak Solutions for Quasilinear Equations . . . .
5.2.1 Conservation Laws and Jump Conditions
5.2.2 Fans and Rarefaction Waves . . . . . . . .
5.3 General Nonlinear Equations . . . . . . . . . . .
5.3.1 Two Spatial Dimensions . . . . . . . . . .
5.3.2 Three Spatial Dimensions . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
10
10
12
12
12
13
13
13
6 SecondOrder Equations
14
6.1 Classication by Characteristics . . . . . . . . . . . . . . . . . . . . . . . 14
6.2 Canonical Forms and General Solutions . . . . . . . . . . . . . . . . . . 14
6.3 WellPosedness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
7 Wave Equation
7.1 The Initial Value Problem . . . . . . . . . .
7.2 Weak Solutions . . . . . . . . . . . . . . . .
7.3 Initial/Boundary Value Problem . . . . . .
7.4 Duhamels Principle . . . . . . . . . . . . .
7.5 The Nonhomogeneous Equation . . . . . . .
7.6 Higher Dimensions . . . . . . . . . . . . . .
7.6.1 Spherical Means . . . . . . . . . . .
7.6.2 Application to the Cauchy Problem
7.6.3 ThreeDimensional Wave Equation .
7.6.4 TwoDimensional Wave Equation . .
7.6.5 Huygens Principle . . . . . . . . . .
7.7 Energy Methods . . . . . . . . . . . . . . .
7.8 Contraction Mapping Principle . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8 Laplace Equation
8.1 Greens Formulas . . . . . . . . . . . . . . . . . . . . .
8.2 Polar Coordinates . . . . . . . . . . . . . . . . . . . .
8.3 Polar Laplacian in R2 for Radial Functions . . . . . .
8.4 Spherical Laplacian in R3 and Rn for Radial Functions
8.5 Cylindrical Laplacian in R3 for Radial Functions . . .
8.6 Mean Value Theorem . . . . . . . . . . . . . . . . . . .
8.7 Maximum Principle . . . . . . . . . . . . . . . . . . .
8.8 The Fundamental Solution . . . . . . . . . . . . . . . .
8.9 Representation Theorem . . . . . . . . . . . . . . . . .
8.10 Greens Function and the Poisson Kernel . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
23
23
24
24
24
24
26
26
26
27
28
28
29
30
.
.
.
.
.
.
.
.
.
.
31
31
32
32
32
33
33
33
34
37
42
44
44
9 Heat Equation
45
9.1 The Pure Initial Value Problem . . . . . . . . . . . . . . . . . . . . . . . 45
9.1.1 Fourier Transform . . . . . . . . . . . . . . . . . . . . . . . . . . 45
9.1.2 MultiIndex Notation . . . . . . . . . . . . . . . . . . . . . . . . 45
9.1.3 Solution of the Pure Initial Value Problem . . . . . . . . . . . . . 49
9.1.4 Nonhomogeneous Equation . . . . . . . . . . . . . . . . . . . . . 50
9.1.5 Nonhomogeneous Equation with Nonhomogeneous Initial Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
9.1.6 The Fundamental Solution . . . . . . . . . . . . . . . . . . . . . 50
10 Schr
odinger Equation
52
54
12 Problems: Shocks
75
102
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
127
127
128
130
136
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
139
139
141
155
156
171
174
187
.
.
.
.
.
.
.
.
196
196
205
216
221
223
232
242
249
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
271
273
282
302
305
309
323
333
338
365
31 Laplace Transform
385
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
393
393
393
393
393
394
394
Trigonometric Identities
cos(a + b) = cos a cos b sin a sin b
cos(a b) = cos a cos b + sin a sin b
sin(a + b) = sin a cos b + cos a sin b
sin(a b) = sin a cos b cos a sin b
cos a cos b =
sin a cos b =
sin a sin b =
cos(a + b) + cos(a b)
2
sin(a + b) + sin(a b)
2
cos(a b) cos(a + b)
2
0
mx
nx
cos
dx =
cos
L
L
L
L
L
0
mx
nx
sin
dx =
sin
L
L
L
L
L
mx
nx
cos
dx = 0
sin
L
L
L
L
0
nx
mx
cos
cos
dx =
L
L
L
0
mx
nx
sin
dx =
sin
L
L
1
cos2 t =
2
1
sin2 t =
2
einx eimx dx =
0
0
cot t + 1 = csc t
sin x =
eix + eix
2
eix eix
2i
cosh x =
sinh x =
ex + ex
2
ex ex
2
d
cosh x = sinh(x)
dx
d
sinh x = cosh(x)
dx
cosh2 x sinh2 x = 1
du
+ u2
du
a2 u2
a2
1
u
tan1 + C
a
a
u
= sin1 + C
a
=
0
n = m
n=m
n = m
n=m
L
2
n = m
n=m
L
2
n = m
n=m
0
0
L
n = m
n=m
0
L
n = 0
n=0
x sin x cos x
2
2
x sin x cos x
cos2 x dx = +
2
2
tan2 x dx = tan x x
cos2 x
sin x cos x dx =
2
cos x =
einx dx =
1 + tan2 t = sec2 t
2
sin2 x dx =
z2
ez dz =
x2
x2
ln x
2
4
e 2 dz =
A=
a b
c d
,
1
=
det(A)
d b
c a
Eigenvalues n
n 2
L1 2
X(0) = X(L) = 0
Eigenfunctions Xn
(n 2 )
L
(n 12 ) 2
L
n 2
L
2n
2
L
X(0) = X (L) = 0
X (0) = X(L) = 0
X (0) = X (L) = 0
X(0) = X(L), X (0) = X (L)
X(L) = X(L), X (L) = X (L)
n 2
L
sin n
L x
(n 1 )
sin L2 x
(n 1 )
cos L2 x
cos n
L x
2n
sin L x
cos 2n
L x
n
sin L x
cos n
L x
n = 1, 2, . . .
n = 1, 2, . . .
n = 1, 2, . . .
n = 0, 1, 2, . . .
n = 1, 2, . . .
n = 0, 1, 2, . . .
n = 1, 2, . . .
n = 0, 1, 2, . . .
X X = 0
Boundary conditions
X(0) = X(L) = 0, X (0) = X (L) = 0
X (0) = X (L) = 0, X (0) = X (L) = 0
Eigenvalues n
n 4
L
n
4
L
Eigenfunctions Xn
sin n
L x
cos n
L x
n = 1, 2, . . .
n = 0, 1, 2, . . .
Separation of Variables:
Quick Guide
u = 0.
Laplace Equation:
X (x)
Laplace Equation:
(y)
Y
= .
X(x)
Y (y)
X + X = 0.
=
uxx + uyy + u = 0.
X Y
+
+ = 0. ( = 2 + 2 )
X
Y
Y + 2 Y = 0.
X + 2 X = 0,
X (t)
Y ()
=
= .
X(t)
Y ()
Y () + Y () = 0.
uxx + uyy + k2 u = 0.
utt uxx = 0.
Wave Equation:
T (t)
X (x)
=
= .
X(x)
T (t)
X + X = 0.
Y
X
=
+ k 2 = c2 .
X
Y
X + c2 X = 0,
Y + (k2 c2 )Y = 0.
( > 0)
X
= ,
X
T
T X
1 T
+
=
1
+
.
c2 T
c2 T
c2 T X
4th Order: utt = k uxxxx.
1 T
X
=
= .
X
k T
X X = 0.
Heat Equation:
T
T
= k
ut = kuxx .
X
= .
X
X = 0.
k
ut = uxxxx .
X +
4th Order:
uxx + uyy + k2 u = 0.
X
Y
=
+ k 2 = c2 .
Y
X
Y + c2 Y = 0,
utt uxx + u = 0.
X
T
=
= .
T
X
X X = 0.
X + (k2 c2 )X = 0.
10
FirstOrder Equations
5.1
Quasilinear Equations
Consider the Cauchy problem for the quasilinear equation in two variables
a(x, y, u)ux + b(x, y, u)uy = c(x, y, u),
with parameterized by (f (s), g(s), h(s)). The characteristic equations are
dy
= b(x, y, z),
dt
dx
= a(x, y, z),
dt
dz
= c(x, y, z),
dt
y(s, 0) = g(s),
z(s, 0) = h(s).
dy
In a quasilinear case, the characteristic equations for dx
dt and dt need not decouple from
the dz
dt equation; this means that we must take the z values into account even to nd
the projected characteristic curves in the xyplane. In particular, this allows for the
possibility that the projected characteristics may cross each other.
The condition for solving for s and t in terms of x and y requires that the Jacobian
matrix be nonsingular:
xs ys
= xs yt ys xt = 0.
J
xt yt
(5.1)
(5.2)
The characteristic projection in the xtplane1 passing through the point (s, 0) is the
line
x = h(s)t + s
along which u has the constant value u = h(s). Two characteristics x = h(s1 )t + s1
and x = h(s2 )t + s2 intersect at a point (x, t) with
t=
1
s2 s1
.
h(s2 ) h(s1 )
11
ux =
h (s)
1 + h (s)t
1
.
h (s)
The smallest t for which this happens corresponds to the value s = s0 at which h (s)
has a minimum (i.e.h (s) has a maximum). At time T = 1/h (s0 ) the solution u
experiences a gradient catastrophe.
5.2
5.2.1
12
(5.3)
b
ut (x, t) dx +
ut(x, t) dx
= (t)ul ((t), t) (t)ur ((t), t) +
a
f (ur ) f (ul )
.
ur ul
1
u2 = 0,
ut +
2
x
x x
x x
=
u
= .
we have f (u) = u, f u
t
t
t
t
For a rarefaction fan emanating from (s, 0) on xtplane, we have:
xs
ul ,
t f (ul ) = ul ,
u(x, t) = xs
ul xs
t ,
t ur ,
xs
ur ,
t f (ur ) = ur .
5.3
13
5.3.1
We need to complete to a strip. Find (s) and (s), the initial conditions for p(s, t)
and q(s, t), respectively:
We need to complete to a strip. Find 1 (s1 , s2 ), 2 (s1 , s2 ), and 3 (s1 , s2 ), the initial
conditions for p1 (s1 , s2 , t), p2 (s1 , s2 , t), and p3 (s1 , s2 , t), respectively:
s1
s1
s1
s1
h
f1
f2
f3
= 1
+ 2
+ 3
s2
s2
s2
s2
The characteristic equations are
dx2
dx3
dx1
= Fp1
= Fp2
= Fp3
dt
dt
dt
dz
= p1 Fp1 + p2 Fp2 + p3 Fp3
dt
dp1
dp2
dp3
= Fx1 p1 Fz
= Fx2 p2 Fz
= Fx3 p3 Fz
dt
dt
dt
6
6.1
14
SecondOrder Equations
Classification by Characteristics
Consider the secondorder equation in which the derivatives of secondorder all occur
linearly, with coecients only depending on the independent variables:
a(x, y)uxx + b(x, y)uxy + c(x, y)uyy = d(x, y, u, ux, uy ).
(6.1)
b b2 4ac
dy
=
.
dx
2a
b2 4ac > 0 two characteristics, and (6.1) is called hyperbolic;
b2 4ac = 0 one characteristic, and (6.1) is called parabolic;
b2 4ac < 0 no characteristics, and (6.1) is called elliptic.
These denitions are all taken at a point x0 R2 ; unless a, b, and c are all constant,
the type may change with the point x0 .
6.2
x, y, u, ux , uy),
uxx uyy = d(
, u, u, u ).
u + u = d(,
Example 1. Reduce to canonical form and nd the general solution:
uxx + 5uxy + 6uyy = 0.
Proof. a = 1, b = 5, c = 6 b2 4ac = 1 > 0
characteristics.
The characteristics are found by solving
51
3
dy
=
=
dx
2
2
to nd y = 3x + c1 and y = 2x + c2 .
(6.2)
hyperbolic two
15
(x, y) = 3x y, (x, y) = 2x y.
Let
x = 3,
x = 2,
y = 1,
y = 1.
u = f (),
u = F () + G(),
u(x, y) = F (3x y) + G(2x y),
General solution.
(6.3)
x = 0,
y = y,
y = 1.
= (yu + u )y = u + y(u y + u y ) + (u y + u y )
= u + y 2 u + 2yu + u .
u = 3 + f (),
u = 3 + f () + g(),
y2
y2
+x +g
+x ,
u(x, y) = y 3 + y f
2
2
General solution.
16
17
Problem (F03, #4). Find the characteristics of the partial dierential equation
xuxx + (x y)uxy yuyy = 0,
x > 0, y > 0,
(6.4)
and then show that it can be transformed into the canonical form
( 2 + 4)u + u = 0
whence and are suitably chosen canonical coordinates. Use this to obtain the general
solution in the form
g( ) d
u(, ) = f () +
1
( 2 + 4 ) 2
where f and g are arbitrary functions of and .
Proof. a = x, b = x y, c = y b2 4ac = (x y)2 + 4xy > 0 for x > 0,
y > 0 hyperbolic two characteristics.
The characteristics are found by solving
2x
dy
b b2 4ac
x y (x y)2 + 4xy
x y (x + y)
2x = 1
=
=
=
=
2y
dx
2a
2x
2x
= xy
2x
dx
dy
= ,
y
x
ln y = ln x1 + c2 ,
c2
y= .
x
y = x + c1 ,
Let = x y and = xy
x = 1,
x = y,
y = 1,
y = x.
(u )
d,
d =
u
2 + 4
1
ln u = ln (2 + 4) + g(),
2
1
ln u = ln (2 + 4) 2 + g(),
g()
u =
1,
(2 + 4) 2
g() d
u(, ) = f () +
1 ,
(2 + 4) 2
General solution.
18
6.3
19
WellPosedness
ux + uyy = 0;
b)
uxx + uyy = 0;
c)
uxx uyy = 0.
Prescribe data for each problem separately on the boundary of so that each of these
problems is wellposed. Justify your answers.
Proof. The initial / boundary value problem for the HEAT EQUATION is wellposed:
x , t > 0,
ut = u
u(x, 0) = g(x)
x ,
u(x, t) = 0
x , t > 0.
Existence  by eigenfunction expansion.
Uniqueness and continuous dependence on the data by maximum principle.
The method of eigenfunction expansion and maximum
principle give wellposedness for more general problems:
x , t > 0,
ut = u + f (x, t)
u(x, 0) = g(x)
x ,
u(x, t) = h(x, t)
x , t > 0.
It is also possible to replace the Dirichlet boundary condition u(x, t) = h(x, t) by a
Neumann or Robin condition, provided we replace n , n by the eigenvalues and eigenfunctions for the appropriate boundary value problem.
a) Relabel the variables (x t, y x).
We have the BACKWARDS HEAT EQUATION:
ut + uxx = 0.
Need to dene initial conditions u(x, 1) = g(x), and
either Dirichlet, Neumann, or Robin boundary conditions.
b) The solution to the LAPLACE EQUATION
u = 0
in ,
u=g
on
exists if g is continuous on , by Perrons method. Maximum principle gives uniqueness.
To show the continuous dependence on the data, assume
u2 = 0
in ,
in ,
u1 = 0
on ;
on .
u1 = g 1
u2 = g 2
20
Thus,
for
f C () and h C (), satisfying the compatibility condition h dS =
f dx. It is unique up to an additive constant.
c) Relabel the variables (y t).
The solution to the WAVE EQUATION
utt uxx = 0,
is of the form u(x, y) = F (x + t) + G(x t).
The existence of the solution to the initial/boundary value problem
u(0, t) = (t),
u(1, t) = (t)
t 0.
is given by the method of separation of variables
(expansion in eigenfunctions)
and by the parallelogram rule.
Uniqueness is given by the energy method.
Need initial conditions u(x, 0), ut (x, 0).
Prescribe u or ux for each of the two boundaries.
21
Thus, u1 = u2 .
x , t > 0,
ut = u
u(x, 0) = g(x)
x ,
u(x, t) = 0
x , t > 0.
Existence  by eigenfunction expansion.
Uniqueness and continuous dependence on the data by maximum principle.
22
23
Wave Equation
(7.1)
b b2 4ac
1
4c2
dt
=
=
= ,
2
dx
2a
2c
c
and thus
1
t = x + c1
c
= x + ct
and
1
x + c2 ,
c
= x ct,
t=
(7.2)
The general solution of (7.2) is u(, ) = F ()+G(), where F and G are C 1 functions.
Returning to the variables x, t we nd that
u(x, t) = F (x + ct) + G(x ct)
(7.3)
7.1
(7.4)
(7.5)
x
If we integrate the second equation in (7.5), we get cF (x) cG(x) = 0 h() d + C.
Combining this with the rst equation in (7.5), we can solve for F and G to nd
x
1
h() d + C1
F (x) = 12 g(x) + 2c
0x
1
1
G(x) = 2 g(x) 2c 0 h() d C1 ,
Using these expressions in (7.3), we obtain dAlemberts Formula for the solution
of the initial value problem (7.4):
1
1
u(x, t) = (g(x + ct) + g(x ct)) +
2
2c
x+ct
h() d.
xct
7.2
24
Weak Solutions
Equation (7.3) denes a weak solution of (7.1) when F and G are not C 2 functions.
Consider the parallelogram with sides that are
segments of characteristics. Since
u(x, t) = F (x + ct) + G(x ct), we have
u(A) + u(C) =
= F (k1 ) + G(k3) + F (k2 ) + G(k4 )
= u(B) + u(D),
which is the parallelogram rule.
7.3
(7.6)
u(x, t) =
nx
nx
a0 (t)
+
+ bn (t) sin
.
an (t) cos
2
L
L
n=1
7.4
Duhamels Principle
ut (x, 0) = 0.
an + n an = fn (t)
an (0) = 0
an (0) = 0
7.5
Utt c Uxx = 0
U (x, 0, s) = 0
Ut (x, 0, s) = f (x, s)
an + n an = 0
an (0, s) = 0
an (0, s) = fn (s)
u(x, t) =
an (t) =
an (t s, s) ds.
t
x+c(ts)
f (, s) d ds.
xc(ts)
25
(7.8)
by adding together dAlemberts formula and Duhamels principle gives the solution:
u(x, t) =
1
1
(g(x + ct) + g(x ct)) +
2
2c
x+ct
h() d +
xct
1
2c
t
0
x+c(ts)
xc(ts)
f (, s) d ds.
7.6
7.6.1
26
Higher Dimensions
Spherical Means
Mu (x, r) =
uxi (x + r) i dS =
r
n =1
i=1
n
so,
i=1
=
=
=
=
r
1
rx u(x + r) d =
x
u(x + r) d
( = r)
n <1
n
<1
r 1
x
u(x + ) d
(spherical coordinates)
n r n
 <r
r
1
n1
x
u(x + ) dS d
n r n1
0
=1
r
r
1
1
n1
M
(x,
)
d
=
n1 Mu (x, ) d.
n x
u
n1 x
n r n1
r
0
0
n1
2
Mu (x, r) = x Mu (x, r).
+
r 2
r r
Note that for a radial function u = u(r), we have Mu = u, so the equation provides the
Laplacian of u in spherical coordinates.
7.6.2
x Rn , t > 0,
ut (x, 0) = h(x)
(7.9)
x Rn .
We use Poissons method of spherical means to reduce this problem to a partial dierential equation in the two variables r and t.
27
Suppose that u(x, t) solves (7.9). We can view t as a parameter and take the spherical
mean to obtain Mu (x, r, t), which satises
1
1
2
M
(x,
r,
t)
=
u
(x
+
r,
t)dS
=
c2 u(x + r, t)dS = c2 Mu (x, r, t).
u
tt
t2
n =1
n =1
Invoking the Darboux equation, we obtain the EulerPoissonDarboux equation:
2
Mu (x, r, t) = c2
t2
2
n1
+
r 2
r r
Mu (x, r, t).
Mu
(x, r, 0) = Mh (x, r).
t
7.6.3
2
(x,
r,
t)
=
c
(x,
r,
t)
.
rM
rM
u
u
t2
r 2
For each xed x, consider V x (r, t) = rMu (x, r, t) as a solution of the onedimensional
wave equation in r, t > 0:
2
2 x
2
V
(r,
t)
=
c
V x (r, t),
t2
r 2
V x (r, 0) = rMg (x, r) Gx(r),
(IC)
(IC)
Vtx (r, 0)
x
(BC)
Gx (0) = H (0) = 0.
We may extend Gx and H x as odd functions of r and use dAlemberts formula for
V x (r, t):
1 x
1 r+ct x
G (r + ct) + Gx (r ct) +
H () d.
V x (r, t) =
2
2c rct
Since Gx and H x are odd functions, we have for r < ct:
r+ct
x
x
x
and
H () d =
G (r ct) = G (ct r)
rct
ct+r
H x () d.
ctr
After some more manipulations, we nd that the solution of (7.9) is given by the
Kirchhos formula:
1
t
t
u(x, t) =
g(x + ct)dS +
h(x + ct)dS .
4 t
4 =1
=1
If g C 3 (R3 ) and h C 2 (R3 ), then Kirchhos formula denes a C 2 solution of (7.9).
2
28
This problem is solved by Hadamards method of descent, namely, view (7.9) as a special
case of a threedimensional problem with initial conditions independent of x3 .
We need to convert surface integrals in R3 to domain integrals in R2 .
t
1
g(x1 + ct1 , x2 + ct2 )d1 d2
h(x1 + ct1 , x2 + ct2 )d1 d2
+
2t
2
u(x1 , x2 , t) =
4 t
4
1 12 22
1 12 22
12 +22 <1
12 +22 <1
If g C 3 (R2 ) and h C 2 (R2 ), then this equation denes a C 2 solution of (7.9).
7.6.5
Huygens Principle
Notice that u(x, t) depends only on the Cauchy data g, h on the surface of the hypersphere {x + ct :  = 1} in Rn , n = 2k + 1; in other words we have sharp signals.
If we use the method of descent to obtain the solution for n = 2k, the hypersurface
integrals become domain integrals. This means that there are no sharp signals.
The fact that sharp signals exist only for odd dimensions n 3 is known as Huygens
principle.
3
For x Rn :
1
f (x + t)dS = n1
f (x + y)dy
t
t
=1
yt
f (x + y)dy
yt
= tn1
f (x + t)dS
=1
7.7
29
Energy Methods
x Rn , t > 0,
x Rn ,
(7.10)
c2
uxi xi ut dx
=
n
Rn
=
Rn
i=1
Rn
i=1
ut (utt c2 u) dx = 0,
or
dE
dt
n
n
2
d 1
2
2
=
uxi dx =
uxi uxi t dx
ut + c
ut utt + c2
dt 2 Rn
Rn
i=1
i=1
ut utt + c2 u ut dx
=
n
R
u
2
ds
ut utt dx + c
ut
ut u dx
=
n
n
Rn
Rn
R
ut (utt c2 u) dx = 0.
=
Rn
7.8
30
x, y X.
31
Laplace Equation
in Rn
(8.1)
in Rn .
(8.2)
u(x) = g(x),
u(x)
= h(x),
n
u
+ u = ,
n
Neumann problem:
Robin problem:
8.1
x
x
x
Greens Formulas
u
ds
u v dx =
v
v u dx
u
v
u
ds = (vu uv) dx
v
n
n
u
ds =
u dx
n
u
ds
u2 dx =
u
u u dx
(8.3)
(v = 1 in (8.3))
(u = v in (8.3))
n = (n1 , n2 ) R2
ux vx dxdy =
vux n1 ds
vuxx dxdy
uxk v dx =
uvnk ds
uvxk dx
u 2 v dx =
u
u v v u dx =
v
ds
n
n = (n1 , . . ., nn ) Rn .
v
u
ds +
n
v
u
v
ds +
u
n
n
uv dx.
u
u
v
v
ds.
n
n
8.2
32
Polar Coordinates
Br (x0 )
Rn .
In particular
for each x0
d
f dx =
f dS
dr Br (x0 )
Br (x0 )
for each r > 0.
u = u(x(r, ), y(r, ))
x(r, ) = r cos
y(r, ) = r sin
ur = ux xr + uy yr = ux cos + uy sin ,
= ux x + uy y = ux r sin + uy r cos ,
urr = (ux cos + uy sin )r = (uxxxr + uxy yr ) cos + (uyx xr + uyy yr ) sin
= uxx cos2 + 2uxy cos sin + uyy sin2 ,
= (ux r sin + uy r cos )
1
u
r 2
= uxx cos2 + 2uxy cos sin + uyy sin2 + uxx sin2 2uxy cos sin + uyy cos2 1r (ux cos + uy sin )
= uxx + uyy 1r ur .
1
1
uxx + uyy = urr + ur + 2 u .
r
r
1 2
2
2
1
2
+
+
=
+
.
x2 y 2
r 2 r r r 2 2
8.3
8.4
2
1
u.
+
r 2 r r
In R3 :
2
n1
u.
+
r 2
r r
1
1
ru rr =
u = 2 r 2 ur r =
r
r
4
2
2
u.
+
r 2 r r
8.5
33
8.6
2
1
+
u.
r 2 r r
8.7
Maximum Principle
for all .
Proof. We may assume A = supx u(x) , so by continuity of u we know that
{x : u(x) = A} is relatively closed in . But since
n
u( + rx) dx,
u()
n x1
if u() = A at an interior point , then u(x) = A for all x in a ball about , so
{x : u(x) = A} is open. The connectedness of implies u() < A or u() A for
all .
The maximum principle shows that u C 2 () with u 0 can attain an interior
maximum only if u is constant. In particular, if is compact, and u C 2 () C()
satises u 0 in , we have the weak maximum principle:
max u(x) = max u(x).
x
8.8
34
(8.4)
(8.5)
(8.6)
Let
K =
K =
K =
1
w(r),
r
1
1
w 2 w,
r
r
1
2
2
w 2 w + 3 w.
r
r
r
or
Thus,
w = c1 r + c2 ,
c2
1
w(r) = c1 + .
K =
r
r
See the similar problem, F99, #2, where the fundamental solution for ( I) is
found in the process.
35
1
r w(r),
w = c1 r + c2 , which implied:
c2
.
r
Suppose v(x) 0 for x R and let = BR (0); for small > 0 let
= B (0).
K(x) is harmonic (K(x) = 0) in . Consider Greens identity ( =
B (0)):
K(x)
K(x)
v
v
v
dS +
v
dS.
K(x)v dx =
K(x)
K(x)
n
n
n
n
B(0)
=0, since v0 f or xR
K(x)v dx =
K(x)v dx.
lim
0
Since K(r) = c1 +
c2
is integrable at x = 0.
r
K(x)
+
c
42 max v 0, as 0.
1
n
B(0)
B(0) n
K(x)
dS =
v(x)
n
B(0)
c2
v(x) dS
2
B(0)
c2
c2
v(0)
dS
+
[v(x) v(0)] dS
=
2
2
B(0)
B(0)
c2
= 2 v(0) 42 + 4c2 max v(x) v(0)
xB(0)
0, (v is continuous)
that is K(r) =
5
1
4r
In R3 , for x = ,
K(x)
K() = c1 +
K(x)
n
c2
.
K()
c2
= 2,
r
36
Show that the Fundamental Solution of the Laplace Operator is given by.
1
if n = 2
2 log r
(8.7)
K(x) =
1
2n
if n 3.
(2n)n r
Proof. For v C0 (Rn ), we want to show
K(x) v(x) dx = v(0).
Rn
Suppose v(x) 0 for x R and let = BR (0); for small > 0 let
= B (0).
K(x) is harmonic (K(x) = 0) in . Consider Greens identity ( =
B (0)):
K(x)
K(x)
v
v
v
dS +
v
dS.
K(x)v dx =
K(x)
K(x)
n
n
n
n
B(0)
=0, since v0 f or xR
lim
K(x)v dx =
0
K(x)v dx.
Since K(r) is integrable at x = 0.
K(x)
dS =
n
1
v(x) dS
n1
n
B (0)
1
1
v(0) dS +
[v(x) v(0)] dS
=
n1
n1
B (0) n
B(0) n
1
n1
v(x) v(0)
v(0)
max
=
n
n n1
xB(0)
0, (v is continuous)
= v(0).
Thus,
K(x)v dx = lim
0
1
2
K() =
K()
=
r
K(x)v dx = v(0).
log
1
2n
(2n)n
K(x)
n
1
2
1
n n1
if n = 2
if n 3.
if n = 2
if n 3,
1
,
n n1
8.9
37
Representation Theorem
Representation Theorem, n = 3.
Let be bounded domain in R3 and let n be the unit exterior normal to . Let
u C 2 (). Then the value of u at any point x is given by the formula
1
1
u(y)
1 u(y)
1
u(y)
dS
dy. (8.8)
u(x) =
4 x y n
n x y
4 x y
Proof. Consider the Greens identity:
w
u
u
w
dS,
(uw wu) dy =
n
n
dy =
dS
(8.9)
u(y)
n x y x y n
x y
1
1 u(y)
dS.
(8.10)
u(y)
+
n x y x y n
B(x)
We will show that formula (8.8) is obtained by letting 0.
1
u(y)
u(y)
dy =
dy.
Since
is integrable at x = y.
lim
0
x y
x y
x y
The rst integral on the right of (8.10) does not depend on . Hence, the limit as 0
of the second integral on the right of (8.10) exists, and in order to obtain (8.8), need
1
1 u(y)
dS = 4u(x).
u(y)
lim
0 B(x)
n x y x y n
B(x)
1
1 u(y)
1 u(y)
dS =
dS
u(y)
u(y)
2
n x y x y n
n
B(x)
1 u(y)
1
1
dS
u(x) dS +
[u(y) u(x)]
=
2
2
n
B(x)
B(x)
1 u(y)
1
dS.
[u(y) u(x)]
= 4u(x) +
2
n
B (x)
7
The last integral tends to 0 as 0:
1 u(y)
1
dS
[u(y) u(x)]
2
n
B(x)
38
1
u(y)
dS
u(y) u(x) +
n
B(x)
B (x)
4 max u(y) u(x) + 4 max u(y) .
yB(x)
y
1
2
0, (u continuous in )
and
1
1
= 2.
n x y
0, (u is f inite)
39
Representation Theorem, n = 2.
Let be bounded domain in R2 and let n be the unit exterior normal to . Let
u C 2 (). Then the value of u at any point x is given by the formula
u(y)
1
1
log x y log x y
dS.
(8.11)
u(y) log x y dy +
u(y)
u(x) =
2
2
n
n
Proof. Consider the Greens identity:
w
u
(uw wu) dy =
u
w
dS,
n
n
u(y)
log x y log x y
dS
u(y)
=
n
n
u(y)
log x y log x y
dS.
u(y)
+
n
n
B (x)
We will show that formula (8.11) is obtained by letting 0.
u(y) log x y dy = u(y) log x y dy.
lim
0
since log x y is integrable at x = y.
The rst integral on the right of (8.12) does not depend on . Hence, the limit as 0
of the second integral on the right of (8.12) exists, and in order to obtain (8.11), need
u(y)
log x y log x y
dS = 2u(x).
u(y)
lim
0 B(x)
n
n
B (x)
u(y)
u(y)
1
log x y log x y
dS =
u(y) log
dS
u(y)
n
n
n
B(x)
u(y)
1
1
u(x) dS +
[u(y) u(x)] log
dS
=
n
B(x)
B(x)
u(y)
1
[u(y) u(x)] log
dS.
= 2u(x) +
n
B(x)
8
The last integral tends to 0 as 0:
u(y)
1
[u(y) u(x)] log
dS
n
B(x)
40
u(y)
dS
u(y) u(x) + log
n
B(x)
B(x)
2 max u(y) u(x) + 2 log max u(y) .
yB (x)
y
1
0, (u continuous in )
and
log x y = .
n
0, (u is f inite)
41
Representation Theorems, n > 3 can be obtained in the same way. We use the
Greens identity with
w(y) =
1
,
x yn2
n
n
u(y)
K(x y)
K(x y)
dS.
(8.15)
u(y)
+
n
n
B (x)
We will show that formula (8.13) is obtained by letting 0.
K(x y)u(y) dy = K(x y)u(y) dy.
since K(x y) is integrable at x = y.
lim
0
The rst integral on the right of (8.15) does not depend on . Hence, the limit as 0
of the second integral on the right of (8.15) exists, and in order to obtain (8.13), need
u(y)
K(x y)
K(x y)
dS = u(x).
u(y)
lim
0 B(x)
n
n
42
u(y)
u(y)
K(x y)
K()
K(x y)
dS =
K()
dS
u(y)
u(y)
n
n
n
n
B (x)
B(x)
K()
u(y)
K()
dS +
[u(y) u(x)] K()
dS
u(x)
=
n
n
n
B(x)
B (x)
1
u(y)
1
dS
u(x) dS
[u(y) u(x)] dS
K()
=
n1
n1
n
n
n
B(x)
B (x)
B(x)
1
u(y)
1
n1
dS.
u(x)n
[u(y) u(x)] dS
K()
=
n1
n1
n
n
B (x)
B(x)
n
u(x)
1
n n1
u(y)
dS
K()
n
B(x)
B(x)
max u(y) u(x)n n1 + K() max u(y)n n1 .
yB (x)
y
0, (u is f inite)
0, (u continuous in )
8.10
With a slight change in notation, the Representation Theorem has the following special
case.
Theorem. If Rn is bounded, u C 2 () C 1 () is harmonic, and , then
u(x)
K(x )
K(x )
dS.
(8.16)
u(x)
u() =
n
n
ds.
u
0 =
n
n
(8.17)
(8.18)
Suppose that for each we can nd a function (x) that is harmonic in and
satises (x) = K(x ) for all x . Then G(x, ) = K(x ) + (x) is a
fundamental solution such that
x .
G(x, ) = 0
9
K(x y)
n
2n
(2n)n
K()
=
r
1
2
1
n n1
if n = 2
if n 3.
if n = 2
if n 3,
1
,
n n1
43
G is called the Greens function and is useful in satisfying Dirichlet boundary conditions.
The Greens function is dicult to construct for a general domain since it requires
solving the Dirichlet problem = 0 in , (x) = K(x ) for x , for each
.
From (8.18) we nd 10
u() =
u(x)
G(x, )
dS.
n
Thus if we know that the Dirichlet problem has a solution u C 2 (), then we can
calculate u from the Poisson integral formula (provided of course that we can compute
G(x, )).
10
If we did not assume u = 0 in our derivation, we would have (8.13) instead of (8.16), and an
extra term in (8.17), which would give us a more general expression:
u() =
G(x, ) u dx +
u(x)
G(x, )
dS.
n
8.11
44
8.12
in
on ,
(8.19)
where is a bounded domain and is a (complex) number. The values of for which
(8.19) admits a nontrivial solution u are called the eigenvalues of in , and the
solution u is an eigenfunction associated to the eigenvalue . (The convention
u + u = 0 is chosen so that all eigenvalues will be positive.)
Properties of the Eigenvalues and Eigenfunctions for (8.19):
1. The eigenvalues of (8.19) form a countable set {n}
n=1 of positive numbers with
n as n .
2. For each eigenvalue n there is a nite number (called the multiplicity of n ) of
linearly independent eigenfunctions un .
3. The rst (or principal) eigenvalue, 1 , is simple and u1 does not change sign in .
4. Eigenfunctions corresponding to distinct eigenvalues are orthogonal.
5. The eigenfunctions may be used to expand certain functions on in an innite
series.
45
Heat Equation
ut = ku
(9.1)
9.1
9.1.1
by
If u C0 (Rn ), dene its Fourier transform u
1
eix u(x) dx
for Rn .
u
() =
n
(2) 2 Rn
We can dierentiate u
:
1
u
() =
eix (ixj )u(x) dx = (ix
n
j ) u ().
j
(2) 2 Rn
Iterating this computation, we obtain
k
k
u
() = (ix
j ) u ().
j
(9.2)
1
u
ix u
e
(x) dx =
(eix )u(x) dx
() =
n
n
xj
xj
(2) 2 Rn
(2) 2 Rn xj
1
(ij )eix u(x) dx
=
n
(2) 2 Rn
u().
= (ij )
Iterating this computation, we obtain
ku
().
() = (ij )k u
xkj
(9.3)
Formulas (9.2) and (9.3) express the fact that Fourier transform interchanges dierentiation and multiplication by the coordinate function.
9.1.2
MultiIndex Notation
u
= x11 xnn u.
xn n
x1 1
e
D u(x) dx =
Dx (eix )u(x) dx
D u() =
n
n
(2) 2 Rn
(2) 2 Rn
1
(i)eix u(x) dx
=
n
n
2
(2)
R
().
= (i) u
(i) = (i1 )1 (in )n .
46
Also,
u(x) dx =

u()2 d.
u(x) v(x) dx =
or
u
()
v() d.
The properties (9.2) and (9.3) make it very natural to consider the fourier transform
on a subspace of L1 (Rn ) called the Schwartz class of functions, S, which consists of the
smooth functions whose derivatives of all orders decay faster than any polynomial, i.e.
S = {u C (Rn ) : for every k N and Nn , xk Du(x) is bounded on Rn }.
For u S, the Fourier transform u
exists since u decays rapidly at .
is bounded. (ii) If u S, then u
S.
Lemma. (i) If u L1 (Rn ), then u
Dene the inverse Fourier transform for u L1 (Rn ):
1
eix u(x) dx
for Rn ,
u () =
n
(2) 2 Rn
1
eix u
() d
for x Rn .
u(x) =
n
2
n
(2)
R
or
ix
Rn
1
u
() d =
(2)n
R2n
ei(xy) u(y) dy d = (
u) (x).
47
u(x
a)() =
v() =
=
1
1
eix v(x) dx =
ei(y+a) u(y) dy
2 R
2 R
1
().
u(x
a)() = eia u
Delta function:
1
1
eix (x) dx = ,
2
R
1
= eia .
(x
a)() = eia ()
2
(x)()
=
Convolution:
(f g)(x) =
(f
g)() =
=
=
since u(x) =
R
(x y) u(y) dy .
f (x y)g(y) dy,
1
1
ix
e
f (x y) g(y) dy dx =
eix f (x y) g(y) dy dx
n
n
(2) 2 Rn
(2) 2 Rn Rn
Rn
1
i(xy)
iy
f
(x
y)
dx
e
g(y)
dy
e
n
(2) 2 Rn Rn
n
1
iz
e
f (z) dz
eiy g(y) dy = (2) 2 f()
g().
n
(2) 2 Rn
Rn
Rn
n
g().
(f
g)() = (2) 2 f()
2
2
1
1
1
=
e 2 dx e 2 =
e 2 dy e 2 =
2e 2 = e 2 .
2 R
2 R
2
2
x2
e 2 () = e 2 .
Multiplication by x:
1
ixu()
=
2
d
u
().
eix ixu(x) dx =
d
R
d
().
xu(x)()
= i u
d
48
1
1 ix
1
ix
e
e
xu
xux (x) dx =
(i)eix x + eix u dx
xux (x)() =
2 R
2
2 R
=0
1
1
eix u dx
= i eix x u dx
2
2
R
R
d
d
() u
() = u
() u
().
= i xu(x)()
u
() = i i u
d
d
d
u
() u
().
xu
x (x)() =
d
1
() =
2
a + x2
H(a
x)() =
H(x)()
=
H(x)
H(x) () =
1() =
11
11
2
1
e 2a ,
(Gaussian)
a
1
b
f
,
a
a
1, x L
1 2 sin(L)
,
f
(x)() =
2
0, x > L,
2a
1
,
(a > 0)
2
2 a + 2
2 a
e
,
(a > 0)
2a
2 1
sin a,
1
1
() +
,
i
2
2 1
,
(sign)
i
2().
Results with marked with were taken from W. Strauss, where the denition of Fourier Transform
is dierent. An extra multiple of 12 was added to each of these results.
49
(9.4)
)(,
t)
=
eix ut (x, t) dx = u
(u
(, t)
n
t
n
t
2
(2)
R
n
(ij )2 u
(, t) = 2 u
(, t).
u(, t) =
j=1
u
(, t) = 2 u
(, t),
t
2
u
(, t) = g() e t,
2
=
u(x, t) =
g() e t
1
1
g
n
n
(2) 2
(2) 2 Rn
x2
n
1
1
2
2
4t
eix t d =
n g
n g e
2
2
t
(4 ) 2
(4 ) 2
Rn
2
2
x
xy
1
1
4t
e 4t g(y) dy.
=
n g e
n
n
(4t) 2
(4t) 2 R
=
=
=
Thus,
12
1 2 t
n g e
(2) 2
2
e t eix d
1
u(x, t) =
K(x, y, t) g(y) dy =
n
n
(4t) 2
R
Rn
xy2
4t
g(y) dy.
Uniqueness of solutions for the pure initial value problem fails: there are nontrivial
solutions of (9.4) with g = 0. 13 Thus, the pure initial value problem for the heat
equation is not wellposed, as it was for the wave equation. However, the nontrivial
solutions are unbounded as functions of x when t > 0 is xed; uniqueness can be
regained by adding a boundedness condition on the solution.
12
13
k=0
2
1
dk
x2k k e1/t .
(2k)!
dt
50
Nonhomogeneous Equation
Consider the pure initial value problem with homogeneous initial condition:
for t > 0, x Rn
ut = u + f (x, t)
u(x, 0) = 0
for x Rn .
(9.5)
9.1.5
Rn
y, t s) f (y, s) dy ds.
K(x
Combining two solutions above, we nd that the solution of the initial value problem
for t > 0, x Rn
ut = u + f (x, t)
(9.6)
u(x, 0) = g(x)
for x Rn .
is given by
u(x, t) =
9.1.6
Rn
y, t) g(y) dy +
K(x
t
0
Rn
y, t s) f (y, s) dy ds.
K(x
(9.7)
51
For operators of the form t L, the fundamental solution of the initial value problem,
K(x, t) as dened in (9.8), coincides with the free space fundamental solution, which
satises
t L K(x, t) = (x, t),
provided we extend K(x, t) by zero to t < 0. For the heat equation, consider
x2
1
4t
e
t>0
n/2
(4t)
K(x,
t) =
0
t 0.
(9.9)
(9.10)
Rn+1
x2
1
4t
e
t>
(x, t) =
(4t)n/2
K
0
t .
as distributions, so it suces to show that (t )K
as distribu K
Then K
tions. Now
t v dx dt =
K(x,
t) t v(x, t) dx dt
K
n
R
K(x, t) t v(x, t) dx dt
K(x, t) v(x, t) dx dt
=
Rn
Rn
t=
+
t K(x, t) v(x, t) dx dt
K(x, t) v(x, t) dx dt
K(x, t) v(x, t) dx
=
Rn
Rn
Rn
t=
K(x,
) v(x, ) dx.
t K(x, t) v(x, t) dx dt +
=
Rn
Rn
10
52
Schr
odinger Equation
1 e
(xy)2
4t
(4t) 2
G(0, x, y) = (x y).
By analogy with the heat equation, nd the fundamental solution H(t, x, y) of the
Schr
odinger equation
Ht = iHxx,
H(0, x, y) = (x y).
Show that your expression H(x) is indeed the fundamental solution for the
Schr
odinger equation. You may use the following special integral
ix2
e 4 dx = i4.
Proof. Remark: Consider the initial value problem for the Schr
odinger equation
x Rn , t > 0,
ut = iu
u(x, 0) = g(x)
x Rn .
If we formally replace t by it in the heat kernel, we obtain the Fundamental
Solution of the Schr
odinger Equation: 15
x2
1
4it
(x Rn , t = 0)
n e
(4it) 2
xy2
1
e 4it g(y) dy.
u(x, t) =
n
(4it) 2 Rn
H(x, t) =
u
(, t) = i 2 u
(, t),
t
2
u
(, t) = g() ei t ,
1 i 2 t
2
g e
=
u(x, t) =
g() ei t
2
1
2
1
ei t eix d
= g
2
2 R
2
1
g
eixi t d
= (need some work) =
=
2
R
x2
xy2
1
1
g e 4it =
e 4it g(y) dy.
=
4it
4it R
15
53
x2
1
4it
e
t>0
(4it)n/2
(x,
t) =
0
t 0.
(10.1)
(10.2)
t i (x, t) = (x, t).
To verify (10.2) as distributions, we must show that for any v C0 (Rn+1 ): 16
Rn+1
x2
1
4it
e
t>
n/2
(x, t) =
(4it)
0
t .
as distributions, so it suces to show that (t i)
as distribu
Then
tions. Now
(x, t) t i v(x, t) dx dt
t i v dx dt =
Rn
t) v(x, t) dx dt +
(x,
) v(x, ) dx.
t i (x,
=
Rn
Rn
11
54
Problem (F90, #7). Use the method of characteristics to nd the solution of the
rst order partial dierential equation
x2 ux + xyuy = u2
which passes through the curve u = 1, x = y 2 . Determine where this solution becomes
singular.
Proof. We have a condition u(x = y 2 ) = 1.
dx
dt
= x2
dy
dt
dz
dt
= xy
= z2
x=
1
t c1 (s)
x(0, s) =
1
1
= s2 x =
c1 (s)
t +
1
s2
dy
s2 y
c2 (s)
s
=
y=
y(s, 0) = c2 (s) = s y =
,
dt
1 ts2
1 ts2
1 ts2
1
1
1
z(0, s) =
=1 z=
.
z=
t c3 (s)
c3 (s)
1t
Thus,
x
=s
y
y=
x
y
1 t xy2
u(x, y) =
1
1
y2
x2
t=
1
x
y2
1
.
2
x
x
x2
.
x2 + x y 2
s2
,
1 ts2
= 1
x = t,
= x2 y
dy
= t2 y
dt
= z
z = s2 et .
t3
Thus, x = t and s = ye 3 = ye
u(x, y) = (ye
x3
3
t3
x3
3
)2 ex = y 2 e 3 x
y = se 3 ,
, and
3 x
y4
y4 +y2 y2
= 1.
55
< x < , t 0
u(x, 0) = sin x
< x <
and solve it by the method of characteristics. Discuss the properties of the solution; in
particular investigate the behavior of ux(, t) for t .
Proof. is parametrized by : (s, 0, sin s). We have
dx
= x x = set ,
dt
dy
= 1 y = t,
dt
1 sin s
dz
= 1z z = 1
.
dt
et
Thus, t = y, s = xey , and
sin(xey )
1
+
.
y
e
ey
It can be checked that the solution satises the PDE and the initial condition.
As t , u(x, t) 1. Also,
u(x, y) = 1
t > 0,
56
t0
u(x, 0) = f (x),
< x < .
x(s, 0) = s, x(s,
0) = z(0) = f (s).
and thus,
z(s, t) = x(t)
u(x, 0) = f (x)
= z 2
= 1
= 3z
dx
= f 2 (s)e6t
dt
y = t,
z = f (s)e3t.
1
1
x = f 2 (s)e6t + f 2 (s) + s,
6
6
x = 16 f 2 (s)e6y + 16 f 2 (s) + s,
f (s) = ez3y
z2
z2
s = x 6y + .
6e
6
2
z 2 3y
z
e .
z = f x 6y +
6e
6
u2 3y
u2
e .
u(x, y) = f x 6y +
6e
6
x=
57
1 z 2 6y 1 z 2
z2
z2
+ s,
e
+
+
s
=
6 e6y
6 e6y
6e6y
6
58
u3
3
=0
(11.1)
u3
3
= cu.
(11.2)
How large does the constant c > 0 has to be, so that a smooth solution (with no discontinuities) exists for all t > 0? Explain.
Proof. a) Characteristic form: ut + u2 ux = 0.
: (s, 0, h(s)).
dz
dy
dx
= z2,
= 1,
= 0.
dt
dt
dt
x = h(s)2 t + s, y = t, z = h(s).
u(x, y) = h(x u2 y)
(11.3)
The characteristic projection in the xtplane17 passing through the point (s, 0) is the
line
x = h(s)2 t + s
along which u has the constant value u = h(s).
The derivative of the initial data is discontinuous, and that leads to a
rarefactionlike behavior at t = 0. However, if the question meant to ask to
determine the rst time when a shock forms, we proceed as follows.
Two characteristics x = h(s1 )2 t + s1 and x = h(s2 )2 t + s2 intersect at a point (x, t)
with
s2 s1
.
t=
h(s2 )2 h(s1 )2
From (11.3), we have
ux = h (s)(1 2uuxt)
ux =
h (s)
1 + 2h(s)h (s)t
Hence for 2h(s)h (s) < 0, ux becomes innite at the positive time
t=
1
.
2h(s)h (s)
The smallest t for which this happens corresponds to the value s = s0 at which h(s)h (s)
has a minimum (i.e.h(s)h (s) has a maximum). At time T = 1/(2h(s0 )h (s0 )) the
17
=0
2a2 ex (1 ex ),
2a2 ex (1 ex ),
x<0
x>0
x = ln( 12 ) = ln(2), x < 0
x = ln(2),
x>0
t=
a2
2 ,
x<0
2
1
= 2
min{2h(s)h (s)}
a
59
z 2 = h(s)2 e2ct
x=s+
y = t,
cz
z = h(s)ect
: (s, 0, h(s)).
1
h(s)2 (1 e2ct ),
2c
( h(s) = uecy ).
Thus,
1
u(x, y) = h x u2 e2cy (1 e2cy ) ecy .
2c
1
cy
(1 uux e2cy (1 e2cy )),
ux = h (s)e
c
cy
h (s)ecy
h (s)e
=
.
ux =
1 + 1c h (s)ecy u (1 e2cy )
1 + 1c h (s)h(s)(1 e2cy )
Thus, c > 0 that would allow a smooth solution to exist for all t > 0 should satisfy
1
1 + h (s)h(s)(1 e2cy ) = 0.
c
We can perform further calculations taking into account the result from part (a):
min{2h(s)h (s)} =
a2
.
2
60
u3x
=0
3
(11.4)
3 3s
x>0
aes = (s),
(s) = a e3 ,
2
= Fp = p =
a2 e2s
a2 e2s
x(s, t) =
x<0
x>0
x<0
x>0
a2 e2s t + c4 (s)
a2 e2s t + c5 (s)
x=
a2 e2s t + s
a2 e2s t + s
y(s, t) = t + c1 (s) y = t
3 3s
a3 e3s a 3e
= 23 a3 e3s ,
x<0
3
= pFp + qFq = p + q =
a3 e3s
2 3 3s
3
3s
+ 3 = 3a e ,
x>0
a e
2 3 3s
2 3 3s
s
x<0
3 a e t + c6 (s),
3 a e t a(1 e ),
z
=
z(s, t) =
x>0
23 a3 e3s t + c7 (s),
23 a3 e3s t a(1 es ),
=
dp
dt
Fq = 1
= Fx Fz p = 0
p(s, t) = c2 (s)
p=
aes ,
aes ,
x<0
x>0
x<0
x>0
dq
dt
= Fy Fz q = 0
q(s, t) = c3 (s)
q=
3 e3s
3 ,
a3 e3s
,
3
61
x<0
x>0
Thus,
u(x, y) =
2 3 3s
s
3 a e y a(1 e ),
23 a3 e3s y a(1 es ),
where s is dened as
a2 e2s y + s,
x=
a2 e2s y + s,
x<0
x>0
x<0
x > 0.
u3x
= cu.
3
(11.5)
3 3s
x>0
aes = (s),
(s) = a e3 + ca(1 ex ),
x<0
x>0
x<0
x>0
dx
= Fp = p2
dt
dy
= Fq = 1
dt
dz
= pFp + qFq = p3 + q
dt
dp
= Fx Fz p = cp
dt
dq
= Fy Fz q = cq
dt
We can proceed solving the characteristic equations with initial conditions above.
62
63
for x < 0
for x > 0
where a > 0 is a constant. Determine the rst time when a shock forms.
Proof. a) is parameterized by : (s, 0, h(s)).
dz
dy
dx
= z3,
= 1,
= 0.
dt
dt
dt
x = h(s)3 t + s, y = t, z = h(s).
u(x, y) = h(x u3 y)
(11.6)
ux =
h (s)
1 + 3h(s)2 h (s)t
Hence for 3h(s)2 h (s) < 0, ux becomes innite at the positive time
t=
1
.
3h(s)2 h (s)
The smallest t for which this happens corresponds to the value s = s0 at which
h(s)2 h (s) has a minimum (i.e.h(s)2 h (s) has a maximum). At time T = 1/(3h(s0 )2 h (s0 ))
the solution u experiences a gradient catastrophe.
Therefore, need to nd a minimum of
3a2 (1 ex )2 aex
= 3a3 ex (1 ex )2 ,
x<0
f (x) = 3h(x)2 h (x) =
2
x 2 x
3 x
x 2
= 3a e (1 e ) , x > 0
3a (1 e ) ae
3a3 ex (1 ex )2 ex 2(1 ex )ex
x<0
= 3a3 ex (1 ex )(1 3ex ),
=0
f (x) =
3
x
x 2
x
x x
3 x
x
x
= 3a e (1 e )(1 + 3e ), x > 0
3a e (1 e ) + e 2(1 e )e
x = 0, x = ln 3, x < 0,
We check which ones give the minimum of f (x) :
The zeros of f (x) are
x = 0, x = ln 3, x > 0.
18
x<0
x>0
t=
1
1
1
=
= 3.
2
min{3h(s) h (s)}
min f (s)
3a
64
65
b) Now consider
ut + u3 ux + cu = 0
with the same initial data and a positive constant c. How large does c need to be in
order to prevent shock formation?
b) Characteristic form: ut + u3 ux = cu.
dx
dt
dy
dt
dz
dt
z 3 = h(s)3 e3ct
y = t,
cz
z = h(s)ect
x=s+
: (s, 0, h(s)).
1
h(s)3 (1 e3ct ),
3c
( h(s) = uecy ).
1
z(s, t) = h x h(s)3 (1 e3ct ) ect ,
3c
1
u(x, y) = h x u3 e3cy (1 e3cy ) ecy .
3c
=
1
ux = h (s) ecy 1 u2 ux e3cy (1 e3cy ) ,
c
h (s)ecy
h (s)ecy
=
.
ux =
1 + 1c h (s)u2 e2cy (1 e3cy )
1 + 1c h (s)h(s)2 (1 e3cy )
Thus, we need
1
1 + h (s)h(s)2 (1 e3cy ) = 0.
c
We can perform further calculations taking into account the result from part (a):
min{3h(s)2 h (s)} = 3a3 .
66
a(x)dy
u(x, t) = h x
0
< x < , t 0,
< x < .
b) Find a class of initial data such that this problem has a global solution for all t.
Compute the critical time for the existence of a smooth solution for initial data, f ,
which is not in the above class.
Proof. a) is parameterized by : (s, 0, f (s)).
dx
dt
dy
dt
dz
dt
Check:
xz
dx
= xf (s)
dt
x = sef (s)t ,
y = t,
z = f (s).
z = f xef (s)t ,
u(x, y) = f xeuy .
=
ux =
uy =
f (s)euy
1f (s)xyeuy
f (s)euy xu
1f (s)xyeuy
uy xuux =
ux f (s)xeuy ux y = f (s)euy
uy f (s)xeuy uy y = f (s)xeuy u
f (s)euy xu
f (s)euy
xu
= 0.
1 f (s)xyeuy
1 f (s)xyeuy
u(x, 0) = f (x).
1
f (s)xeutc
67
68
Problem (F96, #6). Find an implicit formula for the solution u of the initialvalue
problem
ut = (2x 1)tux + sin(x) t,
u(x, t = 0) = 0.
Evaluate u explicitly at the point (x = 0.5, t = 2).
Proof. Rewrite the equation as
uy + (1 2x)yux = sin(x) y.
is parameterized by : (s, 0, 0).
dx
dt
dy
dt
dz
dt
= (1 2x)y = (1 2x)t
= 1
x=
1
1
2
(2s 1)et + ,
2
2
1 2 1
s = (x )et +
,
2
2
y = t,
= sin(x) y = sin
(2s 1)et +
t
t.
2
t dt + z(s, 0),
2
2
0
t
2
(2s 1)et +
t dt.
sin
z(s, t) =
2
2
0 y
2
(2s 1)ey +
y dy
sin
u(x, y) =
2
2
0 y
2
2
(2x 1)ey ey +
y dy
sin
=
2
2
y
0 y
(2x 1) +
y dy =
sin
sin(x) y dy,
=
2
2
0
0
2
y
u(x, y) = y sin(x) .
2
Note: This solution does not satisfy the PDE.
z(s, t) =
sin
(2s 1)et +
< x < , t 0,
f (x) C .
u(x, 0) = f (x),
x = set ,
z = f (s)et .
dy
=1
dt
y = t,
1
ey .
69
uy + uux = 0
= z=s
x = st + s
= 1
y = t,
= 0
z = s.
u(x, y) =
s=
x
x
=
.
t+1
y+1
x
; solution is smooth for all positive time y.
y+1
= z = s
x = st + s s =
= 1
y = t,
= 0
z = s.
u(x, y) =
x
x
=
.
1t
1y
x
; solution blows up at time y = 1.
y1
70
71
u(x, 0) = f (x)
0 < x < +
u(0, t) = g(t)
0 < t < +.
u(x, 0) = f (x)
0 < x < +
u(0, y) = g(y)
0 < y < +.
(x + 1)2
x=
s+1
1,
(s + 1)t 1
y = t,
s+1
1,
(s + 1)t 1
z = ln(s + 1)t 1 t + c1 (s),
=
x=
we have
x xy y
x xy y
+ 1 y 1 y + f
,
u(x, y) = ln
xy + y + 1
xy + y + 1
x xy y
1
.
u(x, y) = ln
y+f
xy + y + 1
xy + y + 1
Since t = y, s =
(x + 1)2
x=
1
1,
t1
y = t + s,
1
1,
t1
z = lnt 1 t + c2 (s),
=
x=
z = lnt 1 t + g(s).
19
x
s = y x+1
, we have
x
x
x
u(x, y) = ln
1
+g y
.
x+1
x+1
x+1
x
, both solutions are equal if f (0) = g(0).
Note that on y = x+1
Since t =
x
x+1 ,
72
73
u=x+y
2
and
for x = 0, t 0.
u= t +y
u(x1 , x2 , 0) = x1 + x2 ,
u(0, x2, x3 ) = x23 + x2 .
For region I, we solve the following characteristic equations with is parameterized
20
by : (s1 , s2 , 0, s1 + s2 ).
dx1
dt
dx2
dt
dx3
dt
dz
dt
x2
sin x3 = sin t
x1 = t + s1 ,
x2 = s2 et ,
x3 = t,
z = cos t + s1 + s2 + 1.
u(x, y, t) = cos t + x t + ye
+ 1,
or
x t.
For region II, we solve the following characteristic equations with is parameterized
by : (0, s2, s3 , s2 + s23 ).
dx1
= 1 x1 = t,
dt
dx2
= x2 x2 = s2 et ,
dt
dx3
= 1 x3 = t + s3 ,
dt
dz
= sin x3 = sin(t + s3 ) z = cos(t + s3 ) + cos s3 + s2 + s23 .
dt
Since t = x1 , s3 = x3 x1 , s2 = x2 ex3 , we have
u(x1 , x2 , x3 ) = cos x3 + cos(x3 x1 ) + x2 ex3 + (x3 x1 )2 ,
x t.
u(x, y, t) = cos t + cos(t x) + yet + (t x)2 ,
or
74
c2 (s) = 1.
y = sx + x ln x
t = ln x,
y x ln x
.
s =
x
u(x, y) = t + s = ln x +
u(x, y) =
y x ln x
.
x
y
.
x
We have found that the characteristics in the xyplane are of the form
y = sx + x ln x,
where s is such that 0 s 1. Also, the characteristics originate from .
Thus, u is uniquely determined in the region between the graphs:
y = x ln x,
y = x + x ln x.
12
Problems: Shocks
1
u2 = 0,
t>0
ut +
2
x
with initial data
1
u(x, 0) = h(x) =
0
if x < 1,
if 1 < x < 1,
if x > 1.
(t) =
F (ur ) F (ul )
=
ur ul
1 2
2 ur
0 12
12 u2l
1
= .
=
ur ul
01
2
shock speed:
(t) =
F (ur ) F (ul )
=
ur ul
1 2
2 ur
1
12 u2l
0
1
= .
= 2
ur ul
1 0
2
(t) =
F (ur ) F (ul )
=
ur ul
1 2
2 ur
1
12 u2l
1
= 2 2 = 0.
ur ul
1 1
75
For t < 2,
1
u(x, t) =
0
if x < 12 t 1,
if 12 t 1 < x < 12 t + 1,
if x > 12 t + 1.
and for t > 2,
u(x, t) =
1
1
if x < 0,
if x > 0.
76
77
1
u2 = 0,
t>0
ut +
2
x
with initial data
1
u(x, 0) = h(x) =
0
if x < 1,
if 1 < x < 1,
if x > 1.
xs
ul ,
t ul ,
xs
xs
u(x, t) =
ul t ur ,
t ,
xs
ur ,
t ur .
1,
x+1
t ,
u(x, t) =
0,
x1
t ,
1,
x < t 1,
t 1 < x < 1,
1 < x < 1,
1 < x < t + 1,
x > t + 1.
i.e. 1 <
x+1
t
<0
i.e.
x1
t
<1
0<
78
79
1
u2 = 0,
t>0
ut +
2
x
with initial data
u(x, 0) = h(x) =
2
0
if 0 < x < 1,
if otherwise.
(t) =
F (ur ) F (ul )
=
ur ul
1 2
2 ur
12 u2l
02
= 1.
=
ur ul
02
0
if x < 0,
x
if 0 < x < 2t,
t
For 0 < t < 1,
u(x, t) =
2
if 2t < x < t + 1.
0
if x > t + 1.
Rarefaction catches up to shock at t = 1.
Shock: At (x, t) = (2, 1), ul = x/t, ur = 0. RankineHugoniot shock condition:
F (ur ) F (ul )
=
ur ul
x
dxs
=
,
dt
2t
x = c t,
(t) =
1 2
2 ur
0 12 ( xt )2
12 u2l
1x
,
=
=
ur ul
0 xt
2t
and since the jump occurs at (x, t) = (2, 1), x(1) = 2 = c. Therefore, x = 2 t.
if x < 0,
0
x
For t > 1,
u(x, t) =
if 0 < x < 2 t,
t
0
if x > 2 t.
80
81
1
u2 = 0,
t>0
ut +
2
x
with initial data
1+x
u(x, 0) = h(x) =
0
if x < 0,
if x > 0.
x+1
.
t+1
xt
.
1+t
F (ur ) F (ul )
=
ur ul
x+1
t+1 ,
xt
x+1
=
.
1+t
t+1
2
0 12 ( x+1
12 u2l
1x+1
t+1 )
,
=
=
x+1
ur ul
2 t+1
0 t+1
1 2
2 ur
1x+1
dxs
=
,
dt
2 t+1
x = c t + 1 1,
and since the jump occurs
at (x, t) = (0, 0), x(0) = 0 = c 1, or c = 1. Therefore,
the shock curve is x = t + 1 1.
x+1
if x < t + 1 1,
t+1
u(x, t) =
0
if x > t + 1 1.
82
83
1
u2 = 0,
t>0
ut +
2
x
with initial data
u0
u(x, 0) = h(x) = u0 (1 x)
if x < 0,
if 0 < x < 1,
if x 1,
where u0 > 0.
Proof. Characteristic form: ut + uux = 0.
The characteristic projection in xtplane passing through the point (s, 0) is the line
x = h(s)t + s.
For s > 1, the characteristics are x = s.
For 0 < s < 1, the characteristics are x = u0 (1 s)t + s.
For s < 0, the characteristics are x = u0 t + s.
The characteristics emanating from (s, 0), 0 < s < 1 on xtplane intersect at (1, u10 ).
Also, we can check that the characteristics do not intersect before t = u10 for this
problem:
1
1
= .
tc = min
h (s)
u0
To nd solution in a triangular domain between x = u0 t and x = 1, we note that
characteristics there are x = u0 (1 s)t + s. Solving for s we get
x u0 t u0 (1 x)
x u0 t
.
Thus, u(x, t) = h(s) = u0 (1 s) = u0 1
=
.
s=
1 u0 t
1 u0 t
1 u0 t
We can also nd a solution in the triangular domain as follows. Note, that the characteristics are the straight lines
dx
= u = const.
dt
Integrating the equation above, we obtain
x = ut + c
Since all characteristics in the triangular domain meet at (1, u10 ), we have c = 1
and
u0 (1 x)
u
.
or
u=
x = ut + 1
u0
1 u0 t
if
x < u0 t,
u0
1
u0 (1x)
u(x, t) =
For 0 < t < ,
if u0 t < x < 1,
1u0 t
u0
0
if
x > 1.
u
u0 ,
84
(t) =
1
,
For t >
u0
u(x, t) =
u0
0
if x <
if x >
85
u0 t+1
2 ,
u0 t+1
2 .
x x
2
t
t
and
ux = f
x 1
.
t
t
x
x 1
x x
2 +a f
f
=0
t
t
t
t
t
or, assuming f is not identically 0 to rule out the constant solution, that
x x
= .
a f
t
t
This shows the functions a and f to be inverses of each other.
13
86
13.1
x2
p2
q+
= 0.
2
2
(s) =
s2 + 1
.
2
2 +1
).
= Fp = p,
= Fq = 1
y(s, t) = t + c1 (s)
y = t,
= pFp + qFq = p2 q,
= Fx Fz p = x,
s2 + 1
.
2
found y and q in terms of s and t. Note that we have a coupled system:
= Fy Fz q = 0
q(s, t) = c2 (s)
q=
x = p,
p = x,
x(s, 0) = s,
x (s, 0) = p(s, 0) = 1,
p + p = 0,
p(s, 0) = 1,
p (s, 0) = x(s, 0) = s.
87
+ s,
2
2
2
2
2
0
sin t cos t
s2 sin t cos t t
+ s,
+ s cos2 t
=
2
2
0
sin t cos t
s2 sin t cos t
=
+ s cos2 t
s+s =
2
2
s2 sin t cos t
sin t cos t
+ s cos2 t
.
=
2
2
Plugging in x and y found earlier for s and t, we get
(x sin(y))2 sin(y) cos(y)
sin(y) cos(y) x sin(y)
+
cos2 (y)
2
cos(y)
cos2 (y)
2
2
(x + sin y) sin y cos y
sin y cos y x + sin y
+
cos2 y +
=
2
cos y
cos2 y
2
2
(x + sin y) sin y
sin y cos y
+ (x + sin y) cos y +
=
2
2 cos y
2
sin y cos y (x + sin y) sin y
+
.
= x cos y +
2
2 cos y
u(x, y) =
88
t 0, < x <
ut +
< x < ,
where h(x) is smooth function which vanishes for x large enough.
Proof. Rewrite the equation as
F (x, y, z, p, q) =
x2
p2
+q+
= 0.
2
2
(s) =
h (s)2 + s2
.
2
= Fp = p,
= Fq = 1
y(s, t) = t + c1 (s)
y = t,
= pFp + qFq = p2 + q,
= Fx Fz p = x,
h (s)2 + s2
.
2
found y and q in terms of s and t. Note that we have a coupled system:
= Fy Fz q = 0
q(s, t) = c2 (s)
q=
x = p,
p = x,
x(s, 0) = s,
89
h (s)2 + s2
= h (s)2 cos2 t 2sh (s) cos t sin t + s2 sin2 t
.
2
t
h (s)2 + s2
dt + z(s, 0)
h (s)2 cos2 t 2sh (s) cos t sin t + s2 sin2 t
z(s, t) =
2
0
t
h (s)2 + s2
dt + h(s).
h (s)2 cos2 t 2sh (s) cos t sin t + s2 sin2 t
=
2
0
We integrate the above expression similar to S 01#3 to get an expression for z(s, t).
Plugging in x and y found earlier for s and t, we get u(x, y).
90
u(x, y) +
u(x, y)
= 2 + y,
x
y
u(x, 0) = u0 (x) = x.
Proof. We have
u2x + u2y = 2 + y
u(x, 0) = u0 (x) = x.
Rewrite the equation as
F (x, y, z, p, q) = p2 + q 2 y 2 = 0.
is parameterized by : (s, 0, s, (s), (s)).
We need to complete to a strip. Find (s) and (s), the initial conditions for p(s, t)
and q(s, t), respectively:
(s) = 1.
Fp = 2p
Fq = 2q
dx
=2
dt
dy
= 2t + 2
dt
x = 2t + s,
y = t2 + 2t,
dz
= 2t2 + 4t + 4,
dt
2
2
2
z = t3 + 2t2 + 4t + s = t3 + 2t2 + 4t + x 2t = t3 + 2t2 + 2t + x,
3
3
3
=
dp
= Fx Fz p = 0 p = 1,
dt
dq
= Fy Fz q = 1 q = t + 1.
dt
We solve y = t2 + 2t, a quadratic equation in t, t2 + 2t y = 0, for t in terms of y to
get:
t = 1 1 + y.
2
91
2y
,
f (r)
1 h (s)
= 0,
1 h (s) 0 1 +
h (s)
h (s) = 0.
Thus, h (s) = 0 ensures that the problem is noncharacteristic.
To show that one can solve y = (f (s))2 (x s) for (x, y) in a suciently small
neighborhood of (x0 , 0) with s(x0 , 0) = x0 , let
G(x, y, s) = (f (s))2 (x s) y = 0,
G(x0 , 0, x0) = 0,
Gr (x0 , 0, x0) = (f (s))2 .
Hence, if f (s) = 0, s, then Gs (x0 , 0, x0) = 0 and we can use the implicit function
theorem in a neighborhood of (x0 , 0, x0) to get
G(x, y, h(x, y)) = 0
and solve the equation in terms of x and y.
92
x2
x2
=0
and
uy x,
> 0.
u x,
2
2
Leave your answer in parametric form.
Proof. Rewrite the equation as
F (x, y, z, p, q) = p2 + q 2 1 = 0.
2
(s)2 =
s2
1
.
+1
2s
= Fp = 2p =
s2 + 1
2
= Fq = 2q =
2
s +1
2st
x=
+ s,
s2 + 1
2t
s2
y=
+ ,
2
s2 + 1
= Fy Fz q = 0
z = 2t,
s
,
s2 + 1
1
.
q=
s2 + 1
p=
13.2
93
x1 (s1 ,s2 ,0) x2(s1 ,s2 ,0) x3(s1 ,s2 ,0) z(s1 ,s2 ,0) p1 (s1 ,s2 ,0)
We need to complete to a strip. Find 1 (s1 , s2 ), 2 (s1 , s2 ), and 3 (s1 , s2 ), the initial
conditions for p1 (s1 , s2 , t), p2 (s1 , s2 , t), and p3 (s1 , s2 , t), respectively:
1 + 22 + 23 = 1.
h
s1
h
s2
f1
f2
f3
+ 2
+ 3
,
s1
s1
s1
s2 = 2 .
f1
f2
f3
= 1
+ 2
+ 3
,
s2
s2
s2
s1 = 3 .
= 1
21
s2 ,
s1
This problem is very similar to an already handwritten solved problem F95 #2.
94
= Fp1 = 1
x1 = t,
= Fp2 = 2p2
= Fp3 = 2p3
dx2
= 2s2
dt
dx3
= 2s1
dt
x2 = 2s2 t + s1 ,
x3 = 2s1 t + s2 ,
dp1
= Fx1 p1 Fz = 0 p1 = s21 s22 + 1,
dt
dp2
= Fx2 p2 Fz = 0 p2 = s2 ,
dt
dp3
= Fx3 p3 Fz = 0 p3 = s1 .
dt
Thus, we have
t = x1
=
t
x
s = x 2s t
x = 2s t + s
2
2
1
1
2
2
x3 = 2s1 t + s2
s2 = x3 2s1 t
z = (s2 + s2 + 1)t + s s
z = (s2 + s2 + 1)t + s s
1 2
1 2
1
2
1
2
u(x1 , x2 , x3 ) =
2x1 x2
s2 = x314x
x 2x x
x 2x x
x2 2x1 x3 2
x3 2x1 x2 2
2
1 3
3
1 2
+
+
1
x
+
.
1
1 4x21
1 4x21
1 4x21
1 4x21
t = x1
s1 = x2 2x12x3
14x
95
Problem (S94, #1). Solve the following PDE for f (x, y, t):
ft + xfx + 3t2 fy = 0
f (x, y, 0) = x2 + y 2 .
Proof. Rewrite the equation as (x x1 , y x2 , t x3 , f u):
x1 ux1 + 3x23 ux2 + ux3 = 0,
0
, s21 + s22 , 1 (s1 , s2 ), 2 (s1 , s2 ), 3 (s1 , s2 )
x1 (s1 ,s2 ,0) x2(s1 ,s2 ,0) x3(s1 ,s2 ,0) z(s1 ,s2 ,0) p1 (s1 ,s2 ,0)
We need to complete to a strip. Find 1 (s1 , s2 ), 2 (s1 , s2 ), and 3 (s1 , s2 ), the initial
conditions for p1 (s1 , s2 , t), p2 (s1 , s2 , t), and p3 (s1 , s2 , t), respectively:
h
s1
h
s2
3 = s1 1 .
f1
f2
f3
= 1
+ 2
+ 3
,
s1
s1
s1
2s1 = 1 .
f1
f2
f3
= 1
+ 2
+ 3
,
s2
s2
s2
2s2 = 2 .
2s2 ,
2s21
x1 (s1 ,s2 ,0) x2(s1 ,s2 ,0) x3(s1 ,s2 ,0) z(s1 ,s2 ,0) p1(s1 ,s2 ,0) p2 (s1 ,s2 ,0) p3 (s1 ,s2 ,0)
= Fp1 = x1
= Fp2 = 3x23
= Fp3 = 1
x1 = s1 et ,
dx2
= 3t2
dt
x2 = t3 + s2 ,
x3 = t,
= Fx3 p3 Fz = 6x3 p2
z = s21 + s22 ,
p1 = 2s1 et ,
p2 = 2s2 ,
dp3
= 12ts2
dt
p3 = 6t2 s2 + 2s21 .
3
f (x, y, t) = x2 e2t + (y t3 )2 .
u(x1 , x2 , x3 ) = x21 e2x + (x2 x33 )2 .
The solution satises the PDE and initial condition.
96
= x1
= x1 + x3
= 1
= x33
x1 = s1 et ,
dx2
= s1 et + t
dt
x2 = s1 et +
t2
+ s2 s1 ,
2
x3 = t,
dz
= t3
dt
z=
Since t = x3 , s1 = x1 ex3 , s2 = x2 s1 et
t4
+ s1 s2 .
4
t2
2
+ s1 = x2 x1
x43
x2
+ x1 ex3 (x2 x1 3 + x1 ex3 ),
4
2
4
2
t
t
+ xet (y x + xet ).
f (x, y, t) =
4
2
The solution satises the PDE and initial condition.
u(x1 , x2 , x3 ) =
x23
2
+ x1 ex3 , we have
or
0
, s1 s2 , 1 (s1 , s2 ), 2 (s1 , s2 ), 3 (s1 , s2 )
x1 (s1 ,s2 ,0) x2(s1 ,s2 ,0) x3(s1 ,s2 ,0) z(s1 ,s2 ,0) p1 (s1 ,s2 ,0)
We need to complete to a strip. Find 1 (s1 , s2 ), 2 (s1 , s2 ), and 3 (s1 , s2 ), the initial
conditions for p1 (s1 , s2 , t), p2 (s1 , s2 , t), and p3 (s1 , s2 , t), respectively:
h
s1
h
s2
3 = s1 (1 + 2 ).
f1
f2
f3
= 1
+ 2
+ 3
,
s1
s1
s1
s2 = 1 .
f1
f2
f3
= 1
+ 2
+ 3
,
s2
s2
s2
s1 = 2 .
s1 , s21 s1 s2
x1 (s1 ,s2 ,0) x2(s1 ,s2 ,0) x3(s1 ,s2 ,0) z(s1 ,s2 ,0) p1(s1 ,s2 ,0) p2 (s1 ,s2 ,0)
97
= Fp1 = x1
x1 = s1 et ,
= Fp3 = 1
dx2
= s1 et + t
dt
= Fp2 = x1 + x3
x2 = s1 et +
x3 = t,
22
z=
t4
+ s1 s2 ,
4
p1 = 2s1 et s1 ,
p3 = t3 s1 t s21 s1 s2 .
x43
x2
3
3
+ x1 ex (x2 x1 3 + x1 ex ),
4
2
2
t
t4
+ xet (y x + xet ).
f (x, y, t) =
4
2
The solution satises the PDE and initial condition.
u(x1 , x2 , x3 ) =
p2 = s1 ,
With t = x3 , s1 = x1 ex , s2 = x2 s1 et t2 + s1 = x2 x1
22
t2
+ s2 s1 ,
2
x23
2
+ x1 ex , we have
or
Variable t in the derivatives of characteristics equations and t in the solution f (x, y, t) are dierent
entities.
98
for t > 0
x1 = t + s1 ,
x2 = t + s2 ,
= 1
x3 = t,
dz
= dt z = (s1 , s2 )et .
z
(x , x , x ) 1 0 0
1 2 3
= 0 1 0 = 1 = 0
J is invertible.
J det
(s1 , s2 , t)
1
= z
Since t = x3 , s1 = x1 x3 , s2 = x2 x3 , we have
u(x1 , x2 , x3 ) = (x1 x3 , x2 x3 )ex3 ,
u(x, y, t) = (x t, y t)et .
or
99
x1 = at + s1 ,
x2 = bt + s2 ,
x3 = t,
dz
c(x1 , x2 , x3 )z
= c(at + s1 , bt + s2 , t)z
dt
t
c(a + s1 , b + s2 , )d + c1 (s1 , s2 ),
ln z =
=
z(s1 , s2 , t) = c2 (s1 , s2 )e
t
0
t
(x , x , x ) 1 0 0
1 2 3
= 0 1 0 = 1 = 0
J det
(s1 , s2 , t)
a b 1
dz
= c(at + s1 , bt + s2 , t)dt
z
J is invertible.
t
0
x3
0
x3
c(x+a(t),y+b(t),)d
or
100
Problem (F89, #4). Consider the rst order partial dierential equation
ut + ( + t)ux + etuy = 0
(13.1)
(13.2)
= ex3
dx1
= + t
dt
= + x3
dx2
= et
dt
= 1
x3 = t,
= 0
z = sin(s1 s2 ).
(x , x , x )
1 2 3
=
J det
(s1 , s2 , t)
Since t = x3 , s1 = x1
x23
2
x1 =
t2
+ t + s1 ,
2
x2 = et + s2 ,
1
0 0
0
1 0 = 1 = 0
t + et 1
J is invertible.
x3 , s2 = x2 ex3 + , we have
x23
x3 )(x2 ex3 + )),
2
t2
t)(y et + )).
u(x, y, t) = sin((x
2
The solution satises the PDE and initial condition.
u(x1 , x2 , x3 ) = sin((x1
or
b) We need a compatibility condition between the initial and boundary values to hold
on yaxis (x = 0, t = 0):
u(x = 0, y, 0) = u(0, y, t = 0),
0 = 0.
101
14
102
u1 (x, t)
u2 (x, t)
, (x, t) R R,
u1 (x, 0)
u2 (x, 0)
=
eixa
0
, a R.
Let U = V . Then,
Ut + AUx = 0,
Vt + AVx = 0,
Vt + 1 AVx = 0,
Vt + Vx = 0.
Thus, the transformed problem is
1 0
Vx = 0,
Vt +
0 3
1
V (x, 0) = U (x, 0) =
1
2
5
2
0
1
eixa
0
1
= eixa
2
1
5
.
U=
u(1)(x, t)
u(2)(x, t)
=
eia(x+t)
52 eia(x+t) + 52 eia(x+3t)
1 ia(x+t)
2e
5 ia(x+3t)
2e
.
.
Can check that this is the correct solution by plugging it into the original equation.
103
104
(14.2)
(14.3)
105
For region I, (14.2) and (14.3) give two initial value problems (since any point in
region I can be traced back along both characteristics to initial conditions):
(1)
(1)
(2)
(2)
vt + vx = 0,
vt 2vx = 0,
v (1)(x, 0) = sin x;
v (2)(x, 0) = 0.
which we solve by characteristics to get traveling wave solutions:
v (1)(x, t) = sin(x + 2t),
v (2)(x, t) = 0.
1 1
0 1
sin(x + 2t)
0
=
sin(x + 2t)
0
.
For region II, solutions of the form F (x + 2t) can be traced back to initial conditions.
Thus, v (1) is the same as in region I. Solutions of the form G(x t) can be traced back
to the boundary. Since from (14.4), u(2) = v (2), we use boundary conditions to get
u(2)(0, t) = t = G(t).
Hence, G(x t) = (x t).
Thus, for region II,
U = V =
1 1
0 1
sin(x + 2t)
(x t)
=
sin(x + 2t) (x t)
(x t)
.
106
u
1 2
u
=
.
v
2
2
t
x v
(14.5)
Find an explicit solution for the following mixed problem for the system (14.5):
u(x, 0)
f (x)
=
for x > 0,
v(x, 0)
0
u(0, t) = 0
for t > 0.
You may assume that the function f is smooth and vanishes on a neighborhood of x = 0.
Proof. Rewrite the equation as
1 2
Ux = 0,
Ut +
2 2
(1)
f (x)
u (x, 0)
=
.
U (x, 0) =
0
u(2)(x, 0)
The eigenvalues ofthe matrix A
are
1
2
, e2 =
vectors are e1 =
2
1
3 0
1
=
,
=
0 2
2
,
1
1
=
=
det
5
1 2
2 1
.
Let U = V . Then,
Ut + AUx = 0,
Vt + AVx = 0,
Vt + 1 AVx = 0,
Vt + Vx = 0.
Thus, the transformed problem is
3 0
Vx = 0,
Vt +
0 2
1
V (x, 0) =
1
U (x, 0) =
5
(14.6)
1 2
2 1
f (x)
0
f (x)
=
5
1
2
.
(14.7)
(14.8)
107
For region I, (14.6) and (14.7) give two initial value problems (since value at any
point in region I can be traced back along both characteristics to initial conditions):
(1)
(1)
(2)
(2)
vt 3vx = 0,
vt + 2vx = 0,
v (1)(x, 0) = 15 f (x);
v (2)(x, 0) = 25 f (x).
which we solve by characteristics to get traveling wave solutions:
1
2
v (2)(x, t) = f (x 2t).
v (1)(x, t) = f (x + 3t),
5
5
1
1
f (x + 3t)
f (x + 3t) + 45 f (x 2t)
1 2
5
5
=
.
Thus, for region I, U = V =
2
2
25 f (x 2t)
2 1
5 f (x + 3t) 5 f (x 2t)
For region II, solutions of the form F (x + 3t) can be traced back to initial conditions.
Thus, v (1) is the same as in region I. Solutions of the form G(x 2t) can be traced back
to the boundary. Since from (14.9),
u(1) = v (1) 2v (2),
we have
1
u(1)(x, t) = F (x + 3t) 2G(x 2t) = f (x + 3t) 2G(x 2t).
5
The boundary condition gives
1
u(1)(0, t) = 0 = f (3t) 2G(2t),
5
1
2G(2t) = f (3t),
5
1
3
G(t) = f t ,
10
2
1
3
f (x 2t) .
G(x 2t) =
10
2
1
1 2
5 f (x + 3t)
=
Thus, for region II, U = V =
1
3
2 1
10 f ( 2 (x 2t))
Solutions for regions I and II satisfy (14.5).
Solution for region I satises both initial conditions.
Solution for region II satises given boundary condition.
1
1
3
5 f (x + 3t) 5 f ( 2 (x 2t))
2
1
3
5 f (x + 3t) + 10 f ( 2 (x 2t))
.
108
Problem (F94, #1; S97, #7). Solve the initialboundary value problem
ut + 3vx = 0,
vt + ux + 2vx = 0
in the quarter plane 0 x, t < , with initial conditions
u(x, 0) = 1 (x),
v(x, 0) = 2 (x),
25
0 < x < +
t > 0.
u(1)(x, 0)
u(2)(x, 0)
=
1 (x)
2 (x)
(14.10)
.
V (x, 0) =
1
U (x, 0) =
4
1 1
1 3
(14.11)
1 (x)
2 (x)
1
=
4
1 (x) + 2 (x)
1 (x) + 32 (x)
. (14.12)
(14.13)
109
For region I, (14.11) and (14.12) give two initial value problems (since value at
any point in region I can be traced back along characteristics to initial conditions):
(1)
(2)
(1)
(2)
vt vx = 0,
vt + 3vx = 0,
v (1)(x, 0) = 14 1 (x) + 14 2 (x);
v (2)(x, 0) = 14 1 (x) + 34 2 (x),
which we solve by characteristics to get traveling wave solutions:
1
1
1
3
v (2)(x, t) = 1 (x 3t) + 2 (x 3t).
v (1)(x, t) = 1 (x + t) + 2 (x + t),
4
4
4
4
Thus, for region I,
1
3 1
4 1 (x + t) + 14 2 (x + t)
U = V =
1
3
1 1
4 1 (x 3t) + 4 2 (x 3t)
1 31 (x + t) 32 (x + t) + 1 (x 3t) + 32 (x 3t)
.
=
1 (x + t) + 2 (x + t) + 1 (x 3t) + 32 (x 3t)
4
For region II, solutions of the form F (x + t) can be traced back to initial conditions.
Thus, v (1) is the same as in region I. Solutions of the form G(x 3t) can be traced back
to the boundary. Since from (14.14),
we have
u(1) = 3v (1) + v (2),
3
3
u(1)(x, t) = 1 (x + t) 2 (x + t) + G(x 3t).
4
4
The boundary condition gives
3
3
u(1)(0, t) = (t) = 1 (t) 2 (t) + G(3t),
4
4
3
3
G(3t) = (t) 1 (t) + 2 (t),
4
4
t 3
t 3
t
1
+ 2
,
G(t) =
3
4
3
4
3
x 3t 3
x 3t 3
x 3t
1
+ 2
.
G(x 3t) =
3
4
3
4
3
Thus, for region II,
3 1
14 1 (x + t) + 14 2 (x + t)
U = V =
3
x3t
3
x3t
( x3t
1 1
3 ) 4 1 ( 3 ) + 4 2 ( 3 )
3
3
x3t
3
x3t
3
x3t
4 1 (x + t) 4 2 (x + t) + ( 3 ) 4 1 ( 3 ) + 4 2 ( 3 )
.
=
3
x3t
3
x3t
14 1 (x + t) + 14 2 (x + t) + ( x3t
3 ) 4 1 ( 3 ) + 4 2 ( 3 )
110
111
Problem (F91, #1). Solve explicitly the following initialboundary value problem for
linear 22 hyperbolic system
ut = ux + vx
vt = 3ux vx ,
where 0 < t < +, 0 < x < + with initial conditions
u(x, 0) = u0 (x),
v(x, 0) = v0 (x),
0 < x < +,
0 < t < +,
where b = 13 is a constant.
What happens when b = 13 ?
Proof. Let us change the notation (u u(1), v u(2)). Rewrite the equation as
1 1
Ux = 0,
(14.15)
Ut +
3 1
!
(1)
(1)
u0 (x)
u (x, 0)
.
=
U (x, 0) =
(2)
u(2)(x, 0)
u0 (x)
The eigenvalues ofthe matrix A
are
1
1
, e2 =
vectors are e1 =
1
3
2 0
1
=
,
=
0 2
1
,
1 =
1
4
3 1
1 1
.
Let U = V . Then,
Ut + AUx = 0,
Vt + AVx = 0,
Vt + 1 AVx = 0,
Vt + Vx = 0.
Thus, the transformed problem is
2 0
Vx = 0,
Vt +
0 2
V (x, 0) =
1
U (x, 0) =
4
3 1
1 1
(14.16)
u(1)(x, 0)
u(2)(x, 0)
1
=
4
(1)
(2)
(14.17)
Equation (14.16) gives traveling wave solutions of the form:
v (1)(x, t) = F (x + 2t),
(14.18)
112
(2)
(1)
(2)
2t)
4 0
4 u0 (x 2t)
!
3 (1)
1 (2)
1 (1)
1 (2)
u
(x
+
2t)
+
u
(x
+
2t)
+
u
(x
2t)
u
(x
2t)
4 0
4 0
4 0
4 0
.
=
3 (1)
1 (2)
3 (1)
3 (2)
u
(x
+
2t)
+
u
(x
+
2t)
u
(x
2t)
+
4 0
4 0
4 0
4 u0 (x 2t)
For region II, solutions of the form F (x + 2t) can be traced back to initial conditions.
Thus, v (1) is the same as in region I. Solutions of the form G(x 2t) can be traced back
to the boundary. The boundary condition gives
u(1)(0, t) + bu(2)(0, t) = (t).
Using (14.19),
v (1)(0, t) + G(2t) + bv (1)(0, t) 3bG(2t) = (t),
(1 + b)v (1)(0, t) + (1 3b)G(2t) = (t),
3
1 (2)
(1)
(1 + b) u0 (2t) + u0 (2t) + (1 3b)G(2t) = (t),
4
4
(1)
(2)
(t) (1 + b) 34 u0 (2t) + 14 u0 (2t)
,
G(2t) =
1 3b
(1)
(2)
( 2t ) (1 + b) 34 u0 (t) + 14 u0 (t)
,
G(t) =
1 3b
3 (1)
1 (2)
)
(1
+
b)
u
((x
2t))
+
u
((x
2t))
( x2t
2
4 0
4 0
.
G(x 2t) =
1 3b
Thus, for region II,
3 (1)
1 (2)
u
(x
+
2t)
+
u
(x
+
2t)
1 1
4 0
4 0
(1)
(2)
U = V =
( x2t
)(1+b) 34 u0 ((x2t))+ 41 u0 ((x2t))
1 3
2
13b
(1)
(2)
x2t
( 2 )(1+b) 34 u0 ((x2t))+ 14 u0 ((x2t))
3 (1)
1 (2)
u (x + 2t) + 4 u0 (x + 2t) +
(1)13b
.
= 4 0
(2)
3( x2t
)3(1+b) 34 u0 ((x2t))+ 14 u0 ((x2t))
1 (2)
3 (1)
2
4 u0 (x + 2t) + 4 u0 (x + 2t)
13b
The following were performed, but are arithmetically complicated:
Solutions for regions I and II satisfy (14.15).
113
114
u(x, 0) = 0, x 0
v(x, 0) = x2 , x 0
v(0, t) = t2 , t 0.
b)
u(x, 0) = 0, x 0
v(x, 0) = x2 , x 0
u(0, t) = t,
c)
t 0.
u(x, 0) = 0, x 0
v(x, 0) = x2 , x 0
u(0, t) = t,
t0
v(0, t) = t , t 0.
Proof. Rewrite the equation as Ut + AUx = BU . Initial conditions are same for
(a),(b),(c):
3 2
0 0
Ux =
U,
Ut +
0
1
0 1
(1)
0
u (x, 0)
=
.
U (x, 0) =
x2
u(2)(x, 0)
The eigenvalues ofthe matrix
A are
1
1
, e2 =
vectors are e1 =
0
2
3 0
1
=
,
=
0 1
0
,
1
=
2
2 1
0 1
Let U = V . Then,
Ut + AUx = BU,
Vt + AVx = BV,
Vt + 1 AVx = 1 BV,
Vt + Vx = 1 BV.
Thus, the transformed problem is
3 0
0 1
Vx =
V,
Vt +
0 1
0 1
1
V (x, 0) =
1
U (x, 0) =
2
2 1
0 1
(14.20)
0
x2
x2
=
2
1
1
.
(14.21)
(14.22)
115
Thus,
a) u(2)(0, t) = t2 , t 0. Wellposed.
b) u(1)(0, t) = t, t 0. Not wellposed.
c) u(1)(0, t) = t, u(2) (0, t) = t2 , t 0. Not wellposed.
116
=
2,
1
2
2 and the corresponding
1
1
, e2 =
. Thus,
eigenvectors are e1 =
1 + 2
1 2
1
1
1
1
+
2
0
2
1
1
,
=
,
=
.
=
1 + 2 1 2
0 2
2 2 1 + 2 1
Let U = V . Then,
Ut + AUx = 0,
Vt + AVx = 0,
Vt + 1 AVx = 0,
Vt + Vx = 0.
Thus, the transformed problem is
2
0
Vx = 0,
Vt +
0 2
1
V (x, 0) =
1
U (x, 0) =
2 2
(14.24)
1
1 + 2
(1 + 2)f (x) + g(x)
f (x)
1
=
.
g(x)
1 + 2 1
2 2 (1 + 2)f (x) g(x)
(14.25)
(14.26)
However, we can continue and obtain the solutions. We have two initial value problems
(2) (2)
(1) (1)
vt 2vx = 0,
vt + 2vx = 0,
v (1)(x, 0) =
(1+ 2)
1
f (x) +
g(x);
2 2
2 2
v (2)(x, 0) =
(1+ 2)
f (x)
2 2
(1 + 2)
1
(1)
(1
+
1
2)
g(x),
2 2
117
v (1)
v (2)
=
1
1
1 + 2 1 2
2 2
(1 + 2)f (x 2t) + g(x 2t)
.
(1 + 2)f (x + 2t) g(x + 2t)
(14.27)
+
2t) can be traced back to
In region III, the solutions of the form v (2) = G(x
(1)
right boundary and those of the form v = F (x 2t), to initial condition. Since by
(14.27), u(1) and u(2) are written in terms of both v (1) and v (2), one initial condition
and one boundary condition at x = 1 need to be prescribed.
To obtain the solution for region IV, two boundary conditions, one for each boundary, should be given.
Thus,
a) No boundary conditions. Not wellposed.
b) u(1)(0, t) = h(x), u(2)(0, t) = k(x). Not wellposed.
c) u(1)(0, t) = h(x), u(2)(1, t) = k(x). Wellposed.
118
f, h
prescribed on x = 0, t 0.
f g, h prescribed on x = 0, t 0.
f , g, h prescribed on x = 0, t 0.
For each of these 3 sets of data, determine whether or not the system is wellposed.
Justify your conclusions.
Proof. The third equation is decoupled from the rst two and can be considered separately. Its solution can be written in the form
h(x, t) = H(x 2t),
and therefore, h must be prescribed on t = 0 and on x = 0, since the characteristics
propagate from both the x and t axis.
We rewrite the rst two equations as (f u1 , g u2 ):
0 1
Ux = 0,
Ut +
1 0
(1)
u (x, 0)
.
U (x, 0) =
u(2)(x, 0)
The eigenvalues ofthe matrix
A are 1 = 1, 2 = 1 and the corresponding eigen
1
1
, e2 =
. Thus,
vectors are e1 =
1
1
1 1 1
1 0
1 1
1
.
=
,
=
,
=
1 1
0 1
1 1
2
Let U = V . Then,
Ut + AUx = 0,
Vt + AVx = 0,
Vt + 1 AVx = 0,
Vt + Vx = 0.
Thus, the transformed problem is
1 0
Vx = 0,
Vt +
0 1
1
V (x, 0) =
1
U (x, 0) =
2
(14.28)
1 1
1 1
u(1) (x, 0)
u(1) (x, 0)
.
(14.29)
(14.30)
119
u(1) prescribed on x = 0.
120
for x = 0
v = v0
for t = 0.
u(1)(x, 0)
u(2)(x, 0)
u(1)(0, t)
u(2)(0, t)
(1)
=
u0 (x)
(2)
u0 (x)
!
,
= 0.
0
1 ab
.
=
0
1 + ab
ab, 2 = 1 +
ab.
1 ab
0
Vx = 0,
Vt +
0
1 + ab
(14.33)
(14.34)
We also have U = V , i.e. both u(1) and u(2) (and their initial and boundary conditions)
are combinations of v (1) and v (2).
In order for this problem to be wellposed, both sets of characteristics should emanate
from the boundary at x = 0. Thus, the eigenvalues of the system are real (ab > 0) and
1,2 > 0 (ab < 1). Thus,
0 < ab < 1.
(1)
(2)
121
v(x, 0),
u(1, t),
v(1, t).
ut (1 cx2 )u(2)
x + F (x t) = 0.
Solving the equation by characteristics, we obtain
the characteristics in the xtplane are of the form
dx
= cx2 1.
dt
We need to determine c such that the prescribed
data u(2)(x, 0) and u(2)(1, t) makes the problem to
be wellposed. The boundary condition for u(2)(1, t)
requires the characteristics to propagate to the
left with t increasing. Thus, x(t) is a decreasing
function, i.e.
dx
<0
c < 1.
We could also do similar analysis we have done in other problems on rst order systems involving nding eigenvalues/eigenvectors of the system and using the fact that
u(1)(x, t) is known at both boundaries (i.e. values of u(1)(1, t) can be traced back either
to initial conditions or to boundary conditions on x = 1).
122
(14.36)
= 1 ,
dx
dt
(14.37)
= 2 .
1 < 0, 2 >
123
1 B.C. on x = 0, 1 B.C. on x = 1.
1 B.C. on x = 0.
1 B.C. on x = 1.
0
1 B.C. on x = 0, 1 B.C. on
124
1
0 9
inun (t)
+
0
invn (t)
1 0
0
9in
un (t)
+
vn (t)
in 0
2
n
in
n2 un (t)
=
n2 vn (t)
n2 0
un (t)
=
0 n2
vn (t)
un (t)
9in
= 0,
n2
vn (t)
0
1
(n2 )2 9n2 = 0,
which gives 1 =
n2 +3n,
2 =
n2 3n,
3i
1
un (t)
vn (t)
un (t)
vn (t)
,
,
3i
, v2 =
,
1
u
v
u
v
u
v
9vx uxx
ux vxx
125
. We have
+L
= 0, L
=
b) We want to solve
t
u
u
u
u
= L
=
, i.e.
= et . We can write the solution as
v
v
v
v t
&
un (t)einx
&
=
an e1 t v1 einx + bn e2 t v2 einx
U (x, t) =
vn (t)einx
n=
3i
3i
(n2 +3n)t
inx
(n2 3n)t
=
an e
e + bn e
einx .
1
1
n=
ix
e
3i
3i
inx
inx
an
,
e + bn
e =
U (x, 0) =
0
1
1
n=
an = bn = 0, n = 1;
1
1
and a1 = b1
a1 = b1 = .
a1 + b1 =
3i
6i
1
1 4t 3i
3i
e
eix + e2t
eix
U (x, t) =
1
1
6i
6i
1 4t
(e
+ e2t)
2
eix .
=
1 4t
2t )
(e
e
6i
26 27
26
126
15
127
15.1
Perturbation
28 29
ut + uux + (F ())x = 0,
t + (u)x = 0.
Here F () is a given C smooth function of . At t = 0, 2periodic initial data
u(x, 0) = f (x),
(x, 0) = g(x).
a) Assume that
f (x) = U0 + f1 (x),
g(x) = R0 + g1 (x)
where U0 , R0 > 0 are constants and f1 (x), g1 (x) are small perturbations. Linearize the equations and given conditions for F such that the linearized problem is
wellposed.
b) Assume that U0 > 0 and consider the above linearized equations for 0 x 1,
t 0. Construct boundary conditions such that the initialboundary value problem is
wellposed.
Proof. a) We write the equations in characteristic form:
ut + uux + F ()x = 0,
t + ux + ux = 0.
Consider the special case of nearly constant initial data
u(x, 0) = u0 + u1 (x, 0),
(x, 0) = 0 + 1 (x, 0).
Then we can approximate nonlinear equations by linear equations. Assuming
u(x, t) = u0 + u1 (x, t),
(x, t) = 0 + 1 (x, t)
remain valid with u1 = O(1), 1 = O(1), we nd that
ut = u1t,
t = 1t ,
ux = u1x ,
x = 1x,
u1t + (u0 + u1 )u1x + F (0 ) + 1 F (0 ) + O(2 ) 1x = 0,
= (u0 )(u0 ) 0 F (0) = 0,
15.2
Stationary Solutions
30
Consider
ut + uux + x = uxx ,
t + (u)x = 0
for t 0, < x < .
Give conditions for the states U+ , U , R+, R, such that the system has
stationary solutions (i.e. ut = t = 0) satisfying
u
U+
u
U
=
,
lim
=
.
lim
x+
x
R+
R
u2
x + uxx = 0,
ut =
2 x
t = (u)x = 0.
Integrating the above equations, we obtain
u2
+ ux = C1 ,
2
u = C2 .
30
128
129
15.3
130
Periodic Solutions
u(x, 0) = (x),
where u(x, t) and (x) are C functions periodic in x with period 2;
i.e. u(x + 2, t) = u(x, t), x, t.
Prove that
u(, t) Ceat 
'
'
2
2
2
2
where u(, t) =
0 u(x, t) dx,  =
0 (x) dx, C, a are some constants.
Proof. METHOD I:
u(x, t) =
an (t)einx .
n=
an (t)einx =
n=
an (t)
n4 an (t)einx + 2
n=
4
n2 an (t)einx ,
n=
2
= (n + 2n )an (t),
4 +2n2 )t
an (t) = an (0)e(n
u(x, 0) =
an (0)einx = (x),
n=
inx
an (0)e = (x).
n=
u(x, t)22
u (x, t) dx =
0
2
0
a2n (t)
n=
an (t)e
n=
einx einx dx = 2
inx
imx
an (t)e
dx
m=
a2n (t) = 2
n=
4 +2n2 )t
a2n (0)e2(n
n=
2(n4 +2n2 )t
2 2t
2
2
2
2
an (0)
e
an (0) e
e2(n 1) t
= 2
n=
= C2 e2t2.
n=
n=
2
n=
= C1 , (convergent)
131
=0
2
(2ab a2 + b2 )
2
2
2
u2xx dx +
(u2 + u2xx ) dx =
u2 dx = u2,
0
u2xx dx
=0
2
2uuxx dx
d
u2 2u2,
dt
u2 u(0)2e2t,
u u(0)et.
METHOD III: Can use Fourier transform. See ChiuYens solutions, that have both
Method II and III.
132
the L2 norm of f .
a) Express dpf /dxp2 in terms of the Fourier coecients of f .
b) Let q > p > 0 be integers. Prove that > 0, K = N (, p, q), constant, such that
dq f 2
dpf 2
p q + Kf 2.
dx
dx
c) Discuss how K depends on .
Proof. a) Let
31
f (x) =
fn einx ,
d f
dxp
fn (in)peinx ,
2
2
dpf 2
2
p inx 2
fn (in) e
dx =
i2 p
fn np einx dx
p =
dx
0
0
2
2
fn np einx dx = 2
fn2 n2p .
=
0
n=0
b) We have
dq f 2
dpf 2
p q + Kf 2,
dx
dx
fn2 n2p 2
fn2 n2q + K 2
fn2 ,
2
n=0
2p
n=0
2q
K,
(1 n )
K,
q
2p
n=0
some q > 0.
< 0, f or n large
Thus, the above inequality is true for n large enough. The statement follows.
31
Note:
L
e
inx imx
dx
0
L
n = m
n=m
32
133
f (x) = f (x + 2) C .
0 < 1,
g = 1,
Linearize the equation. Is the Cauchy problem wellposed for the linearized equation,
i.e., do its solutions v satisfy an estimate
v(, t) Ke(tt0 ) v(, t0)?
c) Determine the best possible constants K, .
Proof. a) The solution to
ut + uux + uxx + uxxxx = 0,
u(x, 0) = f0 = const,
is u(x, t) = f0 = const.
b) We consider the special case of nearly constant initial data
u(x, 0) = 1 + u1 (x, 0).
Then we can approximate the nonlinear equation by a linear equation. Assuming
u(x, t) = 1 + u1 (x, t),
remain valid with u1 = O(1), from , we nd that
u1t + (1 + u1 )u1x + u1xx + u1xxxx = 0.
Dividing by gives
u1t + u1x + u1 u1x + u1xx + u1xxxx = 0.
For small , we have
u1t + u1x + u1xx + u1xxxx = 0.
Multiply this equation by u1 and integrate
u1 u1t + u1 u1x + u1 u1xx + u1 u1xxxx = 0,
d u21
u21
+
+ u1 u1xx + u1 u1xxxx = 0,
dt 2
2 x
2
2
2
u21 2
1 d
2
u1 dx + +
u1 u1xx dx +
u1 u1xxxx dx = 0,
2 dt 0
0
0
20
=0
2 2
2
2 2
1 d
2
2
u1 2 + u1 u1x
u1x dx + u1 u1xxx u1x u1xx +
u21xx dx = 0,
2 dt
0
0
0
0
0
=0
=0
1 d
u1 22 =
2 dt
32
=0
2
u21x dx
u21xx dx.
134
u1 =
n=
u1x = i
u1xx =
an (t)einx .
Then,
u21x
inx
nan (t)e
n=
n=
n2 an (t)einx
u21xx =
n=
Thus,
n2 an (t)einx
2
.
n=
1 d
u1 22 =
2 dt
2
=
nan (t)einx ,
u21x dx
u21xx dx
inx
2
2
dx
n2 an (t)einx dx
0
0
2
2
4
n an (t)2 = 2
an (t)2 (n2 + n4 ) 0.
= 2
n an (t) 2
=
nan (t)e
where K = 1, = 0.
{
u(k, t)} is the Fourier series of u in x on [0, 2].
a) If u0 A, show that u(t) A.
b) Find dierential equations for u
(0, t), u
(1, t), and u
(2, t).
Proof. a) Solving
ut = ux + u4
u(x, 0) = u0 (x)
by the method of characteristics, we get
u(x, t) =
u0 (x + t)
1
(1 3t(u0 (x + t))3 ) 3
u
0k ei
kx
2
k=0
Since
u
k =
1
2
0
u(x, t) ei
kx
2
dx,
k=0
u
k ei
kx
2
135
15.4
136
Energy Estimates
f (x) = f (x + 2) C .
2
2 2
u2xx dx = 0,
+uuxxx ux uxx +
0
0
0
2
2
2
1
1 d
2
2
2
u
Ux u dx
ux dx +
u2xx dx = 0,
2 dt
2 0
0
0
2
2
2
1
1 d
u2 =
Uxu2 dx +
u2x dx
u2xx dx (from S90, #5)
2 dt
2 0
0
0
2
1
1 2
2
max Ux
Ux u dx
u2 dx.
2 0
2 x
0
d
u2 max Ux u2,
x
dt
u(x, t)2 u(x, 0)2 e(maxx Ux )t .
This can also been done using Fourier Transform. See ChiuYens solutions where the
above method and the Fourier Transform methods are used.
33
137
f (x) = f (x + 2) C .
u(x, 0) = f (x),
=0
2
2
1 d
2
2
u =
ux dx
u2xx dx 0,
2 dt
0
0
u(, t) u(, 0).
2 0
0
2
1 2 3
ux (uux )x =
ux .
2 0
0
We have
34
1 d
ux 2 +
2 dt
0
u3x
dx +
0
ux uxxx dx +
ux uxxxxx dx = 0.
2
2
2
1 d
w2 =
w 3 dx
wwxx dx
wwxxxx dx
2 dt
0
0
0
2
2
2
2
3
2
2
w dx +
wx dx
wxx dx
w 3 dx,
=
0
0
0
0
2
d
ux2 =
u3x dx.
dt
0
138
16
139
16.1
utt c uxx = 0,
u(x, 0) =
x3 , ut (x, 0) =
sin x .
g(x)
h(x)
(16.1)
x,
f (x,t)
u(x, 0) =
sin x,
ut (x, 0) = 1
+ x .
g(x)
h(x)
u (x, t) =
=
1
2c
1
2c
t
0
t
x+c(ts)
xc(ts)
1
f (, s) d ds =
2c
t
x+c(ts)
cos d ds
0
xc(ts)
1
sin[x + c(t s)] sin[x c(t s)] ds = 2 (cos x cos x cos ct).
c
1
(cos x cos x cos ct).
c2
140
We can check that the solution satises equation (16.1). Can also check that uA , uD
satisfy
2 A
2 D
uD
uA
tt c uxx = 0,
tt c uxx = cos x,
uA (x, 0) = sin x, uA
uD (x, 0) = 0, uD
t (x, 0) = 1 + x;
t (x, 0) = 0.
16.2
141
u(0, t) = 0,
u(L, t) = 0
t 0.
(16.2)
nx
nx
a0 (t)
+
+ bn (t) sin
.
an (t) cos
2
L
L
u(x, t) =
n=1
a0 (t)
+
an (t)
2
0 = u(0, t) =
an (t) = 0.
Thus,
n=1
u(x, t) =
bn (t) sin
n=1
nx
.
L
(16.3)
nx
nx
n 2
2
+c
= 0,
or
bn (t) sin
bn (t) sin
L
L
L
n=1
n=1
nc 2
bn (t) = 0,
bn (t) +
L
whose general solution is
bn (t) = cn sin
nct
nct
+ dn cos
.
L
L
(16.4)
nct
nc
nct
Also, bn (t) = cn ( nc
L ) cos L dn ( L ) sin L .
The constants cn and dn are determined by the initial conditions:
g(x) = u(x, 0) =
h(x) = ut (x, 0) =
n=1
nx
nx
=
,
bn (0) sin
dn sin
L
L
bn (0) sin
n=1
n=1
nx
nx
nc
=
sin
.
cn
L
L
L
n=1
mx
mx
nx
L
dx =
sin
dx = dm ,
g(x) sin
dn sin
L
L
L
2
0
0 n=1
L
L
nx
mx
mx
nc
mc L
dx =
sin
sin
dx = cm
.
h(x) sin
cn
L
L
L
L
L 2
0
0
n=1
Thus,
2
dn =
L
0
nx
dx,
g(x) sin
L
2
cn =
nc
h(x) sin
0
nx
dx.
L
(16.5)
utt uxx = 0
0<x<
u(x, 0) = 1,
ut (x, 0) = 0
u(0, t) = 0,
u(, t) = 0
t 0.
142
(16.6)
a0 (t)
+
an (t) cos nx + bn (t) sin nx.
u(x, t) =
2
n=1
0 = u(0, t) =
a0 (t)
+
an (t)
2
an (t) = 0.
Thus,
n=1
u(x, t) =
(16.7)
n=1
n=1
bn (t)
bn (t)n2 sin nx = 0,
or
n=1
+ n2 bn (t) = 0,
(16.8)
n=1
bn (0) sin nx =
bn (0) sin nx =
n=1
n=1
dn sin nx,
ncn sin nx.
n=1
sin mx dx = dm ,
2
0
0 dx = ncn .
2
0
Thus,
2
(1 cos n) =
dn =
n
4
n ,
0,
n odd,
n even,
and
cn = 0.
(16.9)
143
144
(2n + 1)
or
n=0
(2n + 1)
(2n + 1)
n=0
(16.10)
n=0
4 sin(2n + 1)x
(2n + 1)
(16.11)
n=0
u(x, t) =
1 1
= 0.
2 2
Since sin[(2n + 1)(x + t)] = sin[(2n + 1)(x + t 2)] = sin[(2n + 1)(2 (x + t))],
and 0 < 2 (x + t) < in R3 ,
in R2 ,
u(x, t) =
1 1
u(x, t) = + = 0.
2 2
Since 0 < (x t) < and 0 < 2 (x + t) < in R4 ,
in R3 ,
in R4 ,
1 1
u(x, t) = = 1.
2 2
145
2
0 < x < L, t > 0
utt c uxx = 0
0<x<L
u(x, 0) = g(x), ut (x, 0) = h(x)
ux (L, t) = 0
t 0.
ux (0, t) = 0,
(16.12)
a0 (t)
nx
nx
u(x, t) =
+
+ bn (t) sin
.
an (t) cos
2
L
L
n=1
n
nx
nx
ux (x, t) =
sin
+ bn (t)
cos
,
an (t)
L
L
L
L
n=1
0 = ux (0, t) =
bn (t)
n=1
bn (t) = 0.
Thus,
nx
a0 (t)
an (t) cos
+
.
u(x, t) =
2
L
(16.13)
n=1
n 2
nx
nx
a0 (t)
2
+
+c
= 0,
an (t) cos
an (t)
cos
2
L
L
L
n=1
n=1
nc 2
and
an (t) +
an (t) = 0,
a0 (t) = 0
L
whose general solutions are
nct
nct
+ dn cos
.
and
an (t) = cn sin
a0 (t) = c0 t + d0
L
L
(16.14)
nct
nc
nct
Also, a0 (t) = c0 and an (t) = cn ( nc
L ) cos L dn ( L ) sin L .
The constants cn and dn are determined by the initial conditions:
g(x) = u(x, 0) =
h(x) = ut (x, 0) =
n=1
d0
nx
nx
a0 (0)
+
=
+
,
an (0) cos
dn cos
2
L
2
L
a0 (0)
2
n=1
n=1
c0
nx
nx
nc
=
+
cos
.
cn
L
2
L
L
n=1
0
nx
dx,
g(x) cos
L
2
cn =
nc
0
nx
dx,
h(x) cos
L
2
c0 =
L
h(x) dx.
0
(16.15)
The formulas (16.13), (16.14), and (16.15) dene the solution.
146
t 0.
ux (0, t) = 0, ux (, t) = 0
Proof. Find u(x, t) in the form
a0 (t)
+
an (t) cos nx + bn (t) sin nx.
u(x, t) =
2
n=1
n=1
0 = ux (0, t) =
bn (t)n
bn (t) = 0.
Thus,
n=1
a0 (t)
u(x, t) =
+
an (t) cos nx.
2
(16.17)
n=1
n=1
n=1
an (t)
a0 (t)
+
an (t) cos nx +
an (t)n2 cos nx = 0,
2
a0 (t)
=0
and
+ n2 an (t) = 0,
and
(16.18)
n=1
n=1
d0
a0 (0)
+
+
an (0) cos nx =
dn cos nx,
2
2
a0 (0)
2
n=1
c0
+
2
cn n cos nx.
n=1
x dx = d0 ,
x cos mx dx = dm ,
2
2
0
0
0 dx = c0 ,
0 cos mx dx = cm m .
2
2
0
0
Thus,
d0 = ,
dn =
2
(cos n 1),
n2
cn = 0.
an (t) =
2
(cos n 1) cos nt,
n2
(16.19)
u(x, t) = +
.
2
n=1
n2
147
148
,
or
u(x, t) = +
2
n2
n=1
1
u(x, t) = +
2
n=1
n2
n=1
2 (cos n 1) cos nx
u(x, 0) = x = +
2
n2
n=1
which implies
1 (cos n 1) cos nx
x
=
2
4
n2
(16.21)
n=1
u(x, t) =
(x t) x + t
+
+
= t.
2
2
4
2
4
Since cos[n(x+t)] = cos[n(x+t2)] = cos[n(2 (x+t))], and 0 < 2 (x+t) <
in R3 ,
in R2 ,
u(x, t) =
x t 2 (x + t)
+
+
= t.
2
2
4
2
4
Since 0 < (x t) < and 0 < 2 (x + t) < in R4
in R3 ,
u(x, t) =
in R4 ,
u(x, t) =
(x t) 2 (x + t)
+
+
= x.
2
2
4
2
4
149
Example (McOwen 3.1 #4). Consider the initial boundary value problem
2
for
x > 0, t > 0
utt c uxx = 0
(16.22)
for x > 0
u(x, 0) = g(x), ut (x, 0) = h(x)
u(0, t) = 0
for t 0,
where g(0) = 0 = h(0). If we extend g and h as odd functions on < x < , show
that dAlemberts formula gives the solution.
Proof. Extend g and h as odd functions on < x < :
g(x),
x0
h(x),
g(x) =
h(x)
=
g(x), x < 0
h(x),
Then, we need to solve
xx = 0
for
u
tt c2 u
u
(x, 0) = g(x), u
t(x, 0) = h(x)
x0
x < 0.
(16.23)
To show that dAlemberts formula gives the solution to (16.23), we need to show that
the solution given by dAlemberts formula satises the boundary condition u(0, t) = 0.
1 x+ct
1
(
g(x + ct) + g(x ct)) +
h() d,
u
(x, t) =
2
2c xct
1 ct
1
(
g(ct) + g(ct)) +
h() d
u
(0, t) =
2
2c ct
1
1
(
g(ct) g(ct)) + (H(ct) H(ct))
=
2
2c
1
= 0 + (H(ct) H(ct)) = 0,
2c
x
h()
d; and since h is odd, then H is even.
where we used H(x) =
0
Example (McOwen 3.1 #5). Find in closed form (similar to dAlembets formula)
the solution u(x, t) of
for x, t > 0
utt c uxx = 0
(16.24)
u(x, 0) = g(x), ut (x, 0) = h(x) for x > 0
u(0, t) = (t)
for t 0,
where g, h, C 2 satisfy (0) = g(0), (0) = h(0), and (0) = c2 g (0). Verify that
u C 2 , even on the characteristic x = ct.
Proof. As in (McOwen 3.1 #4), we can extend g and h to be odd functions. We want
to transform the problem to have zero boundary conditions.
Consider the function:
U (x, t) = u(x, t) (t).
(16.25)
fU (x,t)
gU (x)
0 .
U (0, t) =
u (t)
150
151
2
for x > 0, t > 0
utt c uxx = 0
(16.26)
for x > 0
u(x, 0) = g(x), ut (x, 0) = h(x)
for t > 0.
ux (0, t) = 0
Proof. Extend g and h as even functions on < x < :
g(x),
x0
h(x),
x0
g(x) =
h(x)
=
g(x), x < 0
h(x), x < 0.
Then, we need to solve
xx = 0
for
u
tt c2 u
u
(x, 0) = g(x), u
t(x, 0) = h(x)
(16.27)
To show that dAlemberts formula gives the solution to (16.27), we need to show that
the solution given by dAlemberts formula satises the boundary condition ux (0, t) = 0.
1 x+ct
1
(
g(x + ct) + g(x ct)) +
h() d.
u
(x, t) =
2
2c xct
1
1
ct)],
(
g (x + ct) + g (x ct)) + [h(x
+ ct) h(x
u
x (x, t) =
2
2c
1
1
(
g (ct) + g (ct)) + [h(ct) h(ct)]
= 0.
u
x (0, t) =
2
2c
Since g is even, then g is odd.
Problem (F89, #3). 36 Let = c, constant. Find the solution of
for t > 0,
ut (0, t) = ux (0, t)
(16.28)
Similar to McOwen 3.1 #5. The notation in this problem is changed to be consistent with McOwen.
2
u(0, t) = 0,
u(, t) = /2 for t 0.
152
(16.29)
Proof. If we rst nd a particular solution of the nonhomogeneous equation, this reduces the problem to a boundary value problem for the homogeneous equation ( as in
(McOwen 3.1 #2) and (McOwen 3.1 #3) ).
Hint: You should use a particular solution depending on x!
First, nd a particular solution. This is similar to the method of separation of
variables. Assume
up (x, t) = X(x),
which gives
X (x) = 1,
X (x) = 1.
The solution to the above ODE is
x2
X(x) = + ax + b.
2
The boundary conditions give
up(0, t) = b = 0,
2
2
up (, t) = + a + b = ,
2
2
Thus, the particular solution is
up (x, t) =
a = b = 0.
x2
.
2
up tt up xx = 1
2
up (x, 0) = x2 , up t (x, 0) = 0
up(, t) = 2 .
up (0, t) = 0,
Second, we nd a solution to a boundary value problem for the homogeneous equation:
utt uxx = 0
2
ut (x, 0) = 0
u(x, 0) = x2 ,
u(0, t) = 0,
u(, t) = 0.
This is solved by the method of Separation of Variables. See Separation of Variables
subsection of Problems: Separation of Variables: Wave Equation McOwen 3.1 #2.
The only dierence there is that u(x, 0) = 1.
We would nd uh (x, t). Then,
u(x, t) = uh (x, t) + up (x, t).
153
utt uxx = 0,
0 .
u(x, 0) = a cos x b sin x G(x), ut (x, 0) =
g(x)
h(x)
154
It can be veried that the solution satises the above homogeneous PDE with the
boundary conditions. Thus, the complete solution is:
u(x, t) = uh (x, t) + up (x, t).
t
37
x+(ts)
f () cos s d ds.
0
x(ts)
x+t
h() d = 0.
xt
16.3
155
Similarity Solutions
(16.30)
(16.31)
x
v(x, t) = t w(y),
y= .
t
1
vt = t
w + t w yt = t
w+t
vx = t w yx = t w t = t w ,
x
+1 w = t1 w t1 yw ,
t
vxx = (t w )x = t w yx = t w t = t2 w .
Plugging in the derivatives we calculated into (16.31), we obtain
t1 w t1 yw = t2 w + 2(tw)(t w ),
w yw = t12 w + 2t+1 ww .
The parameters that would eliminate t from equation above are
1
1
= , = .
2
2
With these parameters, we obtain the dierential equation for w(y):
1
1
w yw = w + 2ww ,
2
2
1
1
w + 2ww + yw + w = 0.
2
2
We can write the ODE as
1
w + 2ww + (yw) = 0.
2
Integrating it with respect to y, we obtain the rst order ODE:
1
w + w 2 + yw = c.
2
16.4
156
38
(16.32)
(16.33)
(16.34)
(16.35)
(16.36)
We are looking for solutions f which satisfy f (x), f (x), f (x) 0 as x . (In
which case the function u having the form (16.33) is called a solitary wave.) Then
(16.35) and (16.36) imply a = b = 0, so that
(f )2
c
= 0,
f2 + f3 +
2
2
or f = f
c 2f .
c
c
2
f (x) = sech [ (x x0 )],
2
2
where x0 is the constant of integration. A solution of this form is called a soliton.
38
157
,
t
2
x x3
where n is a positive integer. Solitary wave solutions are sought in which u = f (),
where = x ct and
f, f , f 0,
as  ;
(16.37)
1
(f )2
cf 2
= f n+2
+ af + b.
(16.38)
2
2
2
We are looking for solutions f which satisfy f, f , f 0 as x . Then (16.37)
and (16.38) imply a = b = 0, so that
1 n+2 (f )2
cf 2
=
f
,
2
2
2
(f )2 = f n+2 + cf 2 .
f  f n + c d,
f =
f  f n + c d.
0 =
Thus, either f  0 f = 0, or
f n + c = 0. Since f 0 as x , we have c = 0 f = 0.
Thus, solitary waves do not exist if n is even.
158
159
When n = 1, we have
(f )2 = f 3 + cf 2 .
(16.39)
We show that all conditions of the problem are satised provided c > 0, including
1
c(x ct) ,
or
u = c sech2
2
c 2
c
2 c
f = c sech
= c cosh
=
.
2
2
cosh2 [ 2 c ]
We have
f
= 2c cosh
c 3
2
2
(f )
c sinh 2 c
,
cosh6 2 c
f3 =
cf 2 =
2
c c
c 3
c
= c c cosh
,
sinh
sinh
2
2
2
2
c3
cosh6
c3
cosh4
,
c
2
.
c
2
= c
=
0, as  .
f () = c sech
c
c
2
cosh2 [ 2 c ]
e[ 2 ] + e[ 2 ]
Similarly, f , f 0, as  .
39
cosh2 x sinh2 x = 1.
cosh x =
ex + ex
,
2
sinh x =
ex ex
2
160
Problem (S00, #5). Look for a traveling wave solution of the PDE
utt + (u2 )xx = uxxxx
of the form u(x, t) = v(x ct). In particular, you should nd an ODE for v. Under
the assumption that v goes to a constant as x , describe the form of the solution.
Proof. Since (u2 )x = 2uux, and (u2 )xx = 2u2x + 2uuxx, we have
utt + 2u2x + 2uuxx = uxxxx .
We look for a traveling wave solution
u(x, t) = v(x ct).
We get the ODE
c2 v + 2(v )2 + 2vv = v ,
c2 v + 2((v )2 + vv ) = v ,
c2 v + 2(vv ) = v ,
2
c v + 2vv = v + a,
c2 v + v 2 = v + as + b,
(exact dierentials)
s = x ct
v + c2 v + v 2 = a(x ct) + b.
Since v C = const as x , we have v , v 0, as x . Thus, implies
c2 v + v 2 = as + b.
Since x , but v C, we have a = 0:
v 2 + c2 v b = 0.
c2 c4 + 4b
.
v=
2
161
y = x st.
2
In order to nd and analyze the stationary points of an ODE above, we write it as a
rstorder system.
1 = ,
2 = .
1 = = 2 ,
1
s
s
1
2 = = + 2 = 2 1 + 21 .
2
1 = 2 = 0,
1 = 2 = 0,
1 = 2 = 0,
1 2
1 2
1 = 0;
1 = 0;
2 = 2 s 1 + 2
2 = s 1 + 2
2 = 1 1 (s 12 1 ) = 0.
Stationary points: (0, 0), (2s, 0), s > 0.
1 = 2
= f (1 , 2 ),
s
1
2 = 2 1 + 21 = g(1 , 2 ).
162
For (1 , 2 ) = (0, 0) :
1
det(J(0,0) I) = s
'
2
s
4
= 2
2 .
2
>s
If 4
(0,0)
2
If 4
<s
(0,0)
= 2 + +
= 0.
R, < 0.
is Stable Improper Node.
C, Re( ) < 0.
is Stable Spiral Point.
For (1 , 2 ) = (2s, 0) :
1
= 2 +
det(J(2s,0) I) = s
'
2
s
4
= 2
2 + .
+ > 0, < 0.
(2s,0) is Untable Saddle Point.
= 0.
b) Since
lim (y) = 0 = lim (x st),
we may have
lim (y) = lim (x st) = 2s.
y+
That is, a particle may start o at an unstable node (2s, 0) and as t increases, approach
the stable node (0, 0).
A phase diagram with (0, 0) being a stable spiral point, is shown below.
163
164
y = x st.
suL + f (uL )
+b
suR + f (uR )
+b
0 = =
0 = =
suL f (uL )
.
suR f (uR )
b=
.
b=
f (uL ) f (uR )
.
uL uR
40
40
For the solution for the second part of the problem, refer to ChiuYens solutions.
165
lim u(x, t) = 1.
x+
One class of solutions is the set of wavefront solutions. These have the form u(x, t) =
(x + ct), c 0.
Determine the ordinary dierential equation and boundary conditions which must
satisfy (to be of physical interest). Carry out a phase plane analysis of this equation,
and show that physically interesting wavefront solutions are possible if c 2, but not if
0 c < 2.
Proof. We look for a traveling wave solution
u(x, t) = (x + ct),
s = x + ct.
(s) 0, as s ,
(s) 1, as s +,
0 1.
2
y2 = cy2 y1 + y1 = 0;
y1 (y1 1) = 0.
Stationary points: (0, 0), (1, 0).
y1 = y2
y2
= cy2
= f (y1 , y2 ),
y1 + y12
= g(y1 , y2 ).
166
= 2 c + 1 = 0.
c2 4
.
2
If c 2 R, > 0.
(0,0) is Unstable Improper (c > 2) / Proper (c = 2) Node.
If 0 c < 2 C, Re( ) 0.
(0,0) is Unstable Spiral Node.
For (y1 , y2 ) = (1,0) :
1
det(J(1,0) I) =
1 c
=
= 2 c 1 = 0.
c2 + 4
.
2
If c 0 + > 0, < 0.
(1,0) is Unstable Saddle Point.
By looking at the phase plot, a particle may start o at an unstable node (0, 0) and as
t increases, approach the unstable node (1, 0).
167
168
y = x st.
1
1
u + a u = u ,
su + u u +
s
s
2
Integrating, we get
su + uu + au = u .
s
1
su + u2 + au = u + b.
s
u =
1 2
u + (a s)u b.
s
b) Note that the ODE above may be written in the following form:
u + Au2 + Bu = C,
which is a nonlinear rst order equation.
169
g2 f 2
,
1s
f2 g
.
1 s
2y1 2y2
y1
y2
.
=
J((y1 , y2 ), (y1, y2 )) =
2y1 1
y
y
1
= (1 ) = 0.
1 = 0, 2 = 1;
(0,0) is Unstable Node.
eigenvectors: v1 =
1
0
, v2 =
0
1
.
= 2 3 2 = 0.
17
.
2
< 0, + > 0.
(1,1) is Unstable Saddle Point.
For (y1 , y2 ) = (1,1) :
2
2
det(J(1,1) I) =
2
1
1
7
= i
.
2
2
Re( ) < 0.
(1,1) is Stable Spiral Point.
= 2 + + 2 = 0.
170
16.5
171
Dispersion
(16.40)
with constant .
a) Linearize this equation around u = 0 and nd the principal mode solution of the
form et+ikx . For which values of are there unstable modes, i.e., modes with = 0
for real k? For these values, nd the maximally unstable mode, i.e., the value of k with
the largest positive value of .
b) Consider the steady solution of the (fully nonlinear) problem. Show that the resulting
equation can be written as a second order autonomous ODE for v = ux and draw the
corresponding phase plane.
Proof. a) We have
ut = (f (ux))x uxxxx ,
ut = (u2x ux )x uxxxx ,
ut = 2ux uxx uxx uxxxx .
u2
f (0) + = 0 + u(0 1) + = u + .
2
Thus, we have
ut = uxx uxxxx .
Consider u(x, t) = et+ikx .
et+ikx = (k2 k4 )et+ikx ,
= k2 k4 .
To nd unstable nodes, we set = 0, to get
=
1
.
k2
To nd the maximally unstable mode, i.e., the value of k with the largest positive
value of , consider
(k) = k2 k4 ,
(k) = 2k 4k3 .
To nd the extremas of , we set = 0. Thus,the extremas are at
1
.
k1 = 0, k2,3 =
2
To nd if the extremas are maximums or minimums, we set = 0:
(k) = 2 12k2 = 0,
k = 0 is the minimum.
(0) = 2 > 0
1
1
= 4 < 0 k =
is the maximum unstable mode.
2
2
1
1
=
is the largest positive value of .
2
4
172
b) Integrating , we get
u2x ux uxxx = 0.
Let v = ux . Then,
v 2 v vxx = 0, or
v2 v
.
v =
y1 = y2 = 0,
y2 = 0,
y12 y1
y1 (y1 1) = 0.
y2 = = 0;
Stationary points: (0, 0), (1, 0).
y1 = y2
= f (y1 , y2 ),
2
y y1
= g(y1 , y2 ).
y2 = 1
y1
y2
'
For (y1 , y2 ) = (0, 0), = 1 .
+ > 0, < 0. (0,0) is Unstable Saddle Point.
If < 0, R, '
If > 0, = i
173
16.6
174
Energy Methods
Problem (S98, #9; S96, #5). Consider the following initialboundary value
problem for the multidimensional wave equation:
in (0, ),
utt = u
u
(x, 0) = g(x)
for x ,
u(x, 0) = f (x),
t
u
u
+ a(x)
=0
on .
n
t
Here, is a bounded domain in Rn and a(x) 0. Dene the Energy integral for this
problem and use it in order to prove the uniqueness of the classical solution of the problem.
Proof.
dE
=0 =
dt
u
ut ds +
n
(utt u)ut dx =
utt ut dx
1 2
1
2
(ut ) dx +
u dx +
a(x)u2t ds.
=
2
t
2
t
Thus,
a(x)u2t dx
1
=
2 t
u ut dx
ut (utt u) dx
a(x)u2t dx 0.
=
=0
in (0, ),
for x ,
w(x, 0) = 0,
wt(x, 0) = 0
w
w
+ a(x)
=0
on .
n
t
We have
1
(wt(x, 0)2 + w(x, 0)2) dx = 0.
Ew (0) =
2
175
x
on R.
Assume that (x) is of one sign for all x (i.e. always positive or always negative).
For the energy
1
1
E(t) =
u2 + u2 dx,
2
2 c (x) t
show that the sign of
Proof. We have
dE
dt
1
u
u
+
u
u
t tt
t dx
2
c (x)
1
u
ut utt dx +
ds
ut
ut u dx
=
c2 (x)
n
1 2
1
utt u dx +
ut dx
ut 2
=
c (x)
(x)
=0
> 0, if (x) > 0, x ,
1 2
ut dx =
=
< 0, if (x) < 0, x .
(x)
dE
(t) =
dt
176
Problem (F92, #2). Let Rn . Let u(x, t) be a smooth solution of the following
initial boundary value problem:
utt u + u3 = 0
u(x, t) = 0
=0
2
1 d
1
1 2
1
1 4
2
(ut ) dx +
u dx +
(u ) dx =
ut + u2 + u4 dx.
=
2 dt
2
2 t
2 t
4 t
Thus, the Energy integral is
E(t) =
1
u2t + u2 + u4 dx = const = E(0).
2
Thus, u 0.
177
Thus,
dE
dt
R3
R3
178
Problem (W03, #8). a) Consider the damped wave equation for highspeed waves
(0 < << 1) in a bounded region D
2 utt + ut = u
with the boundary condition u(x, t) = 0 on D. Show that the energy functional
2 u2t + u2 dx
E(t) =
D
=
2 ututt dx +
D
2u ut dx
u
2 ututt dx +
2 ut ds
2uut dx
=
D
D n
D
=0, (u=0 on D)
2
ut 2 dx 0.
= 2 ( utt u)ut dx = = 2
D
2 2
D (ut )
+
Since E
(0) E (t) =
Thus, D u 2 dx 0 as 0.
c) If = 1,
u 2 dx,
2(1) f (x)2 dx =
E (0) =
D
then E(t) 0 as 0.
f (x)2 dx.
179
0
1
ft2 + fx2 + ln(1 + f 2 ) dx.
x Rn , 0 < t < +
u(x) = 0
x .
ut(x, 0) = g(x)
u(x, 0) = f (x),
1 d
u
u2t dx +
a2 u2t dx
ut
u ut dx = 0,
ds +
2 dt
n
1 d
2 dt
=0, (u=0, x)
u2t dx
1 d
2 dt
u2t
a2 u2t dx
+ u
1 d
+
2 dt
dx =
u2 dx = 0,
a2 u2t dx 0.
u2t + u2 dx.
We have
dE
dt
u2 dx <
2
u dx C
u2 dx <
Thus,
,
,
,
by Poincare inequality.
180
181
Problem (S98, #4). a) Let u(x, y, z, t), < x, y, z < be a solution of the
equation
wtt + wt = w,
w(x, y, z, 0) = 0,
wt (x, y, z, 0) = 0.
Multiply the equation by wt and integrate:
wtwtt + wt2 = wtw,
2
wt wtt dx +
wt dx =
wtw dx,
3
3
R3
R
R
w
1 d
2
2
dx
w dx +
wt dx =
wt
w wt dx,
2 dt R3 t
n
R3
R3
R3
=0
1 d
1 d
wt2 dx +
wt2 dx =
w2 dx,
2 dt R3
2 dt R3
R3
2
d
wt2 dx 0,
wt + w2 dx = 2
dt R3
3
R
E(t)
dE
dt
0,
wt(x, 0)2 + w(x, 0)2 dx = 0,
E(t) E(0) =
3
R
2
wt + w2 dx = 0.
E(t) =
R3
182
Problem (F94, #8). The onedimensional, isothermal uid equations with viscosity
and capillarity in Lagrangian variables are
vt ux = 0
ut + p(v)x = uxx vxxx
in which v(= 1/) is specic volume, u is velocity, and p(v) is pressure. The coecients
and are nonnegative.
Find an energy integral which is nonincreasing (as t increases) if > 0 and conserved if = 0.
Hint: if = 0, E = u2 /2 P (v) dx where P (v) = p(v).
Proof. Multiply the second equation by u and integrate over R. We use ux = vt .
Note that the boundary integrals are 0 due to nite speed of propagation.
uut + up(v)x = uuxx uvxxx ,
uut dx +
up(v)x dx = uuxx dx uvxxx dx,
R
R
R
R
2
1
(u ) dx +
up(v) ds + ux p(v) dx
2 R t
R
R
=0
2
uux dx ux dx
uvxx dx +
ux vxx dx,
=
R
R
R
R
=0
=0
2
1
(u ) dx +
vtp(v) dx = u2x dx + vt vxx dx,
2 R t
R
R
R
2
1
2
(u ) dx +
P (v) dx = ux dx +
vt vx dx
vxt vx dx,
2 R t
R t
R
R
R
=0
2
1
(u ) dx +
P (v) dx +
(vx ) dx = u2x dx,
2 R t
2 R t
R t
R
2
2
u
d
+ P (v) + vx dx = u2x dx 0.
dt R 2
2
R
2
u
+ P (v) + vx2 dx
E(t) =
2
2
R
is nonincreasing if > 0, and conserved if = 0.
183
(ux )
x
(16.42)
utx (ux) dx =
dt 0 2
0
=0, (2periodic)
d
Q(ux) = (ux)uxt. Thus,
Let Q (z) = (z), then dt
1
d 1
utx(ux ) dx =
Q(ux ) dx.
=
dt 0
0
d 1
u2t
+ Q(ux) dx = 0.
dt 0
2
b) We have
u2t
+ Q(ux ).
2
41 For F to be convex, the Hessian matrix of partial derivatives must be positive denite.
F (ut , ux) =
41
for all y.
2 f (x)
.
xi xj
For functions with continuous second derivatives, it will always be symmetric matrix: fxi xj = fxj xi .
y F (x) y =
T
Fut ux
Fux ux
y1 y2
=
1
0
0 (ux )
1
0
0 (ux)
y1
y2
184
.
= y12 + (ux )y22
0.
need
Thus, for a Hessian matrix to be positive denite, need (ux) 0, so that the above
inequality holds for all y.
c) We have
d 1
F (ut , ux ) dx
dt 0
1
F (ut , ux ) dx
0
1
F (ut , ux ) dx
0
1
2
ut
+ Q(ux) dx
2
0
= 0,
= const,
1
=
F (ut (x, 0), ux(x, 0)) dx,
0
1
2
v0
+ Q(u0x ) dx.
=
2
0
a2 u4x
u2t
+
,
2
12
= 0,
= const,
1
2
a2 u0 4x
v0
=
+
dx.
2
12
0
42
185
ut + uux = uxxx,
1
1 2
ux (x, t) + u3 (x, t) dx
I2 (u) =
2
6
0
is also independent of the time t.
c) Assume the initial data are such that I1 (f ) + I2 (f ) < . Use (a) + (b) to prove
that the maximum norm of the solution, u = supx u(x, t), is bounded in time.
Hint: Use the following inequalities (here, up is the Lp norm of u(x, t) at xed time
t):
2
186
d 2
1 2 1 3
1
dI2
(u) =
ux + u dx =
ux uxt + u2 ut dx =
dt
dt 0
2
6
2
0
We dierentiate the original equation with respect to x:
ut = uux + uxxx
utx = (uux )x + uxxxx .
1 2 2
ux ((uux )x + uxxxx ) dx +
u (uux + uxxx ) dx
2 0
0
2
2
1 2 3
1 2 2
ux (uux)x dx +
ux uxxxx dx
u ux dx +
u uxxx dx
2 0
2 0
0
0
2
2
2
2
uxxuux dx + ux uxxx 0
uxx uxxx dx
ux uux 0 +
0
0
2 1 2
1 2
u4
1
dx + u2 uxx0
2uux uxx dx
2 0
4 x
2
2 0
2
2
2
1 u4 2
uxx uux dx
uxx uxxx dx
uux uxx dx
2 4 0
0
0
0
2
2
2
2 2
uxx uxxx dx = uxx 0 +
uxxx uxx dx =
uxxx uxx dx = 0,
=
=
=
=
=
since
2
2
uxxuxxx dx = + 0
2
1 2
ux (x, t) +
I2 (u) =
2
0
0
1
1 2 1 3
1
ux + u dx = ux22 + u33.
I2 (u) =
2
6
2
6
0
Using given inequalities, we have
1
(u22 + ux22 )
I1 (u) + 2I2 (u) u33
u2
6
6
3
2
I1 (u) + I2 (u) + u2 u I1 (u) + I2 (u) + I1 (u)u
6
3
18
6
3
18
= C + C1 u.
u2 C + C1 u,
u C2 .
Thus, u is bounded in time.
Also see Energy Methods problems for higher order equations (3rd and
4th) in the section on Gas Dynamics.
16.7
187
ut (x, y, z, 0) =
0 .
h(x)
Proof.
We may use the Kirchhos formula:
t
1
t
g(x + ct) dS +
h(x + ct) dS
u(x, t) =
4 t
4 =1
=1
1
2
2
t
=
(x1 + ct1 ) + (x2 + ct2 ) dS + 0 =
4 t
=1
We may solve the problem by Hadamards method of descent, since initial conditions are independent of x3 . We need to convert surface integrals in R3 to domain
integrals in R2 . Specically, we need to express the surface measure on the upper half
2
in terms of the two variables 1 and 2 . To do this, consider
of the unit sphere S+
'
1 12 22
over the unit disk 12 + 22 < 1.
f (1 , 2 ) =
'
d1 d2
1 + (f1 )2 + (f2 )2 d1 d2 =
.
dS =
1 12 22
188
1
g(x1 + ct1 , x2 + ct2 ) d1 d2
2t
u(x1 , x2 , t) =
4 t
1 12 22
12 +22 <1
h(x1 + ct1 , x2 + ct2 ) d1d2
t
2
+
4
1 12 22
2 + 2 <1
1 2
(x1 + t1 )2 + (x2 + t2 )2
1
d1 d2 + 0,
2t
=
4 t
1 12 22
12 +22 <1
x21 + 2x1 t1 + t2 12 + x22 + 2x2 t2 + t2 22
1
t
d1 d2
=
2 t
1 12 22
12 +22 <1
tx21 + 2x1 t2 1 + t3 12 + tx22 + 2x2 t2 2 + t3 22
1
d1 d2
=
2 t
1 12 22
12 +22 <1
x21 + 4x1 t1 + 3t2 12 + x22 + 4x2 t2 + 3t2 22
1
d1 d2
=
2
1 12 22
12 +22 <1
(x21 + x22 ) + 4t(x1 1 + x2 2 ) + 3t2 (12 + 22 )
1
d1 d2
=
2
1 12 22
12 +22 <1
4t
d d
x + x2 2
1 2
1 2
1 1
+
(x1 + x22 )
d1 d2
=
2
2
2
2
2
2
2
2
1 1 2
1 12 22
1 +2 <1
1 +2 <1
2 + 22
3t2
1
d1 d2 =
+
2 12 +22 <1 1 12 22
2 1
1 2
d1 d2
r dr d
1 2
2
2
(x1 + x2 )
(x1 + x2 )
=
=
2
2
1 r2
1 12 22
0
0
12 +22 <1
2
1 1
2 du d
1 2
(x1 + x22 )
=
2
u = 1 r 2 , du = 2r dr
1
2
0
0
u2
2
1 2
(x + x22 )
1 d = x21 + x22 .
=
2 1
0
1 2
2
4t 1
x1 1 + x2 2
x1 1 + x2 2
4t
d1 d2 =
d1 d2
=
2
2
2 12 +22 <1 1 1 2
2 1 122 1 12 22
= 0.
3t2 2 1 (r cos )2 + (r sin )2
12 + 22
3t2
d1 d2 =
r drd
=
2 12 +22 <1 1 12 22
2 0
1 r2
0
r3
3t2 2 1
drd
u = 1 r 2 , du = 2r dr
=
2 0
1 r2
0
2
2
t2 2
2
3t
d =
d = 2t2 .
=
2 0 3
0
Check:
u(x, y, z, 0) = x2 + y 2 .
ut (x, y, z, t) = (x + t) + (y + t) (x t) (y t),
ut (x, y, z, 0) = 0.
utt (x, y, z, t) = 4,
ux (x, y, z, t) = (x + t) + (x t),
uxx (x, y, z, t) = 2,
uy (x, y, z, t) = (y + t) + (y t),
uyy (x, y, z, t) = 2,
uzz (x, y, z, t) = 0,
utt = uxx + uyy + uzz .
189
190
x R2 , t > 0,
ut (x, 0) = h(x)
u(x, 0) = g(x),
x R2 .
The result is the consequence of the Huygens principle and may be proved by Hadamards
method of descent: 43
1
g(x1 + ct1 , x2 + ct2 ) d1 d2
2t
u(x, t) =
4 t
1 12 22
12 +22 <1
h(x1 + ct1 , x2 + ct2 ) d1d2
t
2
+
4
1 12 22
12 +22 <1
1
th(x + ) + g(x + ) d1 d2
'
=
2 2 <c2 t2
c2 t2
2
1 c2 t2
t
g(x + ) (ct, ct) d1 d2
'
+
.
2 2 <c2 t2
c2 t2
2
1 c2 t2
For a given x, let T (x) be so large that T > 1 and supp(h + g) BT (x). Then for
t > 2T we have:
tM + M + 2M ct d1 d2
1
'
u(x, t) =
2 2
2 2 <c2 T 2
c2 t2
1 c T
2
c T
2
2
c2 T 2 4
M 1
M 1
2M c
+
+
.
c2 t
3/4 c2 t
3/4 c2 T t
191
area of intersection of the sphere of radius ct with the support of the functions g and
h.
192
Problem (S95, #6). Spherical waves in 3d are waves symmetric about the origin;
i.e. u = u(r, t) where r is the distance from the origin. The wave equation
utt = c2 u
then reduces to
1
2
utt = urr + ur .
2
c
r
(16.43)
a) Find the general solutions u(r, t) by solving (16.43). Include both the incoming waves
and outgoing waves in your solutions.
b) Consider only the outgoing waves and assume the nite outux condition
0 < lim r 2 ur <
r0
for all t. The wavefront is dened as r = ct. How is the amplitude of the wavefront
decaying in time?
Proof. a) We want to reduce (16.43) to the 1D wave equation. Let v = ru. Then
vtt = rutt,
vr = rur + u,
vrr = rurr + 2ur .
Thus, (16.43) becomes
1
1 1
vtt =
vrr ,
2
c r
r
1
vtt = vrr ,
c2
vtt = c2 vrr ,
which has the solution
v(r, t) = f (r + ct) + g(r ct).
Thus,
1
1
1
f (r + ct) + g(r ct) .
u(r, t) = v(r, t) =
r
r
r
incoming, (c>0)
outgoing, (c>0)
1
1
0 < lim r 2 g (r ct) 2 g(r ct) < ,
r0
r
r
0 < lim rg (r ct) g(r ct) < ,
r0
t
.
g(t) = G
c
193
194
Problem (S00, #8). a) Show that for a smooth function F on the line, while
u(x, t) = F (ct + x)/x may look like a solution of the wave equation utt = c2 u
in R3 , it actually is not. Do this by showing that for any smooth function (x, t) with
compact support
u(x, t)(tt ) dxdt = 4 (0, t)F (ct) dt.
R3 R
Note that, setting r = x, for any function w which only depends on r one has
w = r 2 (r 2 wr )r = wrr + 2r wr .
b) If F (0) = F (0) = 0, what is the true solution to utt = u with initial conditions
u(x, 0) = F (x)/x and ut(x, 0) = F (x)/x?
c) (Ralston Hw) Suppose u(x, t) is a solution to the wave equation utt = c2 u in
R3 R with u(x, t) = w(x, t) and u(x, 0) = 0. Show that
u(x, t) =
x>
x=
u
u
dS .
n
n
The nal equality is derived by integrating by parts twice in t, and using Greens
theorem:
u
v
u
ds.
(vu uv) dx =
v
n
n
= r
, substituting u(x, t) = F (x + ct)/x
Since dS = 2 sin d d and n
gives:
u (tt ) dxdt =
4F (ct) dt.
R3
and
limr0 w(r, t) does not exist unless F (ct) + G(ct) = 0 for all t. Hence
w(r, t) =
u(x, t) =
F (ct + r) + G(ct r)
,
and
r
F (ct + x) + G(ct x)
.
x
195
17
196
44
K(x) = (x)
The fundamental solution for the Laplace operator is
1
if n = 2
2 log x
K(x) =
1
2n
x
if n 3.
(2n)n
17.1
45
(17.1)
G(x, )
g(x) dSx,
nx
We know that u(x) = Rn K(x y)f (y) dy is a distribution solution of u = f when f is integrable
and has compact support. In particular, we have
K(x y)u(y) dy
whenever u C0 (Rn ).
u(x) =
44
Rn
45
197
198
and hence
K(x ) =
x n ,
xn
n
so that the Poisson kernel is given by
2n
G(x, )xn =0 =
x n ,
xn
n
for x Rn1 .
2n
G(x, )
g(x) dSx =
nx
n
Rn1
g(x)
dx .
x n
If g(x ) is bounded and continuous for x Rn1 , then u() is C and harmonic in Rn+
and extends continuously to Rn+ such that u( ) = g( ).
199
x ,
< x1 < .
1
2
log x , n = 2.
1
log (x x0 )2 + (y y0 )2 + ((x, y), ),
2
1
y y0
G
((x, y), ) =
+ y ((x, y), ),
y
2 (x x0 )2 + (y y0 )2
1
y0
G
((x, 0), ) =
+ y ((x, 0), ) = 0.
y
2 (x x0 )2 + y02
G((x, y), ) =
Let
a
log (x x0 )2 + (y + y0 )2 .
Then,
2
1
y0
y0
a
G
((x, 0), ) =
+
= 0.
2
2
y
2 (x x0 ) + y0
2 (x x0 )2 + y02
((x, y), ) =
Thus, a = 1.
G((x, y), ) =
1
1
log (x x0 )2 + (y y0 )2 +
log (x x0 )2 + (y + y0 )2 .
2
2
46
u
G
dS
(uG G u ) dx =
G
u
n
n
=0
=0
46
Note that for the Dirichlet problem, we would have gotten the  sign instead of + in front of
= f (x), we have
u (x ) dx =
G((x, y), ) f (x) dx,
G((x, y), ) f (x) dx.
u() =
u
n
u
(y)
For y = 0, we have
G((x, y), ) =
=
=
Thus,
u() =
1
2
'
'
1
1
2
2
log (x x0 ) + y0 +
log (x x0 )2 + y02
2
2
'
1
2 log (x x0 )2 + y02
2
1
log (x x0 )2 + y02 .
2
log (x x0 )2 + y02 f (x) dx.
f (x1 ) dx1
= 0.
n
Note that the Greens identity applies to bounded domains .
2
R
u
R d = 0.
f dx1 +
r
R
0
???
200
201
x (1)  x (3)
.
x (2)
1
x (1) x (3)
1
.
log x 
log
2
2
x (2)
G(x, ) = 0, x .
x (0)  x (3)
1
log
2
x (1)  x (2)
(x x0 )2 + (y y0 )2 (x + x0 )2 + (y + y0 )2
1
log
.
2
(x x0 )2 + (y + y0 )2 (x + x0 )2 + (y y0 )2
202
Problem (S96, #3). Construct a Greens function for the following mixed DirichletNeumann problem in = {x = (x1 , x2 ) R2 : x1 > 0, x2 > 0}:
u =
2u 2u
+
= f,
x21 x22
ux2 (x1 , 0) = 0,
x1 > 0,
u(0, x2) = 0,
x2 > 0.
x ,
1
2
log x , n = 2.
1
1
2 2(y y0 )
+ wy (x, y)
2 (x x0 )2 + (y y0 )2
y y0
1
+ wy (x, y).
2 (x x0 )2 + (y y0 )2
a
log (x + x0 )2 + (y y0 )2
2
b
log (x x0 )2 + (y + y0 )2
+
2
c
log (x + x0 )2 + (y + y0 )2 .
+
2
boundary conditions, we have
'
1
log x20 + (y y0 )2 + (0, y)
G((0, y), ) =
2
'
'
'
'
a
b
c
1
log x20 + (y y0 )2 +
log x20 + (y y0 )2 +
log x20 + (y + y0 )2 +
log x20 + (y + y0 )2 .
2
2
2
2
a = 1, c = b. Also,
y0
1
+ wy (x, 0)
Gy ((x, 0), ) =
2 (x x0 )2 + y02
y0
y0
y0
y0
(1)
b
(b)
1
+
+
.
2
2
2
2
2
2
2 (x x0 ) + y0
2 (x + x0 ) + y0
2 (x x0 ) + y0
2 (x + x0 )2 + y02
((x, y), ) =
Using
0 =
=
Thus,
0 =
=
Thus, b = 1, and
1
1
2
2
log (x x0 ) + (y y0 ) + (x) =
log (x x0 )2 + (y y0 )2
G((x, y), ) =
2
2
(x + x0
)2
+ (y y0
203
2
2
2
2
+ log (x x0 ) + (y + y0 ) log (x + x0 ) + (y + y0 ) .
204
for x = a.
x  =

x .
a

1
2
log a x 
if n = 2
n2
K(x ) =
a
K(x )
if n 3.

(17.3)
Dene for x, :
K(x ) 1 log  x 
2
a
G(x, ) =
K(x ) a n2 K(x )
if n = 2
if n 3.

a2 2
G(x, )
g(x) dSx =
nx
an
x=a
(17.2)
g(x)
dSx.
x n
17.2
205
Problem (F99, #2). Given that Ka(x y) and Kb (x y) are the kernels for
the operators ( aI)1 and ( bI)1 on L2 (Rn ), where 0 < a < b, show that
( aI)( bI) has a fundamental solution of the form c1 Ka + c2 Kb .
Use the preceding to nd a fundamental solution for 2 , when n = 3.
Proof. METHOD :
( aI)u = f
( bI)u = f
u = Ka f
fundamental solution
*af
u
=K
( aI)( bI)u
(a + b) + abI u
1
f()
u
= 2
( + a)( 2 + b)
1
new =
K
2
( + a)( 2 + b)
1
(Kb Ka),
Knew =
ba
1
1
, c2 =
.
c1 =
ba
ba
2
*b f
u
=K
if u L2 ,
(
bI)u = (2 b)
u = f
u = f
(
aI)u = (2 a)
1
f()
u
= 2
( + a)
*a = 1
K
2 + a
u = Kb f
kernel
1
f()
( 2 + b)
*b = 1
K
2 + b
u
=
= f,
= f,
new f(),
= K
=
1
1
1
b K
a),
2
+ 2
=
(K
ba
+b +a
ba
1
f,
+ 2)
1
1
1
1
1 K
0.
= 2 2
= 2
+ 2 = K
4
2
( + )
( + 1)
+1
( 4
206
METHOD :
For u C0 (Rn ) we have:
Ka(x y) ( aI) u(y) dy,
u(x) =
Rn
Kb(x y) ( bI) u(y) dy.
u(x) =
Rn
Let
u(x) = c1 ( bI) (x),
u(x) = c2 ( aI) (x),
for
for
Rn
1
, we have:
If c1 = c2 and (c1 b + c2 a) = 1, that is, c1 = ab
1
(Ka Kb ) ( aI) ( bI) (y) dy,
(x) =
a
b
n
R
1
ab (Ka
207
Thus,
w = c1 er + c2 er ,
er
er
1
w(r) = c1 + c2
.
K =
r
r
r
Suppose v(x) 0 for x R and let = BR (0); for small > 0 let
= B (0).
K(x)
K(x)
v
v
v
dS +
v
dS
K(x)v vK(x) dx =
K(x)
K(x)
n
n
n
n
B(0)
=0, since v0 f or xR
v
K(x)
K(x)( I)v v ( I)K(x) dx =
K(x)
v
dS.
n
n
B(0)
= 0, in
lim
0
K(x)( I)v dx =
Since K(r) = c1
er
er
+ c2
is integrable at x = 0.
r
r
+
c
4
K(x)
max
c
1
2
n
B(0)
B(0) n
K(x)
dS
v(x)
n
B (0)
1
c1 e + c2 e + 2 c1 e + c2 e v(x) dS
B (0)
1
1
c1 e + c2 e
v(x) dS
+ 2 c1 e + c2 e
=
B(0)
1
1
c1 e + c2 e + 2 c1 e + c2 e
v(0) dS
=
B(0)
1
1
c1 e + c2 e
[v(x) v(0)] dS
+ 2 c1 e + c2 e
+
B (0)
1
2 c1 e + c2 e v(0) 42
4(c1 + c2 )v(0) = v(0).
1
1
, which gives
Thus, taking c1 = c2 , we have c1 = c2 = 8
K(x)( I)v dx = lim
K(x)( I)v dx = v(0),
0
47
In R3 , for x = ,
K(x)
K() = c1
K(x)
n
e
e
+ c2
.
e
e
K()
e
e
1
1
= c1
2 c2
2
c1 e + c2 e + 2 c1 e + c2 e ,
=
r
since n points inwards. n points toward 0 on the sphere x = (i.e., n = x/x).
208
1 er
er
1
+
cosh(r) is the fundamental solution of
= 4r
that is K(r) = 8
r
r
( I).
1
(Ka Kb ) is a fundamental solution of ( aI)( bI).
By part (a), ab
1
1
(K0 K) = 4r
+
Here, the fundamental solution of ( 0I)( 1I) is 1
1
1
4r cosh(r) = 4r 1 cosh(r) .
209
1 cos kx
4 x
is a fundamental solution for ( + k2 ) in R3 where x = x21 + x22 + x23 ,
i.e. prove that for any smooth function f (x) with compact support
cos kx y
1
f (y) dy
u(x) =
4 R3 x y
is a solution to ( + k2 )u = f .
1
Proof. For v C0 (Rn ), we want to show that for K(x) = 4
we have ( + k2 )K = , i.e.
K(x) ( + k2 )v(x) dx = v(0).
cos kx
x ,
Rn
Suppose v(x) 0 for x R and let = BR (0); for small > 0 let
= B (0).
K(x)
K(x)
v
v
v
dS +
v
dS
K(x)v vK(x) dx =
K(x)
K(x)
n
n
n
n
B(0)
=0, since v0 f or xR
K(x)
v
2
2
v
dS.
K(x)( + k )v v ( + k )K(x) dx =
K(x)
n
n
B(0)
= 0, in
K(x)( + k )v dx =
lim
0
K(x)( + k2 )v dx.
cos kr
is integrable at x = 0.
Since K(r) =
4r
4
K(x)
max
n
4
B(0)
B(0) n
48
In R3 , for x = ,
K(x)
K() =
K(x)
n
cos k
.
4
cos k
1
K()
1
k sin k
cos k
k sin k +
,
=
=
2
r
4
4
since n points inwards. n points toward 0 on the sphere x = (i.e., n = x/x).
210
cos k
K(x)
1
dS =
k sin k +
v(x) dS
n
B (0)
B(0) 4
1
cos k
=
k sin k +
v(x) dS
4
B (0)
cos k
cos k
1
k sin k +
k sin k +
v(0) dS
[v(x) v(0)] dS
=
4
4
B (0)
B(0)
cos k
cos k
1
k sin k +
v(0) 42
k sin k +
[v(x) v(0)] 42
=
4
4
v(x)
0, (v is continuous)
K(x)( + k )v dx = lim
0
K(x)( + k2 )v dx = v(0),
1 cos kr
is the fundamental solution of + k2 .
that is, K(r) = 4
r
x R3
1
u
,
iku = O 2 , x = r .
u=O
r
r
r
Prove that u 0.
Hint: A fundamental solution to the Helmholtz equation is
Use the Green formula.
1 ikr
.
4r e
1 ikr
e , a fundamental solution. Thus, ( + k2 )K = .
Proof. Denote K(x) = 4r
Let x0 be any point and = BR (x0 ); for small > 0 let
= B (x0 ).
( + k2 )K(x) = 0 in . Consider Greens identity ( = B (x0 )):
K
u
u
K
2
2
K
dS +
K
dS .
u ( + k )K K( + k )u dx =
u
u
n
n
n
n
B (x0 )
=0
u(x0 ), as 0
(It can be shown by the method previously used that the integral over B (x0 ) approaches u(x0 ) as 0.) Taking the limit when 0, we obtain
u
eikxx0  u
K
eikxx0 
K
dS =
dS
u
u
u(x0 ) =
n
n
r 4x x0 
4x x0  r
eikxx0 
eikxx0  u
eikxx0 
ik
iku dS
u
=
r 4x x0 
4x x0 
4x x0  r
= O(
1
x2
); (can be shown)
1
2
O
4R
4R2 = 0.
= O
R
R2
R
R2
Taking the limit when R , we get u(x0 ) = 0.
211
1
log x,
2
w
=0
n
in D.
and
2 w = f
on D,
,
r log r dr =
rur =
2
4
8
r
r log r
,
ur =
4
8
1
1 2
1
r log r 1 .
r log r dr
r dr =
u =
4
8
8
1 2
r log r 1 .
8
u(r) =
1 2
r log r is the
See the next page that shows that u dened as u(r) = 8
1 2
2
Fundamental Solution of . (The 8 r term does not play any role.)
R2
u(x y) 2 (y) dy =
1
8
R2
x y2 log x y 1 f (y) dy.
212
b) Let
1
x 2 log x  1 .
8
We use the method of images to construct the Greens function.
Let G(x, ) = K(x ) + (x). We need G(x, ) = 0 and G
n (x, ) = 0 for x .
w
2
Consider w (x) with w (x) = 0 in , w (x) = K(x ) and n (x) = K
n (x )
on .
Note, we can nd the Greens function for the upperhalf plane, and then
make a conformal map onto the domain.
K(x ) =
213
1 2
r ln(r),
8
r = x ,
x = (x1 , x2).
1 2
r r (r r ) + r 2 2 ;
for example, V =
1
2 (1 + ln(r)).
We have
1 2
r log(r),
8
In polar coordinates: (here, V = 0)
1
1 2
1 1
1
rVr r =
r log(r)
r
r 2r log(r) + r
=
V =
r
r
8
r r
8 r
r
1
1 1
1
2r 2 log(r) + r 2
4r + 4r log r
=
=
8 r
8 r
r
1
(1 + log r).
=
2
V (x) =
The fundamental solution V (x) for 2 is the distribution satisfying: 2 V (r) = (r).
1
1
1 1
(1 + log r) =
(1 + log r) =
r(1 + log r)r r
2 V = (V ) =
2
2
2 r
1 1
1
1 1
r
(1)r = 0
=
=
for r = 0.
2 r r r
2 r
Thus, 2 V (r) = (r) V is the fundamental solution.
The approach above is not rigorous. See the next page that shows that
V dened above is the Fundamental Solution of 2 .
The solution of
2 = F (x),
if given by
(x) =
1
V (x y) (y) dy =
8
R2
2
R2
214
1 2
r ln(r),
8
r = x ,
(17.4)
Suppose v(x) 0 for x R and let = BR (0); for small > 0 let
= B (0).
K(x) is biharmonic (2 K(x) = 0) in . Consider Greens identity ( =
B (0)):
K(x)
K(x)
v
v
2
v
ds +
v
ds
K(x) v dx =
K(x)
K(x)
n
n
n
n
=0, since v0 f or xR
+
B(0)
K(x)2v dx =
K(x)v 2 dx.
lim
0
K(x)
v
v
ds +
K(x)
n
n
B(0)
K(x)
K(x)
v
v
ds.
n
n
Since K(r) is integrable at x = 0.
K(x)
n
x
B (0)
B(0) n
1 2
= log()n max (v) 0, as 0.
8
x
K(x)
dS =
v(x)
n
B (0)
1
v(x) dS
B(0) 2
1
1
v(0) dS +
[v(x) v(0)] dS
=
B(0) 2
B(0) 2
1
v(0) 2 max v(x) v(0) = v(0).
=
2
xB(0)
0, (v is continuous)
v
1
v
dS = K()
K(x)
dS (1 + log )2 max v 0, as 0.
n
2
x
B(0)
B (0) n
K(x)
dS =
v
n
B(0)
49
1
1
log
v(x) dS
4
8
B(0)
1
log + 2 max v 0, as 0.
4
2
xB(0)
K(x)
n
1 2
log ,
8
K()
1
1
= log
,
r
4
8
K() =
1
(1 + log ),
2
K
K
1
=
=
.
n
r
2
K =
K(x) v dx = lim
0
K(x)2 v dx = v(0).
215
17.3
216
Radial Variables
Problem (F99, #8). Let u = u(x, t) solve the following PDE in two spatial dimensions
u = 1
for r < R(t), in which r = x is the radial variable, with boundary condition
u=0
on r = R(t). In addition assume that R(t) satises
dR
u
= (r = R)
dt
r
with initial condition R(0) = R0 .
a) Find the solution u(x, t).
b) Find an ODE for the outer radius R(t), and solve for R(t).
Proof. a) Rewrite the equation in polar coordinates:
1
1
u = (rur )r + 2 u = 1.
r
r
For a radially symmetric solution u(r), we have u = 0. Thus,
1
(rur )r = 1,
r
(rur )r = r,
r2
rur = + c1 ,
2
r c1
ur = + ,
2
r
r2
u(r, t) = + c1 log r + c2 .
4
Since we want u to be dened for r = 0, we have c1 = 0. Thus,
r2
+ c2 .
4
Using boundary conditions, we have
u(r, t) =
u(R(t), t) =
u(r, t) =
R(t)2
+ c2 = 0
4
c2 =
R(t)2
.
4
r 2 R(t)2
+
.
4
4
b) We have
R(t)2
r2
+
,
4
4
r
,
2
u
R
(r = R) =
,
(from )
r
2
dt
,
2
t
,
2
t
R(0) = c1 = R0 .
Thus,
c1 e 2 ,
u(r, t) =
u
r
dR
dt
dR
R
=
=
=
log R =
R(t) =
Thus,
R(t) = R0 e 2 .
217
218
Problem (F01, #3). Let u = u(x, t) solve the following PDE in three spatial dimensions
u = 0
for R1 < r < R(t), in which r = x is the radial variable, with boundary conditions
u(r = R(t), t) = 0,
and
u(r = R1 , t) = 1.
2
1
2
u = 2 (r 2 ur )r = 0.
+
u =
2
r
r r
r
(r 2 ur )r = 0,
r 2 ur = c1 ,
c1
,
ur =
r2
c1
u(r, t) = + c2 .
r
Using boundary conditions, we have
c1
+ c2 = 0
u(R(t), t) =
R(t)
c1
+ c2 = 1.
u(R1 , t) =
R1
This gives
c1 =
R1 R(t)
,
R1 R(t)
u(r, t) =
c2 =
c2 =
c1
,
R(t)
R1
.
R1 R(t)
1
R1
R1 R(t)
+
.
R1 R(t) r R1 R(t)
b) We have
1
R1
R1 R(t)
+
,
R1 R(t) r R1 R(t)
R1 R(t)
1
=
2,
R1 R(t) r
u
R1 R(t)
1
R1
= (r = R) =
=
2
r
R1 R(t) R(t)
(R1 R(t)) R(t)
u(r, t) =
u
r
dR
dt
R0 > R1 .
(from )
219
1
(u )u + p u = 0
on the domain < x1 < , x22 + x23 R2 , where u = (u1 , u2 , u3 ) = (U (x2 , x3), 0, 0)
is the velocity vector, p(x1 , x2 , x3 ) is the pressure, and and are constants.
p
is a constant c, and that U = c/.
a) Show that x
1
b) Assuming further that U is radially symmetric and U = 0 on the surface of the pipe,
determine the mass Q of uid passing through a crosssection of pipe per unit time in
terms of c, , , and R. Note that
Q =
U dx2 dx3 .
{x22 +x23 R2 }
u u u
1
2
3
,
,
= (U (x2 , x3 ), 0, 0) (0, 0, 0) = 0.
(u )u = (u1 , u2 , u3 )
x1 x2 x3
Thus,
1
p u = 0,
p = u,
p p p
,
,
= (u1 , u2 , u3 ),
x1 x2 x3
p p p
,
,
= (Ux2x2 + Ux3 x3 , 0, 0).
x1 x2 x3
We can make the following observations:
p
x1
= (Ux2 x2 + Ux3 x3 ),
indep. of x1
p
x2
p
x3
= 0
p = f (x1 , x3 ),
= 0
p = g(x1, x2 ).
= U,
U
p
x1
c
1 p
= .
x1
is independent of x1 . Therefore,
p
x1
= c.
220
cr
,
rUr r =
cr 2
+ c1 ,
rUr =
2
c1
cr
+ .
Ur =
2
r
For Ur to stay bounded for r = 0, we need c1 = 0. Thus,
cr
Ur =
,
2
cr 2
+ c2 ,
U =
4
cR2
+ c2 ,
0 = U (R) =
4
c 2
cr 2 cR2
=
(r R2 ).
U =
4
4
4
Q =
=
{x22+x23 R2 }
cR4
U dx2 dx3
c
=
4
2
0
0
c
(r R ) rdrd =
4
2
0
R4
d
4
17.4
221
Weak Solutions
2
in
u=f
u=0
on
u
on
n = 0
provided
uv dx =
f v dx
for all test functions v H02 (). Prove that for each f L2 (), there exists a unique
weak solution for this problem. Here, H02 () is the closure of all smooth
functions in
which vanish on the boundary and with nite H 2 norm: u22 = (u2xx + u2xy +
u2yy ) dxdy < .
Hint: use LaxMilgram lemma.
Proof. Multiply the equation by v H02 () and integrate over :
2 u = f,
2
u v dx =
f v dx,
v
u
v ds
ds +
u
uv dx =
f v dx,
n
n
= 0
uv dx =
f v dx .
a(u,v)
L(v)
v2V ,
v V ;
v V.
u, v V ;
222
50
2
a(u, v)2 = uv dx
(u)2 dx
(v)2 dx u2H 2() v2H 2() .
a(v, v) =
(v)2 dx vH02() .
1
1
2
2
2
2
L(v) = f v dx
f  v dx
f dx
v dx
uH02()
v V.
CauchySchwarz Inequality:
(u, v)
a(u, v)
vdx
a(u, u) 2 a(v, v) 2 ;
1
1
v 1 dx = ( v2 dx) 2 ( 12 dx) 2 .
uvdx (
u2 dx) 2 (
v 2 dx) 2 ;
Poincare Inequality:
vH 2() C (v)2 dx.
Greens formula:
(u)2 dx =
(u2xx + u2yy + 2uxx uyy ) dxdy = (u2xx + u2yy 2uxxy uy ) dxdy = (u2xx + u2yy + 2uxy 2 ) dxdy u2H 2().
17.5
223
Uniqueness
Setting w = u = v gives
w
2
ds
w dx =
w
w w dx .
n
=0
in .
Setting w = u = v gives
w2 dx =
=0,
ds
w
n
Dirichlet or Neumann
w w dx.
224
w dx =
0
q(x)w 2 dx .
0
Thus, w 0, and u1 u2 . Hence, the solution to the Dirichlet and Neumann problems
are unique.
Problem (F02, #8; S93, #5).
Let D be a bounded domain in R3 . Show that a solution of the boundary value problem
2 u = f
in D,
u = u = 0
on D
is unique.
Proof. Method I: Maximum Principle. Let u1 , u2 be two solutions of the boundary
value problem. Dene w = u1 u2 . Then w satises
2 w = 0 in D,
w = w = 0 on D.
So w is harmonic and thus achieves min and max on D w 0.
So w is harmonic, but w 0 on D w 0. Hence, u1 = u2 .
Method II: Greens Identities. Multiply the equation by w and integrate:
w2 w = 0,
w2 w dx = 0,
(w)
ds
w
w()w dx = 0,
n
=0
w
w ds + (w)2 dx = 0.
=0
and
u = 0.
Prove that u 0.
b) What can you say about u(x) when the condition u(x) 0 in is dropped?
225
=0
3
u + u2 dx = 0.
226
n
k=1
u
u3 = 0
xk
u=0
in ,
on ,
u
u u4 = 0,
xk
k=1
n
u
uu dx +
k
u dx
u4 dx = 0,
x
k
k=1
n
u
u
ds u2 dx +
u
k
u dx
u4 dx = 0.
xk
n
k=1
=0
=0
u
k
u dx = 0,
2
xk
n
u
k
u dx = 0.
xk
uu +
k=1
Thus, we have
u4 dx = 0,
u2 dx
2
u4 dx = 0.
u +
Note that
n
k=1
u
u dx =
xk
u u dx =
nu2 ds
=0
and thus,
u u dx = 0.
u u dx,
227
in D,
u
u(x1 , 0) =
(0, x2) = 0
x1
can have only one bounded solution.
b) Find an explicit Greens function for this boundary value problem.
Proof. a) Let u1 , u2 be two solutions of the boundary value problem. Dene w =
u1 u2 . Then w satises
w = 0
in D,
w
(0, x2 ) = 0.
w(x1 , 0) =
x1
Consider Greens formula:
u
ds
u v dx =
v
v u dx.
D
D n
D
Setting w = u = v gives
w
2
ds
w dx =
w
w w dx,
n
D
D
D
w
w
w
ds +
ds +
ds
w2 dx =
w
w
w
w w dx
n
n
n
D
Rx1
Rx2
x>R
D
w
w
w
ds
w(x1 , 0)
ds +
w(0, x2)
ds +
w
w w dx,
=
x1
n
Rx1 x2
Rx2
x>R
D
=0
=0
=0
=0
w2 dx = 0 w2 = 0 w = const.
D
x (0) =
So,
Problem (F98, #4). In two dimensions x = (x, y), dene the set a as
a = +
in which
+ = {x x0  a} {x 0}
= {x + x0  a} {x 0} = +
and x0 = (1, 0). Note that a consists of two components when 0 < a < 1
and a single component when a > 1. Consider the Neumann problem
2 u = f,
u/n = 0,
x a
x a
in which
+
f (x) dx = 1
f (x) dx = 1
a) Show that this problem has a solution for 1 < a, but not for 0 < a < 1.
(You do not need to construct the solution, only demonstrate solveability.)
b) Show that maxa u as a 1 from above. (Hint: Denote L to be
the line segment L = + , and note that its length L goes to 0 as a 1.)
Proof. a) We use the Greens identity. For 1 < a,
u
ds =
u dx =
f (x) dx
0 =
n
a
a
a
f (x) dx +
f (x) dx = 1 1 = 0.
=
+
We get contradictions.
Thus, the solution does not exist for 0 < a < 1.
228
229
L
u
u
u
u dx =
ds =
ds =
ds.
n
L n
L n
u
u dx
u dx = 2
ds,
+
+
L n
u
f (x) dx
f (x) dx = 2
ds.
+
+
L
u
ds,
2 = 2
+
L n
u 2 u 2
u
u
ds
+
L max u L max u.
1 =
+ ds
+
L
a
n+
L n
L n
L
Thus,
max u
a
1
.
L
As a 1 (L 0)
maxa u .
230
in D,
on D.
a) Assuming that a(x) is small enough, prove the uniqueness of the classical solution.
b) Prove the existence of the solution in the Sobolev space H 1 (D) assuming that f
L2 (D).
Note: Use Poincare inequality.
Proof. a) By Poincare Inequality, for any u C01 (D), we have u22 Cu22.
Consider two solutions of the Dirichlet problem above. Let w = u1 u2 . Then, w
satises
w + a(x)w = 0 in D,
w=0
on D.
ww + a(x)w 2
ww dx + a(x)w 2 dx
w2 dx + a(x)w 2 dx
w2 dx
a(x)w 2 dx =
1
w 2 dx
a(x)w 2 dx
C
1
w 2 dx
a(x) w 2 dx
C
1
w 2 dx
a(x)
C
If a(x) < C1
b) Consider
F (v, u) =
= 0,
= 0,
= 0,
1
C
w 2 dx,
0,
0,
0.
w 0.
(vu + a(x)vu) dx =
vf (x) dx = F (v).
v H01,2 (D).
231
in
on
1
(rur )r 4(r 2 + 1)u,
r
1
1
2
2
(rr )r 4(r 2 + 1) = (2r 2 er )r 4(r 2 + 1)er
r
r
1
2
2
r2
2
r2
(4re + 2r 2re ) 4r 2 er 4er = 0.
r
2
b) We have = er = ex
51
2 +y2
L = 0,
2
2
2
2
2
Multiply the rst equation by and the second by u and subtract the two equations:
Lu = u 4(r 2 + 1)u,
uL = u 4(r 2 + 1)u,
Lu uL = u u.
Then, we start from the LHS of the equality we need to prove and end up with RHS:
f dx =
Lu dx =
(Lu uL) dx =
(u u) dx
u
u
u ) ds =
ds =
(
ds.
=
n
n
51
The only shortcoming in the above proof is that we assume
n = (y, x), without giving an explanation why it is so.
17.6
232
SelfAdjoint Operators
n = (n1 , . . . , nn ) Rn ,
m
(D u)v dx = (1)
uD v dx,
(m = ).
We have
(Lu) v dx =
a (x)D u v dx =
a (x) v D u dx
m
(1)

m
m
D (a(x) v) u dx =
L (v) u dx,
L (v) =
(1)D (a(x) v)
m
vL(u) dx =
52
L = L
uL(v) dx.
(1)D (a(x) v) u dx
m
L (v)
233
=0
x
y
on x = 0
on x = y.
2 /2
cos
f ds
r
CR
in which CR is the circular arc of radius R connecting the boundaries x = 0 and x = y.
Proof. a) We have
Lu = u = 0
u
=0
x
u
u
=0
x
y
on x = 0
on x = y.
(u Lv v Lu) dx =
=
=
=
v
u
v
ds
u
n
n
v
u
u
v
v
ds +
v
ds
u
u
n
n
n
n
x=0
x=y
u (v n) v (u n) ds +
u (v n) v (u n) ds
x=0
x=y
u (vx , vy ) (1, 0) v (ux , uy ) (1, 0) ds
x=0
u (vx , vy ) (1/ 2, 1/ 2) v (ux, uy ) (1/ 2, 1/ 2) ds
+
x=y
u (0, vy ) (1, 0) v (0, uy ) (1, 0) ds
x=0
= 0
+
u (vy , vy ) (1/ 2, 1/ 2) v (uy , uy ) (1/ 2, 1/ 2) ds
x=y
uv
vu
y ( 1) y ( 1) ds
= 0.
2
2
x=y
(uv vu) dx =
need
234
n
CR n
x=0
x=y
CR r
=0
Thus,
CR
f
ds =
r
f dx =
er
2 /2
=0
cos r drd
1
2
1 R r2 /2
e
r dr = 1 (1 eR /2 ).
=
1
2 0
2
235
Problem (F99, #1). Suppose that u = 0 in the weak sense in Rn and that there
is a constant C such that
u(y) dy < C, x Rn .
{xy<1}
For v = 1, we have
u
ds =
u dx.
n
x=1
If xy<1 u(y) dy < C x Rn , we have u(x) < C in Rn .
Since u is harmonic and bounded in Rn , u is constant by Liouvilles theorem.
53
53
236
Problem (S01, #1). For bodies (bounded regions B in R3 ) which are not perfectly
conducting one considers the boundary value problem
3
u
0 = (x)u =
(x)
xj
xj
j=1
u = f
on B.
The function (x) is the local conductivity of B and u is the voltage. We dene
operator (f ) mapping the boundary data f to the current density at the boundary by
(f ) = (x)
u
,
n
and /n is the inward normal derivative (this formula denes the current density).
a) Show that is a symmetric operator, i.e. prove
g(f ) dS =
f (g) dS.
B
b) Use the positivity of (x) > 0 to show that is negative as an operator, i.e., prove
f (f ) dS 0.
B
Proof. a) Let
(x)u = 0
on ,
u=f
on .
(x)v = 0
on ,
v=g
on .
v
u
,
(g) = (x) .
n
n
Since /n is inward normal derivative, Greens formula is:
u
dS
v
(x)
v
(x)u
dx
=
v (x)u dx.
n
(f ) = (x)
=g
We have
u
dS =
g(f ) dS =
g(x)
v (x)u dx
v (x)u dx
n
=0
v
dS +
u(x)
u (x)v dx
=
n
=0
v
dS =
f (x)
f (g) dS.
=
n
u
dS = u (x)u dx
u(x)
(x)u u dx
n
=0
= (x)u2 0.
0
237
Problem (S01, #4). The Poincare Inequality states that for any bounded domain
in Rn there is a constant C such that
2
u dx C
u2 dx
&
Expanding u in the eigenfunctions of the Laplacian, u(x) = an n (x), the formula
above gives
2
u dx =
an n (x)
m am m (x) dx =
m an am
n m dx
n=1
m=1
n an 2 .
m,n=1
n=1
Also,
u dx =
n=1
an n (x)
am m (x) =
m=1
an 2 .
n=1
2
2
2
u dx = 1
an 
n an  =
u2 dx.
1
u2 dx
n=1
1
1
n=1
u2 dx,
with C = 1/1 .
b) For the rectangular domain = {(x, y) : 0 x a, 0 y b} R2 , the
eigenvalues of the Laplacian are
m2 n2
m, n = 1, 2, . . ..
mn = 2 2 + 2 ,
a
b
1
1
1 = 11 = 2 2 + 2 ,
a
b
1
1
1
.
= 2 1
C =
11
( a2 + b12 )
238
239
W = 0 in R /B
W =0
on B
W 0 as x .
Consider Greens formula:
u
ds
u v dx =
v
v u dx.
B
B n
B
Setting W = u = v gives
2
W  dx =
B
W
ds
W
n
B
=0
Thus, W 2 = 0 W = const.
B
W W dx = 0.
=0
Since W = 0 on B, W 0, and V1 = V2 .
b & c) For (b)&(c), see the solutions from Ralstons homework (a few pages
down).
240
Problem (W03, #2). Let L be the second order dierential operator L = a(x)
in which x = (x1 , x2 , x3 ) is in the threedimensional cube C = {0 < xi < 1, i = 1, 2, 3}.
Suppose that a > 0 in C. Consider the eigenvalue problem
Lu = u
for
xC
u=0
for
x C.
=0
(u2 + a(x)u2 ) dx
< 0.
=
2
u dx
(17.5)
v a(x)v = v.
(17.6)
Multiply (17.5) by v and (17.6) by u and subtract equations from each other
vu a(x)uv = uv,
uv a(x)uv = uv.
vu uv = ( )uv.
Integrating over gives
uv dx,
vu uv dx = ( )
v
u
u
uv dx.
v
dx = ( )
n n
=0
241
17.7
242
Spherical Means
2
2 2
2
u
+
+
u.
x2 y 2 z 2
a) Show that 2 is selfadjoint on x < 1 with the following boundary conditions on
x = 1:
u = 0,
u = 0.
Proof. a) We have
Lu = 2 u = 0
u = 0
on x = 1
u = 0
on x = 1.
(u 2 v v 2 u) dx
v
u
ds u (v) dx
ds +
u
v
v (u) dx
=
n
n
=0
=0
u
v
ds +
ds
v
uv dx +
u
vu dx = 0.
=
n
n
(u Lv v Lu) dx =
=0
=0
243
x=r
4 3
r
3
1
xr
u(x) dx.
Show that
d
r
S(r) = V (r).
dr
3
Hint: Rewrite S(r) as an integral over the unit sphere before dierentiating; i.e.,
1
u(rx ) dx.
S(r) = (4)
x =1
=
=
54
1
1
1
2
u(x) dSr =
u(rx ) r dS1 =
u(rx ) dS1 .
4r 2 x=r
4r 2 x =1
4 x =1
1
1
1
u
u
u
(rx ) dS1 =
(rx ) dS1 =
(x) dSr
2
4 x =1 r
4 x=1 n
4r x=r n
1
u dx.
4r 2 xr
r2
r2
u(0) + S(0) = u(0) +
u(0).
6
6
54
Change of variables:
Surface integrals: x = rx in R3 :
u(x) dS =
u(rx ) r2 dS1 .
x =1
x=r
Volume integrals: = r in Rn :
h(x + ) d =
h(x + r) rn d.
 <r
<1
244
d
dr
R 2
d
f (x) dx =
f (, , r) r 2 sin d d dr = ???
dr
0
0
0
xR
2
f (, , r) R2 sin d d
=
0
0
f (x) dS.
=
x=R
55
55
245
246
1
d
1 d
2
f (rx ) r dS1 =
f (rx ) dS1
dr 4r 2 x =1
4 dr
x =1
1
f
f
1
(rx ) dS1 =
(rx ) dS1
4 x =1 r
4 x =1 n
1
1
f
(x) dS =
f n dS
4r 2 x=r n
4r 2 x=r
1
f dx.
4r 2 xr
56
Note that part (a) was not used. We use exactly the same derivation as we did in S95 #4.
247
1
M (r) k2 Mx (r) =
r x
=
1
urr + ur k2 u d
r
0
2
1 2
1
u d =
u 0 = 0.
2
2r 0
2r 2
1
2
248
17.8
249
xBR1 (0)
u(x) = a < 0
(0) u(x)
1
1
1
.
x y2n =
(2 n)n
4 x y
4 R3 / x y
((u))(y)
1
dy
=
4 BR / x y
1
1
1
1
y (u) dy
u
dSy
y
=
4 BR /
x y
4 (BR/) n
x y
1
1
1
1
1
1
u dy +
u dSy
u
dSy
=
4 BR /
x y
4 (BR/) n x y
4 (BR/) n
x y
1
1
u
dSy
= ??? = u(x)
u dSy
4R2 B
4R B n
0, as R
= u(x).
c) See the next problem.
0, as R
250
Ralston Hw. a) Suppose that u is a smooth function on R3 and u = 0 for x > R.
1
If limx u(x) = 0, show that you can write u as a convolution of u with the 4x
and prove that limx xu(x) = 0 exists.
b) The conductor potential for R3 is the solution to the Dirichlet problem v =
0. The limit in part (a) is called the capacity of . Show that if 1 2 , then the
capacity of 2 is greater or equal the capacity of 1 .
Proof. a) If we dene
u(y)
1
dy,
v(x) =
4 R3 x y
then (u v) = 0 in all R3 , and, since v(x) 0 as x , we have limx (u(x)
v(x)) = 0. Thus, u v must be bounded, and Liouvilles theorem implies that it is
identically zero. Since we now have
1
x u(y)
xu(x) =
dy,
4 R3 x y
and x/x y converges uniformly to 1 on {y R}, it follows that
1
u(y) dy.
lim xu(x) =
4 R3
x
b) Note that part (a) implies that the limit limx xv(x) exists, because we can
apply (a) to u(x) = (x)v(x), where is smooth and vanishes on , but (x) = 1 for
x > R.
Let v1 be the conductor potential for 1 and v2 for 2 . Since vi as x and
vi = 1 on i , the max principle says that 1 > vi (x) > 0 for x R3 i . Consider
v2 v1 . Since 1 2 , this is dened in R3 2 , positive on 2 , and has limit 0 as
x . Thus, it must be positive in R3 2 . Thus, limx x(v2 v1 ) 0.
0
Proof. Let
2
1
u(x0 + R cos , y0 + R sin ) d,
2 0
w(r, ) = u(x0 + R cos , y0 + R sin ).
M (x0 , R) =
251
59
59
See ChiuYens solutions and Sung Has solutions (in two places). Nicks solutions, as started above,
have a very simplistic approach.
252
is nonempty. This set is closed because u is continuous. We will show it is also open.
This implies that K = because is connected. Thus u is constant on .
Let x0 K. Since is open, > 0, s.t. B (x0 ) = {x Rn : x x0  } . Let
sup u = M . By the mean value property, for 0 r
1
1
u(x
+
r)dS
,
and
0
=
(M u(x0 + r))dS .
M = u(x0 ) =
0
A(S n1 ) =1
A(S n1 ) =1
Sinse M u(x0 + r) is a continuous nonnegative function on , this implies M u(x0 +
r) = 0 for all S n1 . Thus u = 0 on B (x0 ).
b) Since u v has the mean value property in the open interior of Br (), by part
a) it satises the maximum principle. Since it is continuous on Br (), its supremum
over the interior of Br () is its maximum on Br (), and this maximum is assumed at a
point x0 in Br (). If x0 in the interior of Br (), then u v is constant ant the constant
must be zero, since this is the value of u v on the boundary. If x0 is on the boundary,
then u v must be nonpositive in the interior of Br ().
Applying the same argument to v u, one nds that it is either identically zero or
nonpositive in the interior of Br (). Thus, u v 0 on Br ().
Yes, it does follow that u harmonic in . Take v in the preceding to be the harmonic
function in the interior of Br () which agrees with u on the boundary. Since u = v on
Br (), u is harmonic in the interior of Br (). Since is open we can do this for every
. Thus u is harmonic in .
253
Ralston Hw. Assume is a bounded open set in Rn and the Greens function, G(x, y),
for exists. Use the strong maximum principle, i.e. either u(x) < sup u for all x ,
or u is constant, to prove that G(x, y) < 0 for x, y , x = y.
Proof. G(x, y) = K(x, y) + (x, y). For each x , f (y) = (x, y) is continuous on ,
thus, bounded. So (x, y) Mx for all y . K(x y) as y x. Thus,
given Mx , there is > 0, such that K(x y) < Mx when x y = r and 0 < r .
So for 0 < r the Greens function with x xed satises, G(x, y) is harmonic on
Br (x), and G(x, y) 0 on the boundary of Br (x). Since we can choose r as
small as we wish, we get G(x, y) < 0 for y {x}.
Problem (W03, #6). Assume that u is a harmonic function in the half ball
D = {(x, y, z) : x2 + y 2 + z 2 < 1, z 0} which is continuously dierentiable, and satises u(x, y, 0) = 0. Show that u can be extended to be a harmonic function in the whole
ball. If you propose and explicit extension for u, explain why the extension is harmonic.
Proof. We can extend u to all of nspace by dening
u(x , xn ) = u(x , xn )
for xn < 0. Dene
1
(x) =
an
y=1
a2 x2
v(y)dSy
x yn
254
on {0 < x < a/2}. Choose b, 0 < b < min(, a/2), so that /b > M . Then on both
{x = a/2} and {x = b} we have v 2/x < u(x) < v + 2/x.
Thus, by
max principle these inequalities hold on {b x a/2}. Pick x with 0 < x a/2.
u(x) = v(x). v is the extension of u on {x < a/2}, and u is extended on {x < a}.
18
u(0, t) = 0
for t > 0,
where g is continuous and bounded for x 0 and g(0) = 0.
Find a formula for the solution u(x, t).
Proof. Extend g to be an odd function on all of R:
g(x),
x0
g(x) =
g(x), x < 0.
Then, we need to solve
xx
u
t = u
u
(x, 0) = g(x)
for x R, t > 0
for x R.
u
(x, t) =
K(x, y, t)g(y) dy =
e 4t g(y) dy
4t
R
0
2
(xy)
(xy)2
1
e 4t g(y) dy +
e 4t g(y) dy
=
4t 0
(xy)2
(x+y)2
1
4t
4t
e
g(y) dy
e
g(y) dy
=
4t 0
0
2
x +2xyy 2
x2 2xyy 2
1
4t
4t
e
e
g(y) dy
=
4t 0
xy
(x2 +y 2 )
xy
1
e 4t
e 2t e 2t g(y) dy.
=
4t 0
xy
(x2 +y 2 )
1
g(y) dy.
u(x, t) =
e 4t 2 sinh
2t
4t 0
Since sinh(0) = 0, we can verify that u(0, t) = 0.
60
In calculations, we use:
0
ey dy =
0
255
256
ut = uxx
u(x, 0) = g(x)
for x > 0
for t > 0,
ux (0, t) = 0
where g is continuous and bounded for x 0.
Find a formula for the solution u(x, t).
Proof. Extend g to be an even function
g(x),
x0
g(x) =
g(x), x < 0.
Then, we need to solve
xx
u
t = u
u
(x, 0) = g(x)
61
on all of R:
for x R, t > 0
for x R.
4t
e 4t g(y) dy +
e
g(y) dy
=
4t 0
(xy)2
(x+y)2
1
4t
4t
e
g(y) dy +
e
g(y) dy
=
4t 0
0
2
x +2xyy 2
x2 2xyy 2
1
4t
4t
+e
e
g(y) dy
=
4t 0
xy
(x2 +y 2 )
xy
1
e 4t
e 2t + e 2t g(y) dy.
=
4t 0
xy
(x2 +y 2 )
1
g(y) dy.
u(x, t) =
e 4t 2 cosh
2t
4t 0
To check that the boundary condition holds, we perform the calculation:
xy
1
d (x2 +y2 )
4t
e
g(y) dy
2 cosh
ux (x, t) =
2t
4t 0 dx
xy
xy
(x2 +y 2 )
y
2x (x2 +y2 )
1
4t
4t
e
+e
g(y) dy,
2 cosh
2 sinh
=
4t
2t
2t
2t
4t 0
2
2
1
y
y4t
y4t
2 cosh 0 + e
2 sinh 0 g(y) dy = 0.
0e
ux (0, t) =
2t
4t 0
61
Even extensions are always continuous. Not true for odd extensions. g odd is continuous if g(0) =
0.
62
In calculations, we use:
0
ey dy =
0
257
ft = fxx
f (t = 0, x) = f0 (x)
with f0 smooth and bounded.
a) Write down the Greens function G(x, y, t) for this initial value problem.
b) Write the solution f (x, t) as an integral involving G and f0 .
c) Show that the maximum values of f (x, t) and fx(x, t) are nonincreasing
as t increases, i.e.
sup f (x, t) sup f0 (x)
x
63
n
(xy)2
2
1
4k(ts)
G(x, t; y, s) =
e
.
(2)n k(t s)
1
(xy)2
(xy)2
1
4t
4t
f0 (y) dy
e
f
(y)
dy
e
0
4t
4t R
R 2
y
1
e 4t f0 (x y) dy
4t R
y2
dy
1
y
e 4t dy
z = , dz =
sup f0 (x)
x
4t R
4t
4t
1
2
ez 4t dz
sup f0 (x)
x
4t R
1
2
ez dz = sup f0 (x).
sup f0 (x)
R
x
x
=
63
The Greens function for the heat equation on an innite domain; derived in R. Haberman using
the Fourier transform.
258
1
2(x y) (xy)2
1
d (xy)2
4t
4t
e
e
f0 (y) dy =
f0 (y) dy
4t
dy
4t R
4t
R
(xy)2
(xy)2
1
1
4t
e
f0 (y)
+
e 4t f0y (y) dy,
=
4t
4t R
=0
2
1
1
2
(xy)
4t
sup f0x(x) e
sup f0x(x) ez 4t dz
dy =
sup u(x, t)
x
4t x
4t x
R
R
= sup f0x(x).
ux (x, t) =
These inequalities are equalities when f0 (x) and f0x (x) are constants, respectively.
259
Problem (S01, #5). a) Show that the solution of the heat equation
< x <
ut = uxx ,
with squareintegrable initial data u(x, 0) = f (x), decays in time, and there is a constant
independent of f and t such that for all t > 0
1
3
2
f (x)2 dx .
max ux(x, t) t 4
x
f (y) dy = 3
(x y)e 4t f (y) dy.
4t
4t R
4t 2 R
ux (x, t)
=
=
1
3
2
(xy)
4t
f
(y)
y)e
dy
(x
(CauchySchwarz)
4t 2 R
1
(xy)2
1
2
(x y)2 e 2t dy f L2(R)
3
4t 2 R
12
3
1
2
z2
2e
(2t)
z
dz f L2(R)
3
4t 2 R
3
1
(2t) 4
2
2 z2
z
e
dz
f L2(R)
3
4t 2 R
dy
xy
z = , dz =
2t
2t
M <
34
= Ct
1
2
b) Note:
1
2
1
(xy)2
1
2
(xy)
4t
e
f (y) dy
e 2t dy f L2(R)
max u = max
x
x
4t R
4t R
1
dy
1
xy
ez2 2t dz 2 f L2(R)
z = , dz =
4t R
2t
2t
1
1
1
2
(2t) 4
2
ez dz f L2(R) = Ct 4 f L2(R).
=
1 1
2 2 t 2 R
=
65
64
65
1
1
uvdx ( u2 dx) 2 ( v 2 dx) 2
260
lim (x) = c.
x+
e 4t (y) dy.
u(x, t) =
4t R
We want to transform the equation to vt = vxx . Make a change of variables: x = ay.
u(x, t) = u(x(y), t) = u(ay, t) = v(y, t). Then,
vy = ux xy = aux ,
vyy
= auxx xy = a2 uxx ,
1
4t
(yz)2
4t
(az) dz.
x R.
Thus,
z2
dz
z
M
e 4t dz
s = , ds =
4t R
4t
4t
M
2
2
M
es 4t ds =
es ds = M.
R
4t R
2
1
es (y s 4t) 4t ds
=
4t
1
2
es (y s 4t) ds.
=
t+
=
=
261
0
1
2
e
lim (y s 4t) ds +
es lim (y s 4t) ds
t+
t+
0
0
1
1
1
1
2
2
+ b
es c ds +
es b ds = c
0
2
2
c+b
.
2
1
s2
Problem. Consider
ut = kuxx + Q,
0<x<1
u(0, t) = 0,
u(1, t) = 1.
What is the steady state temperature?
Proof. Set ut = 0, and integrate with respect to x twice:
kuxx + Q = 0,
Q
uxx = ,
k
Q
ux = x + a,
k
Q x2
u=
+ ax + b.
k 2
Boundary conditions give
Q
Q
x.
u(x) = x2 + 1 +
2k
2k
262
18.1
263
(18.1)
Show that such solutions, with initial data g L2 (Rn ), are unique, even when c is
negative.
Proof. McOwen. Consider v(x, t) = ect u(x, t).
in Rn (0, )
vt = v
v(x, 0) = g(x)
on Rn .
264
in R2 , t 0
u(x, y, 0) = 0
in
u(x, y, t) = 0
in , t 0.
u
1 d
2
2
ds u dx
u dx =
u
u4 dx,
2 dt
n
=0
1 d
2
2
u2 = u dx
u4 dx 0.
2 dt
Thus,
u(x, y, t)2 u(x, y, 0)2 = 0.
Hence, u(x, y, t)2 = 0, and u 0.
265
a) For a xed domain , show M is a constant in time if the boundary conditions are
u/n = 0.
b) Suppose that = (t) is evolving in time, with a boundary that moves at velocity
v, which may vary along the boundary. Find a modied boundary condition (in terms
of local quantities only) for u, so that M is constant.
Hint: You may use the fact that
d
f (x, t) dx =
ft (x, t) dx +
n v f (x, t) dl,
dt (t)
(t)
(t)
in which n is a unit normal vector to the boundary .
Proof. a) We have
ut u = 0,
on
u
on .
n = 0,
d
M (t) = 0. We have 66
We want to show that dt
d
d
u
M (t) =
ds = 0.
u(x, t) dx =
ut dx =
u dx =
dt
dt
b) We need
0 =
d
dt M (t)
= 0.
u(x, t) dx =
ut dx +
n v u ds
(t)
(t)
(t)
u
ds +
u dx +
n v u ds =
n v u ds
=
(t)
(t)
(t) n
(t)
u n ds +
n v u ds =
n (u + vu) ds.
=
d
M (t) =
dt
d
dt
(t)
(t)
Thus, we need:
n (u + vu) ds = 0,
66
on .
(t)
266
Problem (S95, #3). Write down an explicit formula for a function u(x, t) solving
in Rn (0, )
ut + b u + cu = u
(18.4)
u(x, 0) = f (x)
on Rn .
where b Rn and c R are constants.
Hint: First transform this to the heat equation by a linear change of the dependent
and independent variables. Then solve the heat equation using the fundamental solution.
Proof. Consider
u(x, t) = ex+tv(x, t).
ut = ex+t v + ex+tvt = (vt + v)ex+t,
u = ex+t v + ex+tv = (v + v)ex+t,
(u) = (v + v)ex+t = ( v + v)ex+t + (2 v + v)ex+t
= v + 2 v + 2 v)ex+t.
Plugging this into (18.4), we obtain
vt + v + b (v + v) + cv = v + 2 v + 2 v,
vt + b 2 v + + b + c 2 v = v.
In order to get homogeneous heat equation, we set
b
= ,
2
which gives
vt = v
b2
c,
4
in Rn (0, )
v(x, 0) = e 2 xf (x)
on Rn .
2
b
x( b4
2
+c)t
2
b
1
x( b4 +c)t
v(x, t) =
n e2
(4t) 2
Rn
xy2
4t
e 2 y f (y) dy.
267
(18.5)
for
x > 1,
c(x) = 1
for
x < 1.
u
dx < .
for some a, b, k, l.
Proof. Plug u(x, t) = etv(x) into (18.5) to get:
etv(x) = etv (x) + cetv(x),
v(x) = v (x) + cv(x),
v (x) v(x) + cv(x) = 0.
For x > 1, c = 0. We look for solutions of the form v(x) = aekx .
v (x) v(x) = 0,
ak2 ekx aekx = 0,
k2 = 0,
Thus, v(x) = c1 e
k2 = ,
k = .
x
+ c2 e
u(x, t) = aet e
x .
Since we want
u2 dx < :
For x < 1, c = 1. We look for solutions of the form v(x) = b cos lx.
v (x) v(x) + v(x) = 0,
bl 2 cos lx + (1 )b cos lx = 0,
l 2 + (1 ) = 0,
l 2 = 1 ,
l = 1 .
Thus, (since cos(x) = cos x)
u(x, t) = bet cos (1 )x.
We want v(x) to be continuous on R, and at x = 1, in particular. Thus,
ae = b cos (1 ),
a = be cos (1 ).
Also, v(x) is symmetric:
2
2
u dx = 2
u dx = 2
u dx +
1
u dx < .
268
h(X, T ) = (4T ) 2 e 4T
satises
ut + xu = uxx .
Given that U (x) is bounded and continuous everywhere on x , establish
that
U () h(x , t) d = U (x)
lim
t0
and show that u(x, t) U (x) as t 0. (You may use the fact that
1
2 .)
0
e d =
satises
ut + xu = uxx .
We have
ut =
=
xu =
ux =
=
uxx =
=
d 1 t3 xt
e3
K(x y t2 , t) g(y) dy
dt
1 3
1 3
(t2 x) e 3 t xt K + e 3 t xt KX (2t) + KT g(y) dy,
1 3
x e 3 t xt K(x y t2 , t) g(y) dy,
d 1 t3 xt
e3
K(x y t2 , t) g(y) dy
dx
1 3
1 3
t e 3 t xt K + e 3 t xt KX g(y) dy,
1 3
1 3
d
t e 3 t xt K + e 3 t xt KX g(y) dy
dx
1 3
1 3
1 3
1 3
t2 e 3 t xt K t e 3 t xt KX t e 3 t xt KX + e 3 t xt KXX g(y) dy.
269
Plugging these into , most of the terms cancel out. The remaining two terms cancel
because KT = KXX .
Given that g(x) is bounded and continuous on x , we establish that 67
lim
K(x y, t) g(y) dy = g(x).
Fix
t0
x0 R n ,
if y x0  < , y Rn .
Then if x x0  < 2 , we have: ( R K(x, t) dx = 1)
K(x
y,
t)
g(y)
dy
g(x
)
K(x
y,
t)
[g(y)
g(x
)]
dy
0
0
R
R
K(x y, t) g(y) g(x0 ) dy +
K(x y, t) g(y) g(x0 ) dy
B (x )
RB (x0 )
0
g(y) g(x0 ) <
K(xy,t) dy =
Furthermore, if x x0 
and y x0  , then
y x + y x0 .
2
2
1
Thus, y x 2 y x0 . Consequently,
K(x y, t) dy
= + 2gL
RB (x0 )
xy2
C
e 4t dy
+
t RB (x0 )
yx0 2
C
e 16t dy
+
t RB (x0 )
r2
C
e 16t r dr + 0
as t 0+ .
= +
t
y x0  y x +
Hence, if x x0  <
67
270
(18.6)
Making the change of variable (x, t) (u, t), where u = x/2 t, show that
4t
2T
T
T
=
.
+ 2u
t
u2
u
(18.7)
68
=
=
=
ut =
t = 1,
x
4t
3
2
1
ux = ,
2 t
x = 0.
uxx = 0,
T u T
+
,
u t
T u
,
u x
2 T u u T 2 u
T u
T
2 T
u 2
=
=
+
=
.
x x
x u x
u2 x x
u
x2
u2 x
=0
3 +
u
4t 2
T
4t
T
4t
68
=
=
=
=
=
2 T
u 2
,
u2 x
2 T
1 2
,
u2 2 t
1 2T
x T
,
+ 3
2
4t u
4t 2 u
2T
x T
,
+
u2
t u
2T
T
.
+ 2u
2
u
u
19
271
(19.1)
Problem (W02, #8). a) Find an explicit solution of the following Cauchy problem
2
2
u
xu2 = f (t, x),
t2
(19.2)
u(0, x) = 0, u
x (0, x) = 0.
b) Use part (a) to prove the uniqueness of the solution of the Cauchy problem
2
2
u
xu2 + q(t, x)u = 0,
t2
(19.3)
u(0, x) = 0, u
x (0, x) = 0.
Here f (t, x) and q(t, x) are continuous functions.
Proof. a) It was probably meant to give the ut initially. We rewrite (19.2) as
utt uxx = f (x, t),
u(x, 0) = 0, ut (x, 0) = 0.
(19.4)
2 0
for small t.
t2 qun un+1  un un+1  ,
Thus, T is a contraction a unique xed point.
Since T u = u, u is the solution to the PDE.
272
+ a(x, t)u = 0
t2
x2
in the square D, satisfying the boundary conditions
u1 = ,
u2 = ,
where 1 , 2 are two adjacent sides D. Here a(x, t), and are continuous functions.
Prove the uniqueness of the solution of this Goursat problem.
Proof. The change of variable = x + t, = x t
transforms the equation to
(, )
u = 0.
u
+ a
We integrate the equation:
u
(u, v) du dv =
a
(, ) u
du dv,
0 0
0 0
(0, v) dv =
a
(, ) u
du dv,
u
(, v) u
0
0
0
a(, ) u
du dv.
u
(, ) = u
(, 0) + u
(0, ) u(0, 0)
0
= + Kf,
= + K( + Kf ),
= +
K n ,
n=1
f = Kf
f = 0,
max f  max a max f .
0<x<
For small enough , the operator K is a contraction. Thus, there exists a unique xed
point of K, and f = Kf , where f is the unique solution.
20
273
(20.1)
is given by
u(x, t) =
where
K(x,
t) =
Rn
y, t) g(y) dy +
K(x
n e
(4t) 2
x2
4t
t
0
Rn
y, t s) f (y, s) dy ds
K(x
for t > 0,
for t 0.
x RN , 0 < t < T
u(x, 0) = 0.
Prove the uniqueness of the classical bounded solution assuming that T is small
enough.
Proof. Let {un } be a sequence of approximations to the solution, such that
t
2
=
K(x
y,
t
s)
f
(y,
s)
u
(y,
s)
dy ds.
S(un ) = un+1
n
use Duhamel s principle
Rn
We will show that S has a xed point S(un ) S(un+1 ) un un+1 , < 1
{un } converges to a uniques solution for small enough T .
Since un , un+1 C 2 (Rn ) C 1 (t) un+1 + un  M .
t
K(x y, t s) u2 u2 dy ds
S(un ) S(un+1 )
n+1
n
0 Rn
t
K(x y, t s) un+1 un un+1 + un dy ds
=
0 Rn
t
K(x y, t s) un+1 un dy ds
M
0
Rn
t
un+1 (x, s) un (x, s) ds
M M1
0
for small T.
1
Rn
274
Problem (S97, #3). a) Let Q(x) 0 such that x= Q(x) dx = 1,
and dene Q = 1 Q( x ). Show that (here denotes convolution)
Q(x) w(x)L w(x)L .
In particular, let Qt (x) denote the heat kernel (at time t), then
Qt(x) w1 (x) Qt (x) w2 (x)L w1 (x) w2 (x)L .
b) Consider the parabolic equation ut = uxx + u2 subject to initial conditions
u(x, 0) = f (x). Show that the solution of this equation satises
t
Qts (x) u2 (x, s) ds.
(20.2)
u(x, t) = Qt (x) f (x) +
0
c) Fix t > 0. Let {un (x, t)}, n = 1, 2, . . . the xed point iterations for the solution of
(20.2)
t
Qts (x) u2n (x, s) ds.
(20.3)
un+1 (x, t) = Qt (x) f (x) +
0
Let Kn (t) = sup0mn um (x, t)L . Using (a) and (b) show that
t
un (x, s) un1 (x, s)L ds.
un+1 (x, t) un (x, t)L 2 sup Kn ( )
0 t
Conclude that the xed point iterations in (20.3) converge if t is suciently small.
Proof. a) We have
Q(x) w(x)L
=
Q (x y)w(y) dy
Q (x y)w(y) dy
1
x y
Q
dy
Q (x y) dy = w
w
y
dy
1
y
Q
dy
z = , dz =
= w
Q(z) dz = w(x).
= w
275
x
1 e 4t
4t
(xy)2
1 (xy)2
4t
4t
=
e
w1 (y) dy
e
w2 (y) dy
4t
2
(xy)
1
4t
e
w1 (y) w2 (y) dy
4t
(xy)2
1
e 4t dy
w1 (y) w2 (y)
4t
1
2
xy
dy
ez 4t dz
z = , dz =
= w1 (y) w2 (y)
4t
4t
4t
1
2
=
w1 (y) w2 (y)
ez dz
Qt (x) =
= w1 (y) w2 (y) .
69
Note:
Qt (x) dx =
1
4t
(xy)2
4t
dy =
1
4t
ez
1
4t dz =
ez dz = 1.
276
b) Consider
ut = uxx + u2 ,
u(x, 0) = f (x).
We will show that the solution of this equation satises
t
Qts (x) u2 (x, s) ds.
u(x, t) = Qt (x) f (x) +
0
t
Qts (x) u2 (x, s) ds =
Qts (x y) u2 (y, s) dy ds
0
0
R
t
Qts (x y) us (y, s) uyy (y, s) dy ds
=
0
R
t
d
d
=
Qts (x y)u(y, s)
Qts (x y) u(y, s) Qts (x y)uyy (y, s) dy ds
ds
ds
0 R
Q0 (x y)u(y, t) dy
Qt(x y)u(y, 0) dy
=
R
R
t
d2
d
Qts (x y) u(y, s) + 2 Qts (x y)u(y, s) dy ds
dy
0
R ds
= u(x, t)
Note:
t0+
t0+
(a)
Also,
t
2
2
Qts (x) un (x, s) un1 (x, s) ds
0
t
2
Qts (x) u (x, s) u2 (x, s) ds
0
t
u2 (x, s) u2
n1
n1 (x, s) ds
un (x, s) un1 (x, s) un (x, s) + un1 (x, s) ds
0
t
un (x, s) un1 (x, s) ds
sup un (x, s) + un1 (x, s)
t
0 t
2 sup Kn ( )
0 t
t
0
un+1 (x, t) un (x, t)L 2t sup Kn ( ) un (x, s) un1 (x, s)L .
0 t
277
278
v(x, 0) = v0 (x).
If f and g are uniformly Lipschitz continuous, give a proof of existence and uniqueness of the solution to this problem in the space of bounded continuous functions with
u(, t) = supx u(x, t).
Proof. The space of continuous bounded functions forms a complete metric space so
the contraction mapping
principle
applies.
First, let v(x, t) = w x2 , t , then
ut = uxx + f (u, w)
wt = wxx + g(u, w).
These initial value problems have the following solutions (K is the heat kernel):
t
y, t s) f (u, w) dy ds,
K(x y, t) u0 (y) dy +
K(x
u(x, t) =
n
n
0
R
R
t
y, t) w0(y) dy +
y, t s) g(u, w) dy ds.
K(x
K(x
w(x, t) =
Rn
Rn
y, t s) f (u, w) dy ds,
K(x y, t) u0 (y) dy +
K(x
T1 u =
n
n
0
R
R
t
y, t) w0(y) dy +
y, t s) g(u, w) dy ds.
K(x
K(x
T2 w =
Rn
Rn
t
K(x
y, t s) f (un , w) f (un+1 , w) dy ds
0
Rn
t
K(x
y, t s) un un+1 dy ds
M1
0
Rn
t
y, t s)dy ds
sup un un+1
K(x
M1
0 x
Rn
t
sup un un+1 ds M1 t sup un un+1
M1
x
0 x
for small t.
< sup un un+1
y, t s) dy = 1.
We used the Lipshitz condition and R K(x
Thus, for small t, T1 is a contraction, and has a unique xed point. Thus, the solution
is dened as u = T1 u.
Similarly, T2 is a contraction and has a unique xed point. The solution is dened as
w = T2 w.
21
21.1
279
(21.1)
= {(x, t) U : x or t = 0}.
280
Problem (S98, #7). Prove that any smooth solution, u(x, y, t) in the unit box
= {(x, y)  1 x, y 1}, of the following equation
ut = uux + uuy + u,
t 0, (x, y)
(x, y)
[0,T ]
(x,y)
Then,
= v
v > 0
v
v > 0
[0,T ]
(x,y)
Now
max u =
[0,T ]
max (v + t) max v + T
[0,T ]
[0,T ]
(x,y)
(x,y)
21.2
281
uy = 0.
uxx 0,
uyy 0.
=0
=0
<0
uxx 0,
uyy 0.
=0
=0
>0
22
282
0 < x, y <
uxx + uyy = 0
u(0, y) = 0 = u(, y)
0y
u(x, 0) = 0,
u(x, ) = g(x)
0 x ,
where g is a continuous function satisfying g(0) = 0 = g().
Proof. Assume u(x, y) = X(x)Y (y), then substitution in the PDE gives X Y + XY =
0.
Y
X
=
= .
X
Y
From X + X = 0, we get Xn (x) = an cos nx + bn sin nx.
give
u(0, y) = X(0)Y (y) = 0
X(0) = 0 = X().
u(, y) = X()Y (y) = 0
Boundary conditions
a
n sin nx sinh ny,
n=1
which saties the equation and the three homogeneous boundary conditions. The
boundary condition at y = gives
u(x, ) = g(x) =
g(x) sin mx dx =
0
an sin nx sinh n,
n=1
n=1
an sinh n
sin nx sin mx dx =
0
am sinh m.
2
283
284
in
u = 0
0<y<
ux (0, y) = 0 = ux (, y)
u(x, 0) = 0,
u(x, ) = g(x)
0 < x < .
Proof. Assume u(x, y) = X(x)Y (y), then substitution in the PDE gives X Y + XY =
0.
X
Y
=
= .
X
Y
Consider X + X = 0.
If = 0, X0 (x) = a0 x + b0 .
If > 0, Xn (x) = an cos nx + bn sin nx.
Boundary conditions give
ux (0, y) = X (0)Y (y) = 0
X (0) = 0 = X ().
ux (, y) = X ()Y (y) = 0
Thus, X0 (0) = a0 = 0, and Xn (0) = nbn = 0.
X0 (x) = b0 , Xn (x) = an cos nx, n = 1, 2, . . ..
n an cos nx + an cos nx = 0,
n = n2 , n = 0, 1, 2, . . ..
With these values of n we solve Y n2 Y = 0.
If n = 0, Y0 (y) = c0 y + d0 .
If n = 0, Yn (y) = cn cosh ny + dn sinh ny.
Boundary conditions give
u(x, 0) = X(x)Y (0) = 0 Y (0) = 0.
Thus, Y0 (0) = d0 = 0, and Yn (0) = cn = 0.
Y0 (y) = c0 y, Yn (y) = dn sinh ny, n = 1, 2, . . . .
We have
0 y,
u0 (x, y) = X0 (x)Y0 (y) = b0 c0 y = a
n cos nx sinh ny.
un (x, y) = Xn (x)Yn (y) = (an cos nx)(dn sinh ny) = a
By superposition, we write
u(x, y) = a
0 y +
a
n cos nx sinh ny,
n=1
which saties the equation and the three homogeneous boundary conditions. The fourth
boundary condition gives
u(x, ) = g(x) = a
0 +
n=1
an cos nx sinh n,
&
g(x) dx = 0 a
an cos nx sinh n dx = a
0 2 ,
0 +
n=1
&
g(x) cos mx dx = n=1
an sinh n 0 cos nx cos mx dx =
1
a
0 = 2
g(x) dx,
0
0
0
2
a
n sinh n =
am sinh m.
285
286
Consider X + X = 0.
If = 0, X0 (x) = a0 x + b0 .
If n > 0, Yn (y) = cn e n y + dn e n y .
Since we look for bounded solutions as y , we have c0 = 0, dn = 0. Thus,
u(x, y) = a
0 +
n y
an cos
n x + bn sin n x .
n=1
( n + 1) a
n cos n x + bn sin n x .
n=1
f (x) C0 (R1 ), i.e. has compact support [L, L], for some L > 0. Thus the coecients
a
n , bn are given by
L
f (x) cos n x dx = ( n + 1)
an L.
L
L
f (x) sin
L
n x dx = ( n + 1)bnL.
70
Note that if we change the roles of X and Y in , the solution we get will be unbounded.
287
urr + 1r ur + r12 u = 0
u
r (1, ) = h()
r 2 urr + rur + u = 0.
Let r = et , u(r(t), ).
ut = ur rt = et ur ,
utt = (et ur )t = et ur + e2t urr = rur + r 2 urr .
Thus, we have
utt + u = 0.
Let u(t, ) = X(t)Y (), which gives X (t)Y () + X(t)Y () = 0.
Y ()
X (t)
=
= .
X(t)
Y ()
From Y () + Y () = 0, we get Yn () = an cos n + bn sin n.
n = n2 , n = 0, 1, 2, . . ..
With these values of n we solve X (t) n2 X(t) = 0.
X0 (r) = c0 log r + d0 .
If n = 0, X0 (t) = c0 t + d0 .
If n = 0, Xn (t) = cn ent + dn ent Xn (r) = cn r n + dn r n .
We have
u0 (r, ) = X0 (r)Y0 () = (c0 log r + d0 )a0 ,
un (r, ) = Xn (r)Yn () = (cn r n + dn r n )(an cos n + bn sin n).
But u must be nite at r = 0, so cn = 0, n = 0, 1, 2, . . ..
u0 (r, ) = d0 a0 ,
un (r, ) = dn r n (an cos n + bn sin n).
By superposition, we write
u(r, ) = a
0 +
r n (
an cos n + bn sin n).
n=1
n(
an cos n + bn sin n) = h().
n=1
The coecients an , bn for n 1 are determined from the Fourier series for h().
a0 is not determined by h() and therefore may take an arbitrary value. Moreover,
288
2
the constant term in the Fourier series for h() must be zero [i.e., 0 h()d = 0].
Therefore, the problem is not solvable for an arbitrary function h(), and when it is
solvable, the solution is not unique.
r n (
an cos n + bn sin n).
n=1
n=1
(
an cos n + bn sin n) = g().
289
1
g() d,
0
2
g() cos n d,
0
2
g() sin n d.
0
290
291
Hint:
x2 + y 2 < 1,
uxx + uyy = 0,
u
u = 0,
x2 + y 2 = 1.
n
= 1r r
(r r
) + r12
in polar coordinates.
2
u(r, 0) = u(r, 2)
u (r, 0) = u (r, 2)
Y (0) = Y (2),
r n (
an cos n + bn sin n).
n=1
(n 1)(
an cos n + bn sin n).
n=1
a
0 = 0.
n = n2 .
292
in r < 1
u/r = f ()
on r = 1,
beginning with the Fourier series for f (you may assume that f is continuously differentiable). Give your answer as a power series in x1 + ix2 plus a power series in
x1 ix2 . There is a necessary condition on f for this boundary value problem to be
solvable that you will nd in the course of doing this.
b) Sum the series in part (a) to get a representation of u in the form
2
N (r, )f ( ) d .
u(r, ) =
0
u (r, 0) = u (r, 2)
Y (0) = Y (2),
n = n2 .
293
r n (
an cos n + bn sin n).
n=1
Since
ur (r, ) =
nr n1 (
an cos n + bn sin n),
n=1
n (
an cos n + bn sin n) = f ().
n=1
an =
bn =
2
1
f () cos n d,
n 0
2
1
f () sin n d.
n 0
2
a
0 is not determined by f () (since 0 f () d = 0). Therefore, it may take an
arbitrary
2 value. Moreover, the constant term in the Fourier series for f () must be zero
[i.e., 0 f ()d = 0]. Therefore, the problem is not solvable for an arbitrary function
f (), and when it is solvable, the solution is not unique.
b) In part (a), we obtained the solution and the Fourier coecients:
2
1
f ( ) cos n d ,
an =
n 0
2
bn = 1
f ( ) sin n d .
n 0
u(r, ) = a
0 +
= a
0 +
= a
0 +
= a
0 +
= a
0 +
n=1
r n (
an cos n + bn sin n)
r
n=1
rn
n
n=1
rn
n
1
n
2
1 2
f ( ) cos n d cos n +
f ( ) sin n d sin n
n 0
f ( ) cos n cos n + sin n sin n d
n=1
2
f ( ) cos n( ) d
rn
cos n( ) f ( ) d .
n
n=1
N (r, )
294
for x2 + y 2 = 1.
x2 +y2
(x2 + y 2 )u(x, y) = 0?
u(r, 0) = u(r, 2)
Y (0) = Y (2),
u (r, 0) = u(r, 2)
r n (
an cos n + bn sin n).
n=1
n=1
(
an cos n + bn sin n).
n = n2 .
2
a0 = 0 f () d,
2
2
an = 0 f () cos n d,
2
bn = 0 f () sin n d.
2
1
0 = 2
f () d,
f0 = a
0
1 2
n = 0 f () cos n d,
fn = a
2
f () sin n d.
fn = bn = 1
x2 +y2
(x2 + y 2 )u(x, y) = 0,
or
= 0,
lim r 2 u(r, )
need
lim r 2 f0 +
n=1
r n (fn cos n + fn sin n)
= 0.
need
Since
lim
r 2n (fn cos n + fn sin n) = 0,
n>2
we need
2
2
r 2n (fn cos n + fn sin n)
= 0.
lim r f0 +
n=1
n = 0, 1, 2.
need
295
296
u(r, 0) = u(r, ) = 0,
1 < r < 2,
u(1, ) = sin ,
0 < < ,
u(2, ) = 0,
0 < < .
Hint: Use the formula =
1 2
r r (r r ) + r 2 2
With r = et , we have
utt + u = 0.
Let u(t, ) = X(t)Y (), which gives X (t)Y () + X(t)Y () = 0.
Y ()
X (t)
=
= .
X(t)
Y ()
= n n = n2 .
Thus, 0 = Yn (0) = an , and Yn () = bn sin = 0
Thus, Yn () = bn sin n, n = 1, 2, . . ..
With these values of n we solve X (t) n2 X(t) = 0.
X0 (r) = c0 log r + d0 .
If n = 0, X0 (t) = c0 t + d0 .
If n > 0, Xn (t) = cn ent + dn ent Xn (r) = cn r n + dn r n .
We have,
u(r, ) =
Xn (r)Yn () =
n=1
(
cn r n + dn r n ) sin n.
n=1
n=1
(
cn + dn ) sin n
c1 + d1 = 1,
cn + dn = 0, n = 2, 3, . . . .
(
cn 2n + dn 2n ) sin n
cn 2n + dn 2n = 0, n = 1, 2, . . ..
n=1
u(r, ) =
4
r
sin .
3r 3
297
u(1, ) = u(2, ) = 1,
0 < < .
With r = et , we have
utt + u = 0.
Let u(t, ) = X(t)Y (), which gives X (t)Y () + X(t)Y () = 0.
Y ()
X (t)
=
= .
X(t)
Y ()
From Y () + Y () = 0, we get Yn () = an cos n + bn sin n.
Boundary conditions give
un (r, 0) = 0 = Xn (r)Yn (0) = 0, Yn (0) = 0,
un (r, ) = 0 = Xn (r)Yn () = 0, Yn () = 0.
Thus, 0 = Yn (0) = an , and Yn () = bn sin n.
n = n2 , n = 1, 2, . . ..
With these values of n we solve X (t) n2 X(t) = 0.
X0 (r) = c0 log r + d0 .
If n = 0, X0 (t) = c0 t + d0 .
If n > 0, Xn (t) = cn ent + dn ent Xn (r) = cn r n + dn r n .
We have,
u(r, ) =
Xn (r)Yn () =
n=1
(
cn r n + dn r n ) sin n.
n=1
n=1
(
cn + dn ) sin n,
(
cn 2n + dn 2n ) sin n,
n=1
(
cn + dn ),
sin n d =
2
0
(
cn 2n + dn 2n ),
sin n d =
2
0
that can be solved.
298
299
u = 0
a < r < b, 0 < <
2
subject to the boundary conditions
u
(r, 0) = 0,
u
(a, ) = f1 (),
r
u
(r, ) = 0,
2
u
(b, ) = f2 (),
r
,
2
With r = et , we have
utt + u = 0.
Let u(t, ) = X(t)Y (), which gives X (t)Y () + X(t)Y () = 0.
Y ()
X (t)
=
= .
X(t)
Y ()
Yn () = an n sin n + bn n cos n . Thus, Yn (0) = bn n = 0
n 2 = n n = (2n)2 .
Yn ( 2 ) = an n sin n 2 = 0
Thus, Yn () = an cos(2n), n = 0, 1, 2, . . ..
In particular, Y0 () = a0 t + b0 . Boundary conditions give Y0 () = b0 .
With these values of n we solve X (t) (2n)2 X(t) = 0.
X0 (r) = c0 log r + d0 .
If n = 0, X0 (t) = c0 t + d0 .
2nt
2nt
Xn (r) = cn r 2n + dn r 2n .
If n > 0, Xn (t) = cn e + dn e
u(r, ) = c0 log r + d0 +
(
cn r 2n + dn r 2n ) cos(2n).
n=1
ur (r, ) =
c0
+
(2n
cn r 2n1 + 2ndn r 2n1 ) cos(2n).
r
n=1
bn = 0.
300
c0
+2
n(
cn a2n1 + dn a2n1 ) cos(2n),
a
f1 () = ur (a, ) =
n=1
c0
+2
b
f2 () = ur (b, ) =
n(
cn b2n1 + dn b2n1 ) cos(2n).
n=1
n(
cn a2n1 + dn a2n1 ),
f1 () cos(2n) d =
2
0
2
n(
cn b2n1 + dn b2n1 ).
f2 () sin(2n) d =
2
0
=
u dx =
=
=
u
n
0
u
u
(b, ) d +
(a, ) d +
r
r
2
2
f2 () d +
f1 () d + 0 + 0
0
f1 () d +
b
a
u
(r, 0) dr +
b
u
r,
dr
f2 () d.
Proof. No, since the boundary conditions are Neumann. The solution is unique only
up to a constant.
Problem (S99, #4). Let u(x, y) be harmonic inside the unit disc,
with boundary values along the unit circle
1, y > 0
u(x, y) =
0, y 0.
Compute u(0, 0) and u(0, y).
Proof. Since u is harmonic, u = 0. Use polar coordinates (r, )
1
+ r12 u = 0
0 r < 1, 0 < 2
urr + r ur
1,
0<<
u(1, ) =
0,
2.
r 2 urr + rur + u = 0.
With r = et , we have
utt + u = 0.
Let u(t, ) = X(t)Y (), which gives X (t)Y () + X(t)Y () = 0.
Y ()
X (t)
=
= .
X(t)
Y ()
From Y () + Y () = 0, we get Yn () = an cos n + bn sin n.
n = n2 , n = 1, 2, . . ..
With these values of n we solve X (t) n2 X(t) = 0.
X0 (r) = c0 log r + d0 .
If n = 0, X0 (t) = c0 t + d0 .
If n > 0, Xn (t) = cn ent + dn ent Xn (r) = cn r n + dn r n .
We have
u0 (r, ) = X0 (r)Y0 () = (c0 log r + d0 )a0 ,
un (r, ) = Xn (r)Yn () = (cn r n + dn r n )(an cos n + bn sin n).
But u must be nite at r = 0, so cn = 0, n = 0, 1, 2, . . ..
0 ,
u0 (r, ) = a
an cos n + bn sin n).
un (r, ) = r n (
By superposition, we write
u(r, ) = a
0 +
r n (
an cos n + bn sin n).
n=1
n=1
(
an cos n + bn sin n) =
1,
0,
0<<
2,
and the coecients an and bn are determined from the above equation.
71
71
301
23
302
u(r, 0) = 0
0 r < 1,
0 r < 1,
u (r, 2 ) = 0
u(1, ) = 0
0 2 .
0 r < 1, 0 <
We solve
r 2 urr + rur + u = 0.
Let r = et , u(r(t), ), we have
utt + u = 0.
u(r, 0) = X(r)Y (0) = 0
Y
(0)
=
Y
= 0.
2
u (r, 2 ) = X(r)Y ( 2 ) = 0
(0)
=
a
=
0,
and
Y
(
)
=
b
cos
n 2 = 0
Thus,
Y
n
n
n
n
n
2
n 2 = n 2 , n = 1, 2, . . . n = (2n 1)2 .
n=1
Boundary
303
We now plug this equation into with inhomogeneous term and obtain
= f (t, ),
n=1
Xn (t) (2n 1)2 Xn (t) sin[(2n 1)] = f (t, ),
n=1
Xn (t) (2n 1)2 Xn (t) =
4
Xn(t)
(2n 1) Xn (t) =
0
(2n1)
+ dn r
(2n1)
or
+ unp(r),
dn r (2n1) + unp(r) sin[(2n 1)].
n=1
dn + unp(1) sin[(2n 1)],
n=1
0=
u(r, ) =
unp (1)r (2n1) + unp (r) sin[(2n 1)].
n=1
Note the similar problem on 2D Poisson equation on a square domain. The problem is used by rst nding the eigenvalues and eigenfunctions of the Laplacian, and
then expanding f (x, y) in eigenfunctions, and comparing coecients of f with the general solution u(x, y).
Here, however, this could not be done because of the circular geometry of the domain.
In particular, the boundary conditions do not give enough information to nd explicit
representations for m and n . Also, the condition u = 0 for x > 0, y > 0, x2 +y 2 = 1
72
ChiuYens solutions have attempts to solve this problem using Greens function.
304
24
305
Example (McOwen 3.1 #2). We considered the initial/boundary value problem and
solved it using Fourier Series. We now solve it using the Separation of Variables.
utt uxx = 0
(24.1)
0<x<
u(x, 0) = 1,
ut (x, 0) = 0
u(0, t) = 0,
u(, t) = 0
t 0.
Proof. Assume u(x, t) = X(x)T (t), then substitution in the PDE gives XT X T = 0.
T
X
=
= .
X
T
From X + X = 0, we get Xn (x) = an cos nx + bn sin nx.
give
u(0, t) = X(0)T (t) = 0
X(0) = X() = 0.
u(, t) = X()T (t) = 0
Boundary conditions
n=1
cn sin nt + dn cos nt sin nx,
n
cn cos nt ndn sin nt sin nx.
n=1
n=1
dn sin nx,
n
cn sin nx.
n=1
sin mx dx = dm ,
2
0
0 dx = n
cn ,
2
0
which gives the coecients
2
(1 cos n) =
dn =
n
4
n ,
0,
n odd,
n even,
and
cn = 0.
306
0<x<1
Thus, Xn (0) =
an = 0, and Xn (x)
= bn sin n x.
Xn (1) = bn sin n = 0. Hence, n = n, or n = (n)2 , n = 1, 2, . . ..
n = (n)2,
32 t
Tn (t) = e
5
5
cn cos (n)2 t + dn sin (n)2 t .
4
4
n=1
32 t
5
5
cn cos (n)2 t + dn sin (n)2 t sin nx.
4
4
cn sin nx.
n=1
Thus,
307
5
(n)2 t sin nx.
4
n=1
3
5
5
5
32 t
32 t
2
2
2
dn e
cos (n) t sin nx,
(n)
sin (n) t + dn e
ut (x, t) =
2
4
4
4
n=1
5
sin nx.
dn
(n)2
x sin(2x) = ut (x, 0) =
4
u(x, t) =
32 t
dn e
sin
n=1
,
x sin(2x) sin(mx) dx = dm
(m)2
2
4
0
1
2
dn = '
x sin(2x) sin(nx) dx.
5
2
0
(n) 4
32 t
u(x, t) = e
n=1
dn sin
5
(n)2 t sin nx.
4
Problem (F04, #1). Solve the following initialboundary value problem for the wave
equation with a potential term,
0 < x < ,
u(x, 0) = f (x), ut (x, 0) = 0
where
f (x) =
x
x
if x (0, /2),
if x (/2, ).
The answer should be given in terms of an innite series of explicitly given functions.
Proof. Assume u(x, t) = X(x)T (t), then substitution in the PDE gives
XT X T + XT = 0,
X
T
+1 =
= .
T
X
308
cn cos
1 + n2 t + dn sin
1 + n2 t sin nx.
n=1
ut (x, t) =
cn ( 1 + n2 ) sin 1 + n2 t + dn ( 1 + n2 ) cos 1 + n2 t sin nx.
n=1
cn sin nx.
n=1
dn ( 1 + n2 ) sin nx.
n=1
f (x) sin mx dx = cm ,
2
0
0 dx = dm
1 + m2 ,
2
0
which gives the coecients
2
2 2
2
f (x) sin nx dx =
x sin nx dx +
( x) sin nx dx
cn =
0
0
2
1
1 2
2
1
1
2
2
x cos nx +
cos nx + x cos nx
cos nx dx +
cos nx dx
=
n
0
n 0
n
n
n
2
2
2
n
1
1
n
2
cos
+ 2 sin
2 sin 0
=
2n
2
n
2
n
n
1
1
n
2
cos n + cos
+ cos n
cos
2 sin n + 2 sin
+
n
n
2
n
2n
2
n
n
2
2 1
4
n
n
n
2 1
+
=
sin
sin
sin
=
n2
2
n2
2
n2
2
n = 2k
0,
0,
n = 2k
4
=
=
n = 4m + 1
n1
2,
4
n
(1) 2 n2 ,
n = 2k + 1.
4
n = 4m + 3
n2 ,
dn = 0.
u(x, t) =
n=1
cn cos
1 + n2 t sin nx.
25
309
ut = uxx
2
0x2
u(x, 0) = x x + 1
u(0, t) = 1, u(2, t) = 3
t > 0.
Find limt+ u(x, t).
Proof. First, we need to obtain function v that satises vt = vxx and takes 0
boundary conditions. Let
v(x, t) = u(x, t) + (ax + b),
(25.1)
b = 1,
v(2, t) = 0 = u(2, t) + 2a 1 = 2a + 2
a = 1.
(25.2)
vt = vxx ,
v(x, 0) = (x2 x + 1) x 1 = x2 2x,
v(0, t) = v(2, t) = 0.
We solve the problem for v using the method of separation of variables.
Let v(x, t) = X(x)T (t), which gives XT X T = 0.
T
X
=
= .
X
T
an = 0, and Xn (x)
Hence, Xn (0) =
= bn sin x.
2
Xn (2) = bn sin 2 = 0 2 = n n = ( n
2 ) .
Xn (x) = bn sin
nx
,
2
n =
n 2
2
v(x, t) =
n 2
) t
2
n 2
2
T = 0 to nd
Xn (x)Tn (t) =
n=1
cn e(
n=1
n 2
) t
2
sin
nx
.
2
cn sin
n=1
cn =
2
0
nx
= x2 2x.
2
nx
dx =
(x2 2x) sin
2
v(x, t) =
n=2k1
0
n is even,
32
(n)3 n is odd.
32 ( n )2 t
nx
.
e 2
sin
3
(n)
2
n=2k1
32 ( n )2 t
nx
+ x + 1.
e 2
sin
(n)3
2
lim u(x, t) = x + 1.
t+
310
311
ut = uxx
u(x, 0) = f (x)
0xL
ux (0, t) = ux (L, t) = A
t>0
(A = Const).
Find v(x)  the limit of u(x, t) when t . Show that v(x) is one of the ininitely
many solutions of the stationary problem
vxx = 0
0<x<L
vx (0) = vx (L) = A.
Proof. First, we need to obtain function v that satises vt = vxx and takes 0
boundary conditions. Let
v(x, t) = u(x, t) + (ax + b),
(25.3)
a = A,
vx (L, t) = 0 = ux (L, t) + a = A + a
a = A.
We may set b = 0 (innitely many solutions are possible, one for each b).
Thus, (25.3) becomes
v(x, t) = u(x, t) Ax.
The new problem is
vt = vxx ,
v(x, 0) = f (x) Ax,
vx (0, t) = vx(L, t) = 0.
We solve the problem for v using the method of separation of variables.
Let v(x, t) = X(x)T (t), which gives XT X T = 0.
T
X
=
= .
X
T
(25.4)
312
x + bn cos x.
Xn (x) = an sin
Xn (x) = an cos
n 2
n =
L
n 2
v(x, t) =
n 2
) t
L
Tn (t) = cn e(
T = 0 to nd
, n = 1, 2, . . ..
Xn (x)Tn (t) = c0 +
n=1
cn e(
n 2
) t
L
cos
n=1
nx
.
L
cn cos
n=1
L
c0 =
L
cn =
2
nx
= f (x) Ax.
L
(f (x) Ax) dx =
nx
dx
(f (x) Ax) cos
L
1
v(x, t) =
L
0
1 L
AL
c0 =
,
f (x) dx
L 0
2
1 L
nx
cn =
dx.
(f (x) Ax) cos
L 0
L
AL2
f (x) dx
2
f (x) dx
n 2
AL
nx
+
.
cn e( L ) t cos
2
L
n
L
0
f (x) dx
lim u(x, t) = Ax + b,
t+
n 2
AL
nx
+
+ Ax.
cn e( L ) t cos
2
L
n
b arbitrary.
To show that v(x) is one of the ininitely many solutions of the stationary problem
vxx = 0
0<x<L
vx (0) = vx (L) = A,
we can solve the boundary value problem to obtain v(x, t) = Ax+b, where b is arbitrary.
313
x , t 0
ut = u,
u(x, 0) = f (x),
x
u(x, t) = g(x),
x , t > 0.
Proof. Let w(x) be the solution of the Dirichlet problem:
w = 0,
x
w(x) = g(x),
x
and let v(x, t) be the solution of the IBVP for the heat equation with homogeneous
BC:
x , t 0
vt = v,
v(x, 0) = f (x) w(x),
x
v(x, t) = 0,
x , t > 0.
Then u(x, t) satises
u(x, t) = v(x, t) + w(x).
lim u(x, t) = w(x).
314
x , t 0
ut = u + F (x, t),
u(x, 0) = f (x),
x
u(x, t) = g(x),
x , t > 0.
Proof. We rst nd u1 , the solution to the homogeneous heat equation (no F (x, t)).
Let w(x) be the solution of the Dirichlet problem:
w = 0,
x
w(x) = g(x),
x
and let v(x, t) be the solution of the IBVP for the heat equation with homogeneous
BC:
x , t 0
vt = v,
v(x, 0) = f (x) w(x),
x
v(x, t) = 0,
x , t > 0.
Then u1 (x, t) satises
u1 (x, t) = v(x, t) + w(x).
lim u1 (x, t) = w(x).
Rn
Rn
y, t s) F (y, s) dy ds.
K(x
t 0, 0 < x < 1,
ut = uxx ,
u(x, 0) = 0,
0 < x < 1,
t
u(0, t) = 1 e , ux (1, t) = et 1,
315
t > 0.
(25.6)
(25.7)
sin 11
cos 1 ,
sin 1 1
sin x)et .
cos 1
(25.8)
vt = vxx ,
v(x, 0) = (x 1) + (cos x +
v(0, t) = 0,
vx (1, t) = 0.
sin 1 1
sin 1 1
sin x) = (x 1) + (cos x +
sin x).
cos 1
cos 1
sin 11
cos 1
sin x),
316
bn cos x,
Xn (x) =
Xn (1) =
bn cos = 0,
cos = 0,
= n + .
2
Thus,
x,
Xn (x) = bn sin n +
2
2
n = n +
.
2
2
With these values of n , we solve T + n + 2 T = 0 to nd
2
Tn (t) = cn e(n+ 2 ) t .
v(x, t) =
Xn (x)Tn (t) =
n=1
n=1
bn sin n + x e(n+ 2 )2 t.
2
sin 1 1
sin x)et .
cos 1
bn sin n + x (x 1) (cos x + sin 1 1 sin x).
u(x, 0) =
2
cos 1
n=1
Finally, we can check that the dierential equation and the boundary conditions are
satised:
u(0, t) = 1 (1 + 0)et = 1 et .
bn n + cos n + x e(n+ 2 )2 t 1 + (sin x sin 1 1 cos x)et ,
ux (x, t) =
2
2
cos 1
n=1
sin 1 1
cos 1)et = 1 + et .
cos 1
(n+ )2 t
sin 1 1
2
2
xe
sin x)et = uxx.
=
bn n +
sin n +
+ (cos x +
2
2
cos 1
ux (1, t) = 1 + (sin 1
ut
n=1
317
Problem (F02, #6). The temperature of a rod insulated at the ends with an exponentially decreasing heat source in it is a solution of the following boundary value
problem:
2t
u(x, 0) = f (x).
Find the solution to this problem by writing u as a cosine series,
u(x, t) =
n=0
bn cos nx
with
bn = 2
n=1
n=1
n=1
bn cos nx,
n=1
which gives
a0 (t) = b0 e2t ,
an (t) + n2 2 an (t) = bn e2t ,
n = 1, 2, . . . .
Adding homogeneous and particular solutions of the above ODEs, we obtain the solutions
a0 (t) = c0 b20 e2t ,
2 2
bn
2t
, n = 1, 2, . . . ,
an (t) = cn en t 2n
2 2 e
for some constants cn , n = 0, 1, 2, . . .. Thus,
u(x, t) =
2 2
cn en t
n=0
bn
2t
e
cos nx.
2 n2 2
cn
u(x, 0) =
n=0
bn
cos nx = f (x),
2 n2 2
b0
2
as t .
318
x
s .
ut = uxx + cos x
ux (0, t) = ux (2, t) = 0
x [0, 2]
t>0
x [0, 2].
ux (2) = 0
b) With
v(x, t) = u(x, t) cos x
the problem above transforms to
vt = vxx
vx (0, t) = vx(2, t) = 0
get
X
(x)
=
a
cos
x
+
b
sin
x.
From X +X = 0, we
n
n
n
n = n2 n = ( n2 )2 . Thus,
Xn (2) = n an sin n 2 = 0
Xn (x) = an cos
nx
,
2
n =
n 2
2
n 2
2
319
T = 0 to nd
n 2
Tn (t) = cn e( 2 ) t .
v(x, t) =
Xn (x)Tn (t) =
n=0
n 2
t
a
n e( 2 )
n=0
cos
nx
.
2
an cos
n=0
nx
= cos 2x.
2
n = 0, n = 4. Hence,
Thus, a
4 = 1, a
v(x, t) = e4t cos 2x.
u(x, t) = v(x, t) + cos x = e4t cos 2x + cos x.
ux (2) = 0
x [0, 2]
uxx = cos x,
ux = sin x + C,
u(x) = cos x + Cx + D.
Boundary conditions give:
1 = ux (0) = C,
0 = ux (2) = C
contradiction
n
2
cos x dx = ux (2) ux (0) =
1 .
0
given
=0
We
320
u(0, t) = u(, t) = 0,
v(x, 0) = sin x,
v(0, t) = v(, t) = 0.
x=0
0 x , t > 0
vt = 2vxx,
v(x, 0) = sin x,
v(0, t) = v(, t) = 0.
Assume v(x, t) = X(x)T (t), then substitution in the PDE gives XT = 2X T .
X
T
=2
= .
T
X
X
'
2X
2x
+
= 0, we get Xn (x) = an cos
+ bn sin
From
Boundary conditions give
v(0, t) = X(0)T (t) = 0
X(0) = X() = 0.
v(, t) = X()T (t) = 0
'
Thus, Xn (0) = an = 0, and Xn (x) = bn sin 2 x.
'
'
Xn () = bn sin 2 = 0. Hence 2 = n, or = 2n2 .
= 2n2 ,
'
2 x.
2t
a
n e2n
sin nx.
n=1
a
n sin nx = sin x.
n=1
n = 0, n = 2, 3, . . ..
Thus, a
1 = 1, a
v(x, t) = e2t sin x.
73
Note that if the matrix was fully inseparable, we would have to nd eigenvalues and eigenvectors,
just as we did for the hyperbolic systems.
321
We have
1 2t
sin x,
0 x , t > 0
ut = uxx 2 e
u(x, 0) = sin x,
u(0, t) = u(, t) = 0.
&
Let u(x, t) =
n=1 un (t) sin nx. Plugging this into the equation, we get
n=1
n=1
1
n2 un (t) sin nx = e2t sin x.
2
For n = 1:
1
u1 (t) + u1 (t) = e2t .
2
Combining homogeneous and particular solution of the above equation, we obtain:
1
u1 (t) = e2t + c1 et .
2
For n = 2, 3, . . .:
un (t) + n2 un (t) = 0,
2
un (t) = cn en t .
Thus,
u(x, t) =
1
2
2
e2t + c1 et sin x +
cn en t sin nx = e2t sin x +
cn en t sin nx.
2
2
n=2
n=1
u(x, 0) =
1
sin x +
cn sin nx = sin x.
2
n=1
Thus, c1 = 12 , cn = 0, n = 2, 3, . . ..
u(x, t) =
1
sin x (e2t + et ).
2
x=0
n=1
an sin nx,
v(x, 0) =
n=1
bn sin nx.
322
u(x, t) =
1 2t
2
e sin x +
cn en t sin nx,
2
n=1
v(x, t) =
2t
bn e2n
sin nx.
n=1
1 2t
2
2
2
n2 t
e sin x +
[u (x, t) + v (x, t)] dx =
cn e
sin nx +
bn e2n t sin nx dx
x=0
x=0 2
n=1
n=1
(b2n + a2n )
sin2 nx dx
[u2 (x, 0) + v 2 (x, 0)] dx.
n=1
x=0
x=0
26
323
in
uxx + uyy + u = 0
(26.1)
u(0, y) = 0 = u(a, y)
for 0 y b,
u(x, 0) = 0 = u(x, b)
for 0 x a.
Proof. We can solve this problem by separation of variables.
Let u(x, y) = X(x)Y (y), then substitution in the PDE gives X Y + XY + XY = 0.
X Y
+
+ = 0.
X
Y
Letting = 2 + 2 and using boundary conditions, we nd the equations for X and
Y:
X + 2 X = 0
Y + 2 Y = 0
X(0) = X(a) = 0
Y (0) = Y (b) = 0.
m2
a2
n2
b2
ny
mx
sin
,
a
b
where m, n = 1, 2, . . . .
Observe that the eigenvalues {mn }
m,n=1 are positive. The smallest eigenvalue 11
has only one eigenfunction u11 (x, y) = sin(x/a) sin(y/b); notice that u11 is positive
in . Other eigenvalues may correspond to more than one choice of m and n; for
example, in the case a = b we have nm = nm . For this , there are two linearly
independent eigenfunctions. However, for a particular value of there are at most
nitely many linearly independent eigenfunctions. Moreover,
b a
b a
ny
m x
n y
mx
sin
sin
sin
dx dy
umn (x, y) umn (x, y) dx dy =
sin
a
b
a
b
0
0
0
0
ny
n y
a b
ab
sin
sin
dy
if m = m and n = n
2 0
b
b
4
=
=
0 if m = m or n = n .
0
orthogonal. We could normalize each umn by a
In particular, the {umn } are pairwise
scalar multiple (i.e. multiply by 4/ab) so that ab/4 above becomes 1.
Let us change the notation somewhat so that each eigenvalue n corresponds to a
particular eigenfunction n (x). If we choose an orthonormal basis of eigenfunctions in
each eigenspace, we may arrange that {n }
n=1 is pairwise orthonormal:
1 if m = n
n (x)m(x) dx =
0 if m = n.
n=1
an n (x),
where
an =
324
325
in D
u=0
on D.
in
uxx + uyy + u = 0
u(0, y) = 0 = u(a, y)
for 0 y b,
u(x, 0) = 0 = u(x, b)
for 0 x a.
We may assume that the eigenvalues are positive, = 2 + 2 . Then,
mn = 2
m2
a2
n2
b2
ny
mx
sin
,
a
b
m, n = 1, 2, . . ..
(26.2)
in the strip {(x, y), 0 < y < , < x < +} with boundary conditions
u(x, 0) = 0, u(x, ) = 0.
(26.3)
Find all bounded solutions of the boundary value problem (26.4), (26.5) when
a) = 0,
b) > 0,
c) < 0.
Proof. a) = 0. We have
uxx + uyy = 0.
Assume u(x, y) = X(x)Y (y), then substitution in the PDE gives
X Y + XY = 0.
Boundary conditions give
u(x, 0) = X(x)Y (0) = 0
u(x, ) = X(x)Y () = 0
Y (0) = Y () = 0.
Method I: We have
Y
X
=
= c,
c > 0.
X
Y
cx +
bn sin cx.
From X + cX = 0, we have Xn (x) = an cos
Y () = cn e c cn e c = 0
u(x, y) = X(x)Y (y) = 0.
Method II: We have
X
Y
=
= c,
X
Y
Yn (y) = 0.
cn = 0
c > 0.
c = n c = n2 .
Y () = dn sin c = 0
Yn (y) = dn sin nx = 0.
u(x, y) = X(x)Y (y) = 0.
Xn (x) = 0.
b) > 0. We have
X Y
+
+ = 0.
X
Y
Letting = 2 + 2 , and using boundary conditions for Y , we nd the equations:
X + 2 X = 0
Y + 2 Y = 0
Y (0) = Y () = 0.
326
umn (x, y) =
m,n=1
m,n=1
c) < 0. We have
uxx + uyy + u = 0,
u(x, 0) = 0, u(x, ) = 0.
u 0 is the solution to this equation. We will show that this solution is unique.
Let u1 and u2 be two solutions, and consider w = u1 u2 . Then,
w + w = 0,
w(x, 0) = 0, w(x, ) = 0.
Multiply the equation by w and integrate:
ww + w 2 = 0,
ww dx + w 2 dx = 0,
w
ds w2 dx + w 2 dx = 0,
w
n
=0
2
w dx = w 2 dx .
327
328
Y + 2 Y = 0.
X(0) = X (a) = 0.
The solutions of these onedimensional eigenvalue problems are
(m 12 )
,
a
(m 12 )x
Yn (y) = cn cos n y + dn sin n y,
Xm (x) = sin
a
where m, n = 1, 2, . . . . Thus we obtain solutions of the form
m =
2
kmn
=
(m 12 )
a
2
(m 12 )x
cn cos n y+dn sin n y ,
umn (x, y) = sin
a
+n2 ,
where m, n = 1, 2, . . . .
u(x, y) =
umn (x, y) =
m,n=1
sin
m,n=1
(m 12 )x
cn cos n y + dn sin n y .
a
We can take an alternate approach and prove the second part of the question. We
have
X Y + XY + k2 XY = 0,
X
Y
=
+ k 2 = c2 .
Y
X
We obtain Yn (y) = cn cos cy + dn sin cy.
X + k2 X = c2 X,
X + (k2 c2 )X = 0,
Xm (x) = am e
c2 k2 x
+ bm e
c2 k2 x
329
m = 1, 2, . . .,
k2 c2 a = m ,
2
1 2
m
,
k 2 c2 =
a
2
1 2
+ c2 ,
m
k2 =
a
2
1 2
a2 k2 > 2 m
,
2
1
,
m = 1, 2, . . ..
ak > m
2
Thus, bounded solutions exist only when ak > 2 .
(26.4)
(26.5)
Proof. We have
uxx + uyy + k2 u = 0,
X Y + XY + k2 XY = 0,
Y
X
=
+ k 2 = c2 .
X
Y
We obtain Xm (x) = am cos cx + bm sin cx.
Y + k2 Y = c2 Y,
Y + (k2 c2 )Y = 0,
c2 k2 y
Yn (y) = cn e
+ dn e
c2 k2 y
k 2 c2 = n
Yn () = dn sin k2 c2 = 0
2
2
2
k = n + c , n = 1, 2, . . .. Hence, k > n, n = 1, 2, . . ..
k2 c2 = n2
m,n=1
Xm (x)Yn (y) =
m,n=1
sin ny Xm (x).
330
331
McOwen, 4.4 #7; 266B Ralston Hw. Show that the boundary value problem
a(x)u + b(x)u = u
in
u=0
on
has only trivial solution with 0, when b(x) 0 and a(x) > 0 in .
Proof. Multiplying the equation by u and integrating over , we get
2
u au dx +
bu dx =
u2 dx.
(26.6)
=0
2
2
a u dx +
b u dx =
u2 dx,
>0
=0
2
2
a
u
dx
+
b
u
dx
=
u2 dx.
>0
u2 dx,
which implies = 0. This gives the useful information that for the eigenvalue problem74
a(x)u + b(x)u = u
u/n = 0,
= 0 is an eigenvalue, its eigenspace is the set of constants, and all other s are
positive.
74
In Ralstons Hw#7 solutions, there is no  sign in front of a(x)u below, which is probably a
typo.
332
Similar Problem II: If 0, we show that the only solution to the problem below
is the trivial solution.
u + u = 0
in
u=0
on
u2 dx = 0,
uu dx +
u
u
ds
n
=0
u dx +
u2 dx = 0.
27
333
in
u=0
on ,
an n (x),
where
an =
n=1
&
Proof. Writing u =
n cn n =
n=1
an n (x).
n=1
Thus, cn = an /n , and
u(x) =
an n (x)
n=1
u(L) = 0,
(L) = 0.
The eigenvalues are n = (n/L)2 , and the corresponding eigenfunctions are sin(nx/L),
n = 1, 2, . . .. &
&
Writing u =
cn n = cn sin(nx/L) and inserting into u = f , we get
0
n=1
L
n=1
cn
n 2
L
cn
n 2
L
sin
nx
L
= f (x),
L
mx
nx
mx
sin
dx =
dx,
sin
f (x) sin
L
L
L
0
L
n 2 L
nx
=
dx,
f (x) sin
cn
L
2
L
0
L
2 0 f (x) sin(nx/L) dx
.
cn =
L
(n/L)2
u(x) =
u=
cn sin(nx/L) =
L
n=1
L
0
334
f () sin(nx/L) sin(n/L) d
,
(n/L)2
2 sin(nx/L) sin(n/L)
f ()
d.
L
(n/L)2
n=1
= G(x,)
See similar, but more complicated, problem in SturmLiouville Problems (S92, #2(c)).
335
an (x)n (y),
where
an (x) =
G(x, z)n(z) dz.
G(x, y)
n=1
m=1
&
cm m (x) +
an (x)n (y)
cm m (y) dy = 0,
Suppose, u(x) =
n=1
cm m (x) +
m=1
n=1
an (x)
cm
m=1
m=1
n (y)m(y) dy = 0,
cn n (x) +
n=1
an (x)cn = 0,
n=1
cn n (x) + an (x) = 0,
n=1
an (x) =
Thus,
G(x, y)
n=1
75
n (x)n (y)
.
n
n (x)
.
n
75
336
u(x, 0) = 0 = u(x, b)
for 0 x a.
f (x, y) C 2 , f (x, y) = 0 if x = 0, x = a, y = 0, y = b,
ny
mx
2
sin
.
cmn sin
f (x, y) =
a
b
ab m,n=1
in
uxx + uyy + u = 0
u(0, y) = 0 = u(a, y)
for 0 y b,
u(x, 0) = 0 = u(x, b)
for 0 x a.
Let u(x, y) = X(x)Y (y), then substitution in the PDE gives X Y + XY + XY = 0.
X Y
+
+ = 0.
X
Y
Letting = 2 + 2 and using boundary conditions, we nd the equations for X and
Y:
X + 2 X = 0
Y + 2 Y = 0
X(0) = X(a) = 0
Y (0) = Y (b) = 0.
m2
a2
n2
b2
mx
ny
2
sin
,
mn (x, y) = sin
a
b
ab
cmn mn .
m,n=1
&
mn cmn mn (x, y) =
m,n=1
m,n=1
.
Thus, cmn = cmn
mn
u(x, y) =
cmn
mn (x, y),
mn
n=1
f (x, y)mn dx dy =
cm n m n mn dx dy = cmn .
0
0 m ,n =1
337
28
338
In the section on the wave equation, we considered an initial boundary value problem
for the onedimensional wave equation on an interval, and we found that the solution could be obtained using Fourier series. If we replace the Fourier series by an
expansion in eigenfunctions, we can consider an initial/boundary value problem for the
ndimensional wave equation.
The ND WAVE Equation (eigenvalues/eigenfunctions of the Laplacian).
Consider
for x , t > 0
utt = u
u(x, 0) = g(x), ut (x, 0) = h(x)
for x
u(x, t) = 0
for x , t > 0.
Proof. For g, h C 2 () with g = h = 0 on , we have eigenfunction expansions
g(x) =
an n (x)
and
h(x) =
n=1
bn n (x).
n=1
Assume the solution u(x, t) may be expanded in the eigenfunctions with coecients
&
depending on t: u(x, t) =
n=1 un (t)n (x). This implies
n=1
un (t)
n un (t)n (x),
n=1
+ n un (t) = 0
for each n.
An cos n t + Bn sin n t n (x),
u(x, t) =
ut (x, t) =
u(x, 0) =
ut (x, 0) =
n=1
n An sin n t + n Bn cos n t n (x),
n=1
An n (x)
n=1
= g(x),
n Bn n (x) = h(x).
n=1
bn
Bn = .
n
bn
an cos n t + sin n t n (x),
n
n=1
an =
bn =
g(x)n(x) dx,
h(x)n (x) dx.
339
340
for x , t > 0
utt = uxx + uyy
(28.1)
for x
u(x, 0) = g(x), ut (x, 0) = h(x)
u(x, t) = 0
for x , t > 0.
in
uxx + uyy + u = 0
u(0, y) = 0 = u(a, y)
for 0 y b,
u(x, 0) = 0 = u(x, b)
for 0 x a.
Let u(x, y) = X(x)Y (y), then substitution in the PDE gives X Y + XY + XY = 0.
X Y
+
+ = 0.
X
Y
Letting = 2 + 2 and using boundary conditions, we nd the equations for X and
Y:
X + 2 X = 0
Y + 2 Y = 0
X(0) = X(a) = 0
Y (0) = Y (b) = 0.
m2
a2
n2
b2
mx
ny
2
sin
,
mn (x, y) = sin
a
b
ab
where m, n = 1, 2, . . . .
Second, we solve the Wave Equation (28.1) using the space eigenfunctions.
For g, h C 2 () with g = h = 0 on , we have eigenfunction expansions 76
g(x) =
an n (x)
and
h(x) =
n=1
Assume u(x, t) =
n=1
&
n=1
un (t)n (x).
This implies
for each n.
bn n (x).
An cos n t + Bn sin n t n (x),
u(x, t) =
ut (x, t) =
u(x, 0) =
ut (x, 0) =
n=1
n An sin n t + n Bn cos n t n (x),
n=1
An n (x)
n=1
= g(x),
n Bn n (x) = h(x).
n=1
bn
Bn = .
n
bmn
u(x, t) =
sin mn t mn (x),
amn cos mn t +
mn
m,n=1
341
342
McOwen, 4.4 #3; 266B Ralston Hw. Consider the initialboundary value problem
for x , t > 0
utt = u + f (x, t)
u(x, t) = 0
for x , t > 0
for x .
u(x, 0) = 0, ut (x, 0) = 0
Use Duhamels principle and an expansion of f in eigenfunctions to obtain a (formal)
solution.
Proof. a) We expand u in terms of the Dirichlet eigenfunctions of Laplacian in
.
n + n n = 0
in ,
n = 0
an (t)n (x),
an (t) =
fn (t)n (x),
fn (t) =
Assume
u(x, t) =
f (x, t) =
n=1
n=1
an (t)
=
on .
n (x)u(x, t) dx.
n (x)f (x, t) dx.
n (x)utt dx =
n (u + f ) dx =
n u dx +
n f dx
n u dx +
n f dx = n
n u dx +
n f dx = n an (t) + fn (t).
=
fn
an (0) =
an (0) =
n (x)u(x, 0) dx = 0.
n (x)ut(x, 0) dx = 0.
77
n + n an = 0
t
a
an + n an = fn (t)
a
(t)
=
an (t s, s) ds.
an (0) = 0
an (0, s) = 0
n
0
an (0) = 0
an (0, s) = fn (s)
sin n t
.
an (t, s) = fn (s)
n
Duhamels principle gives
t
t
sin( n (t s))
a
n (t s, s) ds =
fn (s)
ds.
an (t) =
n
0
0
n (x) t
u(x, t) =
fn (s) sin( n (t s)) ds.
n 0
n=1
u
n
We used Greens formula: n n
u
n ds = (n u n u) dx.
On , u = 0; n = 0 since eigenfunctions are Dirichlet.
77
343
0 x , t 0
0 x ,
u(x, 0) = g(x), ut (x, 0) = h(x)
where the coecient a(t) = 0.
a) Express (formally) the solution of this problem by the method of eigenfunction expansions.
b) Show that this problem is not wellposed if a 1.
Hint: Take f = 0 and prove that the solution does not depend continuously on the
initial data g, h.
Proof. a) We expand u in terms of the Dirichlet eigenfunctions of Laplacian in
.
n xx + n n = 0 in ,
n (0) = n () = 0.
un (t)n (x),
un (t) =
n (x)u(x, t) dx.
u(x, t) =
f (x, t) =
g(x) =
h(x) =
n=1
fn (t)n (x),
fn (t) =
n=1
gn =
n (x)g(x) dx.
hn n (x),
hn =
n=1
un (t)
gn n (x),
n=1
n (x)h(x) dx.
n (x)utt dx =
n (a(t)uxx + f ) dx = a(t)
n uxx dx +
n f dx
n xxu dx +
n f dx = n a(t)
n u dx +
n f dx
= a(t)
fn
un + n a(t)un = fn (t)
un (0) = gn
un (0) = hn .
344
Note: The initial data is not 0; therefore, the Duhamels principle is not applicable.
Also, the ODE is not linear in t, and its solution is not obvious. Thus,
u(x, t) =
un (t)n (x),
n=1
345
u2tt + u2xx = 0,
u1tt + u1xx = 0,
u1 (0, t) = u1 (, t) = 0,
u2 (0, t) = u2 (, t) = 0,
That is, the dierence of the two solutions is not bounded by the dierence of initial
data.
By the method of separation of variables, we may obtain
u(x, t) =
u(x, 0) =
ut (x, 0) =
n=1
n=1
n=1
Not complete.
We also know that for elliptic equations, and for Laplace equation in particular, the
value of the function u has to be prescribed on the entire boundary, i.e. u = g on
, which is not the case here, making the problem underdetermined. Also, ut is
prescribed on one of the boundaries, making the problem overdetermined.
29
346
for x , t > 0
ut = u
u(x, 0) = g(x)
for x
u(x, t) = 0
for x , t > 0.
Proof. For g C 2 () with g = 0 on , we have eigenfunction expansion
g(x) =
an n (x)
n=1
Assume the solution u(x, t)&may be expanded in the eigenfunctions with coecients
depending on t: u(x, t) =
n=1 un (t)n (x). This implies
n=1
un (t)
= n
un (t)n (x),
n=1
+ n un (t) = 0,
Thus,
un (t) = An en t.
An en t n (x),
u(x, t) =
u(x, 0) =
n=1
An n (x) = g(x).
n=1
an en tn (x),
n=1
with
an =
Also
u(x, t) =
n t
an e
n=1
n=1
n (x) =
n=1
g(y)n(y) dy en t n (x)
K(x,y,t),
heat kernel
347
for x , t > 0
ut = uxx + uyy
(29.1)
u(x, 0) = g(x)
for x
u(x, t) = 0
for x , t > 0.
in
uxx + uyy + u = 0
u(0, y) = 0 = u(a, y)
for 0 y b,
u(x, 0) = 0 = u(x, b)
for 0 x a.
Let u(x, y) = X(x)Y (y), then substitution in the PDE gives X Y + XY + XY = 0.
X Y
+
+ = 0.
X
Y
Letting = 2 + 2 and using boundary conditions, we nd the equations for X and
Y:
X + 2 X = 0
Y + 2 Y = 0
X(0) = X(a) = 0
Y (0) = Y (b) = 0.
m2
a2
n2
b2
mx
ny
2
sin
,
mn (x, y) = sin
a
b
ab
where m, n = 1, 2, . . . .
Second, we solve the Heat Equation (29.1) using the space eigenfunctions.
For g C 2 () with g = 0 on , we have eigenfunction expansion
g(x) =
an n (x).
n=1
Assume u(x, t) =
&
n=1
un (t)n (x).
This implies
Thus,
un (t) = An en t.
An en t n (x),
u(x, t) =
u(x, 0) =
n=1
n=1
An n (x) = g(x).
m,n=1
348
349
in
uxx + uyy + u = 0
u(0, y) = u(2, y)
for 0 y 2,
u(x, 0) = u(x, 2)
for 0 x 2.
Let u(x, y) = X(x)Y (y), then substitution in the PDE gives X Y + XY + XY = 0.
X Y
+
+ = 0.
X
Y
Letting = 2 + 2 and using periodic BCs, we nd the equations for X and Y :
X + 2 X = 0
Y + 2 Y = 0
X(0) = X(2)
Y (0) = Y (2).
n = n
imx
Yn (y) = einy ,
Xm (x) = e
where m, n = . . . , 2, 1, 0, 1, 2, . . ..
Second, we solve the
using the space eigenfunctions.
&Heat Equationimx
Assume u(x, y, t) = m,n= umn (t)e einy . This implies
umn (t) + (m2 + n2 )umn (t) = 0,
2 +n2 )t
m,n=
Thus,
2 +n2 )t
cmn e(m
eimx einy .
u(x, y, 0) =
m,n=
350
m,n=
2
= 2
1
4 2
cmn =
0
n=
2 2
0
Proof. We have
ut = uxx + uyy
Multiply the equation by u and integrate:
uut = uu,
1 d 2
u = uu,
2 dt
1 d
2
u dxdy =
uu dxdy =
2 dt
=
351
u
u
ds
n
=0, (periodic BC)
u2 dxdy
u2 dxdy 0.
352
(1) +
d
(1) = 0.
dx
2
.
tan =
1
c) Solve the following initialboundary value problem on 0 < x < 1, t > 0
u
2u
=
,
t
x2
u
(0, t) = 0,
u(0, t)
x
u(x, 0) = f (x).
u(1, t) +
u
(1, t) = 0,
x
You may call the relevant eigenfunctions n (x) and assume that they are known.
Proof. a) If = 0, the ODE reduces to = 0. Try (x) = Ax + B.
From the rst boundary condition,
(0) (0) = 0 = B A
B = A.
Thus, the solution takes the form (x) = Ax+A. The second boundary condition gives
(1) + (1) = 0 = 3A
A = B = 0.
Thus the only solution is 0, which is not an eigenfunction, and 0 not an eigenvalue.
(29.2)
and
From (29.2), =
1+
B
= e2 ,
1
A
A+B
AB
or
A
= e .
B
A+B
A
= e BA , which has no solutions.
B
and thus,
B = A .
(0) (0) = 0 = B A
353
= A (1 ) sin( ) + 2 cos( )
A = 0 (since A = 0 implies B = 0 and = 0, which is notan eigenfunction). Therefore,
ut = uxx ,
u(1, t) + ux (1, t) = 0,
u(0, t) ux(0, t) = 0,
u(x, 0) = f (x).
For f , we have an eigenfunction expansion
f (x) =
an n (x).
n=1
Assume u(x, t) =
un (t)
&
n=1
un (t)n (x).
+ n un (t) = 0,
This implies
Thus,
un (t) = An en t.
An en t n (x),
u(x, t) =
u(x, 0) =
n=1
An n (x) = f (x).
n=1
an en tn (x),
n=1
with
an =
354
Problem (W03, #3); 266B Ralston Hw. Let be a smooth domain in three
dimensions and consider the initialboundary value problem for the heat equation
for x , t > 0
ut = u + f (x)
u/n = 0
for x , t > 0
u(x, 0) = g(x)
for x ,
in which f and g are known smooth functions with
for x .
g/n = 0
n
= 0 on .
n
in ,
Note that here 1 = 0 and 1 is the constant V 1/2 , where V is the volume of .
Assume
an (t)n (x),
an (t) =
n (x)u(x, t) dx.
u(x, t) =
f (x) =
g(x) =
n=1
fn n (x),
n=1
fn =
gn n (x),
gn =
n=1
an (t)
=
n (x)ut dx =
n (u + f ) dx =
n u dx +
n f dx
n u dx +
n f dx = n
n u dx +
n f dx = n an + fn .
=
an (0) =
78
fn
n (x)u(x, 0) dx =
n g dx = gn .
fn n t fn
+
.
e
an (t) = gn
n
n
u
n
We used Greens formula: n n
u
n ds = (n u n u) dx.
n
= 0;
= 0 since eigenfunctions are Neumann.
On , u
n
n
78
gn
n=2
If f1 = 0
If f1 = 0
f (x) dx = 0 ,
f (x) dx = 0 ,
fn n t fn
+
e
n (x).
n
n
lim u(x, t) = g1 1 +
lim u(x, t) f1 1 t.
fn n
n=2
355
356
357
Problem (S97, #2). a) Consider the eigenvalue problem for the Laplace operator
in R2 with zero Neumann boundary condition
in
uxx + uyy + u = 0
u
on .
n = 0
Prove that 0 = 0 is the lowest eigenvalue and that it is simple.
b) Assume that the eigenfunctions n (x, y) of the problem in (a) form a complete
orthogonal system, and that f (x, y) has a uniformly convergent expansion
f (x, y) =
fn n (x, y).
n=0
u
u = 0.
n
u
u
ds
u2 dx + u2 dx = 0,
n
=0
2
u dx =
u2 dx,
n=1
an (t)
n (x, y)ut dx =
n (u + f ) dx =
n u dx +
n f dx
=
n u dx +
n f dx = n
n u dx +
n f dx = n an + fn .
=
fn
an (0) =
80
79
n
= 0 on .
n
an (t) =
n (x, y)u(x, y, t) dx.
in ,
358
n (x, y)u(x, y, 0) dx = 0.
fn n t fn
e
+
.
n
n
u(x, t) = f1 1 t +
n=2
If f1 = 0
If f1 = 0
fn n t fn
e
+
n (x).
n
n
f (x) dx = 0 ,
lim u(x, t) =
f (x) dx = 0 ,
fn n
n=2
lim u(x, t) f1 1 t.
v
ds =2 0,
n
79
We use dx dy dx.
u
n
We used Greens formula: n n
u
ds = (n u n u) dx.
n
n
= 0;
= 0 since eigenfunctions are Neumann.
On , u
n
n
80
359
&
&
Assume v(x, y) =
n=0 an n (x, y). Since we have f (x, y) =
n=0 fn n (x, y), we
obtain
n an n +
fn n = 0,
n=0
n=0
n an n + fn n = 0,
fn
.
an =
n
v(x, y) =
&
fn
n=0 ( n )n (x, y).
29.1
360
Problem (F99, #5). In two spatial dimensions, consider the dierential equation
ut = u 2 u
with periodic boundary conditions on the unit square [0, 2]2.
a) If = 2 nd a solution whose amplitude increases as t increases.
b) Find a value 0 , so that the solution of this PDE stays bounded as t , if < 0 .
Proof. a) Eigenfunctions of the Laplacian.
The periodic boundary conditions imply a Fourier Series solution of the form:
amn (t)ei(mx+ny) .
u(x, t) =
m,n
ut =
amn (t)ei(mx+ny) ,
m,n
u = uxx + uyy =
(m2 + n2 ) amn (t)ei(mx+ny) ,
m,n
m,n
m,n
oscillates
When = 2, we have
2
2
2
2
mn e(m +n )[2(m +n )]t ei(mx+ny) .
u(x, t) =
m,n
361
362
(29.4)
2 2
Derive an explicit bound on u(t) and show that it stays nite for all t.
b) If a = 0, construct the normal modes for (29.4); i.e. nd all solutions of the form
u(x, y, t) = et+ikx+ily .
c) Use these normal modes to construct a solution of (29.4) with a = 0 for the initial
data
u(x, y, t = 0) =
1
1
1 ix
2e
1
1 12 eix
u
1 d
2
2
4
ds
u dx =
u
u dx
au dx +
bu2 dx,
2 dt
n
d
u22 2b u22,
dt
u22 u(x, 0)22 e2bt ,
u2 u(x, 0)2 ebt ebt .
Thus, u stays nite for all t.
b) Since a = 0, plugging u = et+ikx+ily into the equation, we obtain:
ut
t+ikx+ily
= (k2 l 2 + b) et+ikx+ily ,
= k2 l 2 + b.
Thus,
2
1
k=0
k=0
= 2+
eix
k
+
1 ikx
e +
2k
k=1
1
k=0
k=0
ikx
ak e
k=
1 ikx
1 ikx
e
+
e .
k
k
2
2
k=1
1 ikx
1 ikx
= 2+
e +
e ,
k
k
2
2
k=1
a0 = 2;
a0 = 2;
k=1
1
, k > 0;
2k
1
ak = k , k =
0.
2
ak =
u(x, y, t) = 2ebt +
ak =
+
81
1
, k<0
2k
1 (k2 +b)t+ikx
e
.
k
2
k=, k=0
81
1 ikx
e
,
2k
eix
363
364
Problem (S00, #3). Consider the initialboundary value problem for u = u(x, y, t)
ut = u u
for (x, y) [0, 2]2, with periodic boundary conditions and with
u(x, y, 0) = u0 (x, y)
in which u0 is periodic. Find an asymptotic expansion for u for t large with terms
tending to zero increasingly rapidly as t .
Proof. Since we have periodic boundary conditions, assume
umn (t) ei(mx+ny) .
u(x, y, t) =
m,n
m,n
umn (t)
Since u0 is periodic,
u0mn ei(mx+ny) ,
u0 (x, y) =
u0mn
m,n
1
=
4 2
2
0
m,n
u(x, y, t) =
amn ei(mx+ny) =
u0mn ei(mx+ny) ,
m,n
amn = u0mn .
2 +n2 +1)t
u0mn e(m
ei(mx+ny) .
m,n
2 +n2 +1)t
u0mn e(m
2 +n2 +1)t
0 as t .
30
= 0,
ut u(1)
x
(2)
(2)
= 0.
ut 5u(1)
x 3ux
Take the Fourier transform in x. The transformed initial value problems are:
(1)
u(1) = 0,
u
t i
(2)
u(1) 3i
u(2) = 0,
u
t 5i
u
(1)(, 0) = f(),
u
(2)(, 0) = 0.
With this u
(1), the second initial value problem becomes
(2)
it
u(2) = 5i f()e
,
u
t 3i
u
(2)(, 0) = 0.
u
h (, t) = c1 e3it .
(2)
With u
p = c2 eit as anzats for a particular solution, we obtain:
ic2 eit 3ic2 eit = 5i f()eit,
2ic2 eit = 5i f()eit,
5
c2 = f().
2
5
it
u
(2)
p (, t) = f ()e .
2
5
(2)
3it
h (, t) + u
(2)
f()eit.
u
(2)(, t) = u
p (, t) = c1 e
2
We nd c1 using initial conditions:
5
5
c1 = f().
u
(2)(, 0) = c1 f() = 0
2
2
Thus,
5
u
(2)(, t) = f() e3it eit .
2
365
366
367
Problem (S02, #4). Use the Fourier transform on L2 (R) to show that
du
+ cu(x) + u(x 1) = f
dx
(30.1)
has a unique solution u L2 (R) for each f L2 (R) when c > 1  you may assume
that c is a real number.
by
Proof. u L2 (R). Dene its Fourier transform u
1
eix u(x) dx
for R.
u
() =
2 R
*
du
() = i
u().
dx
We can nd u(x
1)() in two ways.
Let u(x
1) = v(x), and determinte v():
y
u(x
1)() = v() =
=
1
1
eix v(x) dx =
ei(y+1) u(y) dy
2 R
2 R
1
eiy ei u(y) dy = ei u
().
2 R
u(x
1)() = ei u
Substituting into (30.1), we obtain
() = f(),
i
u() + c
u() + ei u
f()
.
u
() =
i + c + ei
f()
= 1 f B,
= fB
u(x) =
i
i + c + e
2
1
=
,
where B
i + c + ei
1
eix
1
=
d.
B=
i + c + ei
2 R i + c + ei
For c > 1, u
() exists for all R, so that u(x) = (
u()) and this is unique by the
Fourier Inversion Theorem.
1
1
ix
e
v(x) dx =
ei(y+1) u(y) dy
n
n
n
n
(2) 2 R
(2) 2 R
&
1
eiy ei1 u(y) dy = ei1 u
() = e(i j j ) u
().
n
(2) 2 Rn
368
Problem (F96, #3). Find the fundamental solution for the equation
ut = uxx xux.
(30.2)
Hint: The Fourier transform converts this problem into a PDE which can be solved
using the method of characteristics.
by
Proof. u L2 (R). Dene its Fourier transform u
1
eix u(x) dx
for R.
u
() =
2 R
(),
u
*x () = i u
2
() = 2 u
().
u
xx () = (i) u
We nd x
ux () in two steps:
Multiplication by x:
d
1
u
().
eix ixu(x) dx =
ixu() =
d
2 R
d
().
xu(x)()
= i u
d
Using the previous result, we nd:
1
1 ix
1
ix
e
e
xu
xux (x) dx =
(i)eix x + eix u dx
xux (x)() =
2 R
2
2 R
=0
1
1
eix u dx
= i eix x u dx
2
2
R
R
d
d
() u
() = u
() u
().
= i xu(x)()
u
() = i i u
d
d
d
u
() u
().
xu
x (x)() =
d
Plugging these into (30.2), we get:
u
(, t) = 2 u
(, t) u
(, t) ,
(, t) u
t
d
2
+
u + u
,
u
t = u
u = ( 2 1)
u.
u
t
We now solve the above equation by characteristics.
We change the notation: u
u, t y, x. We have
uy xux = (x2 1)u.
dx
dt
dy
dt
dz
dt
x = c1 et ,
(c1 = xet )
y = t + c2 ,
dz
= (c21 e2t 1)dt
z
x2
x2
x2
1
+ t + c3 =
+ y c2 + c3 z = ce 2 +y .
log z = c21 e2t + t + c3 =
2
2
2
=
u
(, t) = ce 2
+t
Thus, we have
2
u
(, t) = ce 2
+t
x2
2
Check:
ut = c et e
x2
2
x2
2
ux = c et xe ,
x2
x2
uxx = c et e 2 + x2 e 2 .
Thus,
ut = uxx xux ,
x2
x2
x2
= c et e 2 + x2 e 2 x c et xe 2 .
c et e
x2
2
82
369
370
x Rn ,
k=1
u(0, x) = f (x)
where ak (t), k = 1, . . ., n, and a0 (t) are continuous functions, and f is a continuous
function. You may assume f has compact support.
b) Solve the initial value problem
u
u
+
ak (t)
+ a0 (t)u = f (x, t),
t
xk
n
x Rn ,
k=1
u(0, x) = 0
where f is continuous in x and t.
Proof. a) Use the Fourier transform to solve this problem.
1
eix u(x, t) dx
for R.
u
(, t) =
n
2
n
(2)
R
u
= ik u
.
xk
Thus, the equation becomes:
&
+ a0 (t)
u = 0,
u
t + i nk=1 ak (t)k u
u
(, 0) = f (),
or
+ a0 (t)
u = 0,
u
t + i a(t) u
.
u
t = i a(t) + a0 (t) u
This is an ODE in u
with solution:
t
u
(, t) = ce 0 (i a(s)+a0 (s)) ds ,
t
u
(, t) = f() e 0 (i a(s)+a0 (s)) ds .
u
(, 0) = c = f().
Thus,
where
g() = e 0 (i a(s)+a0 (s)) ds .
t
1
1
ix
e
g() d =
eix e 0 (i a(s)+a0 (s)) ds d.
g(x) =
n
n
(2) 2 Rn
(2) 2 Rn
t (i a(s)+a
(f g)(x)
1
0 (s)) ds
i(xy)
u(x, t) =
=
e
e 0
d f (y) dy.
n
n
(2) Rn Rn
(2) 2
U
U
+
ak (t)
+ a0 (t)U = 0,
t
xk
n
k=1
U (x, t s, s) ds =
t
0
Rn
Rn
371
ts
ei(xy) e 0 (i a(s)+a0 (s)) ds d f (y, s) dy ds.
372
83
,  < a,
1
f () =
2 ,  = a,
0,  > a,
where a is a real constant, what f (x) does the inversion theorem give?
b) Show that
(x),
f (x
b) = eib f
where b is a real constant. Hence, using part (a) and Parsevals theorem, show that
1 sin a(x + z) sin a(x + )
sin a(z )
dx =
,
x+z
x+
z
where z and are real constants.
Proof. a) The inverse Fourier transform for f L1 (Rn ):
1
eix f (x) dx
for R.
f () =
2
Fourier Inversion Theorem: Assume f L2 (R). Then
1
f (x) =
2
ix
1
f() d =
2
Also,
f (x) g(x) dx =
We can write
,
f () =
0,
83
1
2
or
g() d.
f()
 < a,
 > a.
Note that the Fourier transform is dened incorrectly here. There should be  sign in eix .
Need to be careful, since the consequences of this denition propagate throughout the solution.
373
a
a
1
1
1
1
eix f() d =
0 d +
eix d +
0 d
f (x) = (f()) =
2
2
2 a
2 a
1 ix =a
1 iax
sin ax
1 a ix
e
e
.
e
d =
=
eiax =
=
2 a
2ix
2ix
x
=a
g():
b) Let f (x
b) = g(x), and determinte
y
f (x
b)() =
g() =
eix g(x) dx =
R
R
iy ib
ei(y+b) f (y) dy
sin ax
x
sin a(z s)
ei(zs)a ei(zs)a
=
.
2i(z s)
zs
374
eix/2 y, x y
f (x) =
0,
x > y,
in which y and are constants.
Use this in Parsevals relation to show that
sin2 ( )y
d = y.
( )2
Also,
f (x) dx =
2 d.
f()
f (x) g(x) dx =
or
g() d.
f()
e
f (x) dx =
e
ei()x dx
f () =
dx =
2 y
2 2y y
2 R
2 y
x=y
i()y
1
1
1
ei()x
e
=
ei()y
=
2 2y i( )
2i 2y( )
x=y
sin y( )
.
=
2y( )
Parsevals theorem gives:
2
f() d =
f (x)2 dx,
y 2ix
e
sin2 y( )
dx,
d =
2
2
2y( )
y 4y
y
sin2 y( )
d =
dx,
( )2
2 y
sin2 y( )
d = y.
( )2
We had
f() =
sin y( )
.
2y( )
or
1
1
ix
e g() d =
eix( a) f( ) d
2 R
2 R
ixa
= e
f (x).
f( + a) = g() =
1 2
( 2y)
f () f( + ) d
f (x) e()ixf (x) dx
2y
f (x)2 e()ix dx
2y
y 2ix
e
e()ix dx
2y
4y
y
1 y ()ix
e
dx
2 y
()ixx=y
1
e
x=y
2i( )
()iy
1
e
e()iy
2i( )
sin( )y
.
375
+
f = 0
f
x2 y 2
376
(30.3)
f (x, 0) = h(x).
the Neumann problem y
Assume that f , g and h are 2 periodic in x and that f is bounded at innity.
Find the Fourier transform N of the DirichletNeumann map. In other words,
nd an operator N taking the Fourier transform of g to the Fourier transform of h; i.e.
hk .
N
gk =
f (x, y) =
an (y) einx.
n=
= 0,
n=
f (x, 0) =
n=
fy (x, 0) =
inx
an (0)e
= g(x) =
n=
n=
gn einx
an (0) =
gn .
hn einx
an (0) =
hn .
n=
(Dirichlet)
(Neumann)
a0 (y) = b0 y + c0 .
an (y) = bn eny + cn eny , n = 1, 2, . . .;
ny
ny
, n = 1, 2, . . .;
a0 (y) = b0 .
an (y) = nbn e ncn e
for n > 0,
cn = 0
for n < 0,
b0 = 0,
c0 arbitrary.
377
n > 0:
an (y)
cn eny ,
an (0)
cn =
gn ,
(Dirichlet)
ncn
(Neumann)
an (0)
=
hn .
=
hn .
n
gn
n < 0:
an (y)
bn eny ,
an (0)
an (0)
bn =
gn ,
nbn =
hn .
(Dirichlet)
(Neumann)
hn .
n
gn =
n
gn =
hn ,
n = 0.
n = 0 : a0 (y) = c0 ,
g0 ,
a0 (0) = c0 =
h0 .
a0 (0) = 0 =
(Dirichlet)
(Neumann)
inx
an (y) e
= a0 (y) +
n=
= c0 +
=
1
an (y) e
n=
1
bn eny einx +
n=
an (y) einx
n=1
cn eny einx
n=1
&
gn e e +
gn eny einx ,
n=
n=1
&1
&
hn ny inx
e .
c0 + n= hnn eny einx +
n=1 n e
g0 +
&1
inx
ny inx
(Dirichlet)
(Neumann)
f(, y) = c1 ey + c2 ey .
For > 0, c1 = 0; for < 0, c2 = 0.
fy (, y) = c2 ey ,
c2 ey ,
1
1
eix f (x, 0) dx =
eix g(x) dx =
g(),
c2 = f(, 0) =
2
2
1
1
ix
e
fy (x, 0) dx =
eix h(x) dx =
h().
c2 = fy (, 0) =
2
2
g() =
h().
>0:
f(, y) =
(Dirichlet)
(Neumann)
378
fy (, y) = c1 ey ,
c1 ey ,
1
1
ix
c1 = f (, 0) =
e
f (x, 0) dx =
eix g(x) dx =
g(),
2
2
1
1
eix fy (x, 0) dx =
eix h(x) dx =
h().
c1 = fy (, 0) =
2
2
g() = h().
<0:
f(, y) =

g() =
h().
(Dirichlet)
(Neumann)
379
xn > 0,
u
+ k2 u = 0,
xn > 0
xn
x = (x1 , . . ., xn1 ).
u(x , 0) = f (x ),
u + a
84
( , xn) +
 2 u
Thus, the ODE and initial conditions of the transformed problem become:
u
xn xn + a
uxn + (k2  2 )
u = 0,
u
( , 0) = f( ).
With the anzats u
= cesxn , we obtain s2 + as + (k2  2 ) = 0, and
a a2 4(k2  2 )
.
s1,2 =
2
Choosing only the negative root, we obtain the solution: 85
u
( , xn ) = c( ) e
u
( , xn ) = f( ) e
a2 4(k2  2 )
xn
2
a2 4(k2  2 )
xn
2

u2L2 (Rn1 )
=
=
Rn1

u( , xn)2 d
aa2 4(k2  2 ) x 2
n
f( ) e
2
d
Rn1
2 2 n1
f
L (R
)
Thus,
=
u
( , 0) = c = f( ).
f 2L2(Rn1 )
Rn1
2
f( ) d
C,
84
Note that the last element of x = (x , xn ) = (x1 , . . . , xn1 , xn ), i.e. xn , plays a role of time t.
As such, the PDE may be written as
u + utt + aut + k2 u = 0.
85
Note that a > 0 should have been provided by the statement of the problem.
380
a)
where a(t), b(t), c(t) are continuous functions on [0, +], a(t) > 0 for t > 0.
b)
n
G
G
(x1 , . . . , xn , y1 , . . . , yn , t) =
ak (t)
t
xk
for t > 0,
k=1
y, t) = c e
G(,
0 [a(s)
2 +i b(s)+c(s)] ds
=
eix G(,
G(x, y, t) = G(, y, t)
2
R
1
t
1
2
eix eiy e 0 [a(s) +i b(s)+c(s)] ds d
=
2 R
2
t
1
2
ei(xy) e 0 [a(s) +i b(s)+c(s)] ds d.
=
2 R
y = (y1 , . . . , yn ). Transform in x:
b) Denote x = (x1 , . . . , xn ),
1
,
y, t) =
eix G(x, y, t) dx.
G(
n
(2) 2 Rn
The equation is transformed to an ODE, that can be solved:
y, t) =
t(,
G
n
k=1
,
y, t) = c ei
G(
y , t),
,
ak (t) ik G(
t &n
k=1 ak (s) k ] ds
0[
for t > 0
381
1
i
y
.
n e
2
(2)
382
2
2
+
+
u=f
x2n
x21
in Rn ,
(30.4)
where f is an integrable function (i.e. f L1 (Rn )), satisfying f (x) = 0 for x R.
Solve (30.4) by Fourier transform, and prove the following results.
2
n
a) There
is a solution of (30.4) belonging to L (R ) if n > 4. 2 n
b) If Rn f (x) dx = 0, there is a solution of (30.4) belonging to L (R ) if n > 2.
Proof.
u = f,
() = f(),
 u
1
u
() = 2 f(),

f()
.
u(x) =
2
2
Rn ,
a) Then

uL2 (Rn) =
Rn
f()2
d
4
1
2
<1
1
2
f()2
f()2
d +
d
.
4
4

1 
A
2 B, so B < .
Notice, f 2 = f
Use polar coordinates on A.
1
1
2
f()2
f
n1
2 r n5 dSn1 dr.
d =
r
dSn1 dr =
f
A =
4
4

r
0
0
<1
Sn1
Sn1
If n > 4,
Sn1
1
b) We have
u(x, t) =
=
=
=
=
f()
1
ix f ()
=
e
d
n
2
2
(2) 2
ix
e
1
1
iy
e
f
(y)
dy
d
n
n
(2) 2 2 (2) 2
ei(xy)
1
f
(y)
d dy
(2)n
2
1
ei(xy)r n1
1
f
(y)
r
dS
dr
dy
n1
(2)n
r2
0
Sn1
1
1
i(xy)r n3
f
(y)
e
r
dS
dr
dy.
n1
(2)n
0
Sn1
M < , if n>2.
u(x, t) =
1
M
f
(y)
dy
< .
(2)n
383
384
Problem (F02, #7). For the right choice of the constant c, the function
F (x, y) = c(x + iy)1 is a fundamental solution for the equation
u
u
+i
=f
x
y
in R2 .
Find the right choice of c, and use your answer to compute the Fourier transform
(in distribution sense) of (x + iy)1 .
Proof.
86
=
F1 (x, y) =
+i
i
.
x
y x
y
1
2
F1 (x, y) = ,
+i
i
F (x, y) = .
x
y x
y
hx + ihy = ei(x1 +y2 ) .
Suppose h = h(x1 + y2 ) or h = cei(x1+y2 ) .
c i1 ei(x1 +y2 ) i2 2 ei(x1 +y2 ) = ic(1 i2 ) ei(x1 +y2 ) ei(x1 +y2 ) ,
ic(1 i2 ) = 1,
1
,
c =
i(1 i2 )
1
ei(x1 +y2 ) .
h(x, y) =
i(1 i2 )
Integrate by parts:
1
1
1
() =
+i
dxdy
ei(x1 +y2 )
x + iy
i(1 i2 ) x
y (x + iy) 0
R2
1
1
=
.
=
i(1 i2 )
i(2 + i1 )
86
31
385
Laplace Transform
est u(t) dt
(s > 0).
In practice, for a PDE involving time, it may be useful to perform a Laplace transform
in t, holding the space variables x xed.
The inversion formula for the Laplace transform is:
u(t) = L
1
[u (s)] =
2i
#
c+i
ci
Example: f (t) = 1.
1
t=
1
L[1] =
est 1 dt = est
=
s
s
t=0
0
Example: f (t) = eat .
at
st at
e e dt =
L[e ] =
0
e(as)t dt =
for s > 0.
1 (as)t t=
1
e
=
as
sa
t=0
1
s
for s > a.
1
.
s2 +1
t
1
1
2
1 sin t dt = 1 cos t.
= f g =
s
s
+
1
0
L[f ]
L[g]
0
2
L[u],
est ux dt =
L[ux] =
x
0
2
est uxx dt =
L[u].
L[uxx] =
x2
0
Heat Equation: Consider
in U (0, )
ut u = 0
u=f
on U {t = 0},
and perform a Laplace transform with respect to time:
est ut dt = sL[u] u(x, 0) = sL[u] f (x),
L[ut] =
0
est u dt = L[u].
L[u] =
0
386
in U.
Thus, the solution of this equation with RHS f is the Laplace transform of the solution
of the heat equation with initial data f .
s>0
s>0
s > a
s > a
s>a
s>a
s>0
s>a
s>0
t
0
d2 f
df
2
L[f
]
sf
(0)
f
(0),
f
=
=
s
L
dt2
dt
dn f
= sn L[f ] sn1 f (0) . . . f n1 (0),
L
dtn
1 #
s
f
,
L[f (at)] =
a
a
L[ebtf (t)] = f # (s b),
d
L[tf (t)] = L[f ],
ds
f (t)
=
f # (s ) ds ,
L
t
s
t
1
L[f ],
f (t ) dt =
L
s
0
1
1 + s2
0
387
Example: f (t) = tn .
tn est n
n
n
st n
L[t ] =
e t dt =
+
est tn1 dt = L[tn1 ]
s
s 0
s
0
0
n!
n!
n n1
L[tn2 ] = . . . = n L[1] = n+1 .
=
s
s
s
s
388
t > 0, x > 0
u(x, 0) = 0,
x>0
u(0, t) = g(t),
t > 0,
L[ut] =
est ut dt =
t=
est u(x, t)
+s
t=0
est u dt
2 #
u (x, s) + au# (x, s) = 0.
x2
u# (x, s) = c1 e
s+a x
+ c2 e
s+a x
u# (x, s) = c2 e
s+a x
s+a x
u# (x, s) = g # (s)e
u# (0, s) = c2 = g # (s),
thus,
To obtain u(x, t), we take the inverse Laplace transform of u# (x, s):
s+a x
= gf
u(x, t) = L1 [u# (x, s)] = L1 g # (s) e
L[g]
L[f ]
1 c+i
1 s+a x
st s+a x
e e
ds ,
= g
= gL e
2i ci
t
1 c+i
u(x, t) =
g(t t )
est e s+a x ds dt .
2i ci
0
389
390
Problem (F04, #8). The function y(x, t) satises the partial dierential equation
2y
y
+
+ 2y = 0,
x xt
and the boundary conditions
x
y(0, t) = eat ,
y(x, 0) = 1,
where a 0. Find the Laplace transform, y(x, s), of the solution, and hence derive
an expression for y(x, t) in the domain x 0, t 0.
Proof. We change the notation: y u. We have
xux + uxt + 2u = 0,
u(x, 0) = 1, u(0, t) = eat .
The Laplace transform is dened as:
#
est u(x, t) dt
L[u(x, t)] = u (x, s) =
(s > 0).
est ux dt = x(u# )x ,
t=
st
st
e uxt dt = e ux (x, t)
+s
est ux dt
L[uxt] =
L[xux] =
st
xux dt = x
t=0
(since u(x, 0) = 0)
e(s+a)t dt =
1 (s+a)t t=
e
s+a
t=0
1
.
s+a
c2 =
c2
.
(x + s)2
s2
.
s+a
s2
.
(s + a)(x + s)2
To obtain u(x, t), we take the inverse Laplace transform of u# (x, s):
c+i
s2
1
s2
st
e
=
ds.
u(x, t) = L1 [u# (x, s)] = L1
(s + a)(x + s)2
2i ci
(s + a)(x + s)2
c+i
1
s2
u(x, t) =
est
ds.
2i ci
(s + a)(x + s)2
391
Problem (F90, #1). Using the Laplace transform, or any other convenient method,
solve the Volterra integral equation
x
sin(x y)u(y) dy.
u(x) = sin x +
0
L[sin t] =
1
,
1 + s2
u#
.
1 + s2
1
u#
u# + 1
+
=
.
2
2
1+s
1+s
1 + s2
u# (s) =
1
.
s2
392
Problem (F91, #5). In what follows, the Laplace transform of x(t) is denoted
either by x(s) or by Lx(t). Show that, for integral n 0,
L(tn ) =
n!
sn+1
1
LJ0 (2 ut) = eu/s ,
s
where
J0 (z) =
(1)n ( 1 z)2n
2
n!n!
n=0
1
1
L
.
J0 (2 ut)x(u) du = x
s
s
0
(31.1)
Assuming that
1
,
LJ0 (at) =
2
a + s2
prove with the help of (31.1) that if t 0
1
t
.
J0 (au)J0 (2 ut) du = J0
a
a
0
Hint: For the last part, use the uniqueness of the Laplace transform.
Proof.
L[tn ] =
tn est n
n
st
n
e
t
dt
=
+
est tn1 dt = L[tn1 ]
s
s
s
0
0
f
g
=0
n!
n!
n
n 1 n2
L[t ] = . . . = n L[1] = n+1 .
s
s
s
s
(1)n un tn
(1)nun
(1)n un
=
L[tn ] =
LJ0 (2 ut) = L
n!n!
n!n!
n!sn+1
=
L
n=0
1
(1)
s n=0 n!
J0 (2 ut) x(u) du
x# (s) =
0
u n
s
=
where
n=0
1 u
e s.
s
n=0
1
L[J0 (2 ut)] x(u) du =
s
0
1 #
1
x
,
s
s
0
e s x(u) du
32
393
32.1
Norms
32.2
A Banach space is a normed vector space that is complete in that norms metric. I.e.
a complete normed linear space is a Banach space.
A Hilbert space is an inner product space for which the corresponding normed space
is complete. I.e. a complete inner product space is a Hilbert space.
Examples: 1) Let K be a compact set of Rn and let C(K) denote the space of continuous
functions on K. Since every u C(K) achieves maximum and minimum values on K,
we may dene
u = max u(x).
xK
32.3
CauchySchwarz Inequality
32.4
1
1
uvdx ( u2 dx) 2 ( v 2 dx) 2
H
older Inequality
uv dx upvq,
1
p
1
q
32.5
394
Minkowski Inequality
u + vp up + vp,
Note that C 1 () is not a Banach space since u1, need not be nite for u C 1 ().
32.6
Sobolev Spaces
u1,2 =
(u + u ) dx
< u, u >1 =
1
2
(32.2)
when these expressions are dened and nite. For example, (32.1) and (32.2) are dened
for functions in C01 (). However, C01 () is not complete under the norm (32.2), and so
does not form a Hilbert space.
Divergence Theorem
dx
div A
A n dS =
Trace Theorem
uL2() CuH 1 ()
smooth or square
Poincare Inequality
1p
up Cup
2
u(x) dx C
u(x)2 dx
u u p up
u H01,p()
i.e. p = 2
395
u(x) dx
Notes
u
u
u
= u n = n1
+ n2
n
x1
x2
u
u dx
u dx =
ab
a+b
2
ab
a2 + b2
2
u2
uu = ( )
2
2
(uxy ) dx =
uxxuyy dx
uu
u H02 ()
u2 + u2
2
square
Problem (F04, #6). Let q C01 (R3 ). Prove that the vector eld
q(y)(x y)
1
dy
u(x) =
4 R3 x y3
enjoys the following properties: 87
a) u(x) is conservative;
b) div u(x) = q(x) for all x R3 ;
c) u(x) = O(x2 ) for large x.
Furthermore, prove that the proverties (1), (2), and (3) above determine the vector eld
u(x) uniquely.
Proof. a) To show that
u(x) is conservative, we need to show that curl u = 0.
The curl of V is another vector eld dened by
e1 e2 e3
V3 V2 V1 V3 V2 V1
1 2 3
.
=
curl V = V = det
x2 x3 x3
x1 x1
x2
V1 V2 V3
Consider
(x) =
V
(x1 , x2 , x3 )
x
=
3 .
2
x3
(x1 + x22 + x23 ) 2
Then,
u(x) =
curl u(x) =
curl V (x) = curl
=
1
q(y) V (x y) dy,
4 R3
1
q(y) curlx V (x y) dy.
4 R3
(x1 , x2 , x3 )
3
McOwen, p. 138140.
32 2x3 x2
5 ,
32 2x3 x1
32 2x1 x3
5 ,
32 2x1 x2
1
u(x) is conservative.
Thus, curl u = 4
R3 q(y) 0 dy = 0, and
1
b) Note that the Laplace kernel in R3 is 4r
.
q(y)(x y)
q(r) r
q(r)
1
1
dr = q.
dy =
r dr =
u(x) =
3
3
4 R3 x y
4 R3 r
R3 4r
c) Consider
1
F (x) =
4
R3
q(y)
dy.
x y
396