Electronic Journal of Qualitative Theory of Differential Equations

2007, No. 12, 1-14; http://www.math.u-szeged.hu/ejqtde/

UNIFORM CONTINUITY OF THE SOLUTION MAP FOR

NONLINEAR WAVE EQUATION IN
REISSNER-NORDSTRÖM METRIC
Svetlin Georgiev Georgiev
University of Sofia, Sofia, Bulgaria
email: sgg2000bg@yahoo.com

Abstract
In this paper we study the properties of the solutions to the Cauchy problem

(1) (utt − ∆u)gs = f (u) + g(|x|), t ∈ [0, 1], x ∈ R3 ,

(2) u(1, x) = u0 ∈ Ḣ 1 (R3 ), ut (1, x) = u1 ∈ L2 (R3 ),

where gs is the Reissner-Nordström metric (see [2]); f ∈ C 1 (R1 ), f (0) = 0, a|u| ≤

f 0 (u) ≤ b|u|, g ∈ C(R+ ), g(|x|) ≥ 0, g(|x|) = 0 for |x| ≥ r1 , a and b are positive
constants, r1 > 0 is suitable chosen.
When g(r) ≡ 0 we prove that the Cauchy problem (1), (2) has a nontrivial solution
u(t, r) in the form u(t, r) = v(t)ω(r) ∈ C((0, 1]Ḣ 1 (R+ )), where r = |x|, and the solution
map is not uniformly continuous.
When g(r) 6= 0 we prove that the Cauchy problem (1), (2) has a nontrivial solution
u(t, r) in the form u(t, r) = v(t)ω(r) ∈ C((0, 1]Ḣ 1 (R+ )), where r = |x|, and the solution
map is not uniformly continuous.
Subject classification: Primary 35L10, Secondary 35L50.

1. Introduction
In this paper we study the properties of the solutions to the Cachy problem

(1) (utt − ∆u)gs = f (u) + g(|x|), t ∈ [0, 1], x ∈ R3 ,

(2) u(1, x) = u0 ∈ Ḣ 1 (R3 ), ut (1, x) = u1 ∈ L2 (R3 ),

where gs is the Reissner-Nordström metric (see [2])

r 2 − Kr + Q2 2 r2
gs = 2
dt − 2 dr 2 − r 2 dφ2 − r 2 sin2 φdθ 2 ,
r r − Kr + Q2

EJQTDE, 2007 No. 12, p. 1

K and Q are positive constants, f ∈ C 1 (R1 ), f (0) = 0, a|u| ≤ f 0 (u) ≤ b|u|, g ∈ C(R+ ),
g(|x|) ≥ 0, g(|x|) = 0 for |x| ≥ r1 , a and b are positive constants, r1 > 0 is suitable chosen.
The Cauchy problem (1), (2) we may rewrite in the form
(1)
r2 1 1 1
2 2
utt − 2 ∂r ((r 2 −Kr+Q2 )ur )− 2 ∂φ (sin φuφ )− 2 2 uθθ = f (u)+g(r),
r − Kr + Q r r sin φ r sin φ

(2) u(1, r, φ, θ) = u0 ∈ Ḣ 1 (R+ ×[0, 2π]×[0, π]), ut (1, r, φ, θ) = u1 ∈ L2 (R+ ×[0, 2π]×[0, π]).

When gs is the Minkowski metric; u0 , u1 ∈ C0∞ (R3 ) in [5](see and [1], section 6.3) is
proved that there exists T > 0 and a unique local solution u ∈ C 2 ([0, T ) × R3 ) for the
Cauchy problem

(utt − ∆u)gs = f (u), f ∈ C 2 (R), t ∈ [0, T ], x ∈ R3 ,

u|t=0 = u0 , ut |t=0 = u1 ,
for which
sup |u(t, x)| = ∞.
t<T,x∈R3

When gs is the Minkowski metric, 1 ≤ p < 5 and initial data are in C0∞ (R3 ), in [5](see
and [1], section 6.3) is proved that the initial value problem

(utt − ∆u)gs = u|u|p−1 , t ∈ [0, T ], x ∈ R3 ,

u|t=0 = u0 , ut |t=0 = u1 ,
admits a global smooth solution.
When gs is the Minkowski metric and initial data are in C0∞ (R3 ), in [4](see and [1],
section 6.3) is proved that there exists a number 0 > 0 such that for any data (u0 , u1 ) ∈
C0∞ (R3 ) with E(u(0)) < 0 , the initial value problem

(utt − ∆u)gs = u5 , t ∈ [0, T ], x ∈ R3 ,

u|t=0 = u0 , ut |t=0 = u1 ,
admits a global smooth solution.
When gs is the Minkowski metric in [6] is proved that the Cauchy problem

(utt − ∆u)gs = f (u), t ∈ [0, 1], x ∈ R3 ,

u(1, x) = u0 , ut (1, x) = u1 ,
has global solution. Here f ∈ C 2 (R), f (0) = f 0 (0) = f 00 (0) = 0,

|f 00 (u) − f 00 (v)| ≤ B|u − v|q1

EJQTDE, 2007 No. 12, p. 2

 √
for |u| ≤ 1, |v| ≤ 1, B > 0, 2 − 1 < q1 ≤ 1, u0 ∈ C◦5 (R3 ), u1 ∈ C◦4 (R3 ), u0 (x) = u1 (x) = 0
for |x − x0 | > ρ, x0 and ρ are suitable chosen.
When gs is the Reissner - Nordström metric, n = 3, p > 1, q ≥ 1, γ ∈ (0, 1) are fixed
constants, f ∈ C 1 (R1 ), f (0) = 0, a|u| ≤ f 0 (u) ≤ b|u|, g ∈ C(R+ ), g(|x|) ≥ 0, g(|x|) = 0 for
|x| ≥ r1 , a and b are positive constants, r1 > 0 is suitable chosen, in [7] is proved that the
γ (R+ )) in the form
initial value problem (1), (2), has nontrivial solution u ∈ C((0, 1]Ḃp,q
(
v(t)ω(r) f or r ≤ r1 , t ∈ [0, 1],
u(t, r) =
0 f or r ≥ r1 , t ∈ [0, 1],

where r = |x|, for which limt−→0 ||u||Ḃp,q

γ
(R+ ) = ∞.
In this paper we will prove that the Cauchy problem (1), (2) has nontrivial solution
u = u(t, r) ∈ C((0, 1]Ḣ 1 (R+ )) and the solution map is not uniformly continuous. When
we say that the solution map (u◦ , u1 , g) −→ u(t, r) is uniformly continuous we understand:
for every positive constant  there exist positive constants δ and R such that for any two
solutions u, v of the Cauchy problem (1), (2), with right hands g = g1 , g = g2 of (1), so that

(20 ) E(1, u − v) ≤ δ, ||g1 ||L2 (R+ ) ≤ R, ||g2 ||L2 (R+ ) ≤ R, ||g1 − g2 ||L2 (R+ ) ≤ δ,

the following inequality holds

(200 ) E(t, u − v) ≤  f or ∀t ∈ [0, 1],

where  ∂
E(t, u) := ||∂t u(t, ·)||2L2 (R+ ) +  u(t, ·)||2L2 (R+ ) .

∂r

Our main results are

Theorem 1.1. Let K, Q are positive constants for which

 K 2 > 4Q2 , 1
1−K+Q2
≥ 1,
√
 1 − K + Q2 > 0 K− K 2 −4Q2 p
is enough small such that 2 − 2 1 − K + Q2 > 0.

Let also g ≡ 0, f ∈ C 1 (R1 ), f (0) = 0, a|u| ≤ f 0 (u) ≤ b|u|, a and b are positive constants.
Then the homogeneous Cauchy problem (1), (2) has nontrivial solution u(t, r) = v(t)ω(r) ∈
C((0, 1]Ḣ 1 (R+ )). Also there exists t◦ ∈ [0, 1) for which exists constant  > 0 such that for
every positive constant δ exist solutions u, v of (1), (2), so that

E(1, u − v) ≤ δ,

and
E(t◦ , u − v) ≥ .

EJQTDE, 2007 No. 12, p. 3

 Theorem 1.2. Let K, Q are positive constants for which

 K 2 > 4Q2 , 1
1−K+Q2 ≥ 1,
√
 1 − K + Q2 > 0 K− K 2 −4Q2 p
is enough small such that 2 − 2 1 − K + Q2 > 0.

Let also g 6= 0, g ∈ C(R+ ), g(r) ≥ 0 for r ≥ 0, g(r) = 0 for r ≥ r1 , f ∈ C 1 (R1 ), f (0) = 0,

a|u| ≤ f 0 (u) ≤ b|u|, a and b are positive constants. Then the nonhomogeneous Cauchy
problem (1), (2) has nontrivial solution u(t, r) = v(t)ω(r) ∈ C((0, 1]Ḣ 1 (R+ )). Also there
exists t◦ ∈ [0, 1) for which exists constant  > 0 such that for every pair positive constants
δ and R exist solutions u, v of (1), (2), with right hands g = g1 , g = g2 of (1), so that

E(1, u − v) ≤ δ, ||g1 ||L2 (R+ ) ≤ R, ||g2 ||L2 (R+ ) ≤ R, ||g1 − g2 ||L2 (R+ ) ≤ δ,

and
E(t◦ , u − v) ≥ .
The paper is organized as follows. In section 2 we prove theorem 1.1. In section 3 we prove
theorem 1.2.

2. Proof of theorem 1.1.

For fixed q ≥ 1 and γ ∈ (0, 1) we put
 qγ2qγ  1
q
C=
2qγ − 1
For fixed p > 1, q ≥ 1, γ ∈ (0, 1) and g ∈ C(R+ ), g(r) ≥ 0 for r ≥ 0 we suppose that

EJQTDE, 2007 No. 12, p. 4

the constants A > 0, Q > 0, a > 0, b > 0, B > 0, K > 0, 1 < β < α satisfy the conditions
  2
1 1 2a b
i1) 1−K+Q 2 1−K+Q 2 A2 + AB ≤ 1, A > 1, Aa > 1;

1 a br12
2 (1−αK+α2 Q2 ) 2A4 − AB ≥ 0,

α



2br12

a
− B 2 ≥ 0,


6 2 2 2 A2 

 2A α (1−αK+α Q)


a 1 1 4ar1
i2) 2A4 (1−αK+α2 Q2 ) β
− α − A3 B(1−K+Q 2 )2 ≥ 0,
   2
1 1 1 a 2 2b 2 1 1

β − α (1−αK+α2 Q2 )2 4A6 − r1 (1−K+Q2 )A2 B 2 − r1 1−K+Q2 maxr∈[0,r1 ] g(r) ≥ A2 ,





   2 2ar14
 1 − 1 1 a

2 Q2 )2 2A4 − AB(1−K+Q2 ) > 0,

 β α (1−αK+α 
1 2a 2b 1 22−γ 21−γ
i3)C (1−K+Q 2 )2 A2 + AB(1−K+Q2 ) + A2 1 +C 1 < 1,
   (q(1−γ)) q A (1−K+Q2 )2 (q(1−γ)) q
2

 12 1 a
2 Q2 4A6 − β 2 4A4
1 a
> 0,
i4)  α1 1−αK+α 1 a 1 a

2 Q2 2A6 − β 2 2A4 > 0,
 2
 α2 1−αK+α 2 1
K > 4Q , A ≥ 1−K+Q2 > 1,
√



2Q2 K− K 2 −4Q2




 1> K > 2 ,
 1 − K + Q2 > 0 is enough small such that



√
i5) K− K 2 −4Q2 p

 1 > 2 − 2 1 − K + Q2 > 0,
√ 2

< β < α ≤ 3,





 K− K 2 −4Q2 −2(1−K+Q2 )
 2br 2
a
− A2 B12 − r12 maxr∈[0,r1 ] g(r) ≥ 0,


4A6 α2(1−αK+α
 2 Q2 )  
1 2 1 2a b 2 2
i6) 1−K+Q 2 AB 1−K+Q2 A2 + AB + r1 maxs∈[0,r1 ] g(s) ≤ AB ;
 2
i7) maxs∈[0,r1 ] g(s) ≤ β1 − α1 A14 ,

where p
K 2 − 4Q2 q
K−
r1 = − 1 − K + Q2 .
2
Example. Let 0 <  << 31 is enough small,
√
1 1 3 3 1 1 K− K 2 −4Q2 p
A = 4 , B =  , p = 2 , q = 2 , γ = 3 , α = 3, β = 2 − 2 1 − K + Q2 ,
(
11 (r − r1 )2 f or r ≤ r1 ,
g(r) =
0 f or r ≥ r1 ,
4
K = 3 + 6  − 32 2 ,
1 20

Q2 = 31 + 16 20 − 21 2 ,
a = 4 , b = 3 .
Then
1 − αK + α2 Q2 = 1 − 3K + 9Q2 = 20 ,
1 − K + Q2 = 2 • .
When g(r) ≡ 0 we put
(10 ) u0 := v(1)ω(r) =

EJQTDE, 2007 No. 12, p. 5

 ( R
r1 1 R r1  s4

r τ 2 −Kτ +Q2 τ s2 −Ks+Q2 v 00 (1)ω(s) − s2 f (v(1)ω(s)) dsdτ f or r ≤ r1 ,
=
0 f or r ≥ r1 ,
and u1 ≡ 0. Here v(t) is fixed function which satisfies the conditions

(H1) v(t) ∈ C 3 [0, ∞), v(t) > 0 f or ∀t ∈ [0, 1];

(H2) ( v 00 (t) > 0 f or ∀t ∈ [0, 1], v 0 (1) = v 000 (1) = 0, v(1) 6= 0;
00 (t) 00 (t)
mint∈[0,1] vv(t) ≥ 2Aa 4 , maxt∈[0,1] vv(t) 2a
≤A 2;
(H3) 00 a
v (t) − 2A4 v(t) ≥ 0 f or t ∈ [0, 1].

Bellow we will prove that the equation (10 ) has unique nontrivial solution ω(r) forh which i
2
ω(r) ∈ C 2 [0, r1 ], ω(r) ∈ Ḣ 1 [0, r1 ], |ω(r)| ≤ AB for r ∈ [0, r1 ], ω(r) ≥ A12 for r ∈ α1 , β1 ,
ω(r1 ) = ω 0 (r1 ) = ω 00 (r1 ) = 0.
Example.
There exists function v(t) for which (H1)-(H3) are hold. Really, let us consider the
function
4A4
(t − 1)2 + a −1
(3) v(t) = A3
,
a
2
where the constants A and a satisfy the conditions A > 1, Aa > 1. Then
1) v(t) ∈ C 3 [0, ∞) and v(t) > 0 for all t ∈ [0, 1], i.e. (H1) is hold.
2)
v 0 (t) = 2(t−1)
A3
, v 0 (1) = 0,
a
2
v 00 (t) = A3
≥ 0 ∀ t ∈ [0, 1],
a
v 000 (t) = 0, v 000 (1) = 0,
consequently (H2) is hold. On the other hand we have

v 00 (t) 2
= 4A4
.
v(t) 2
(t − 1) + a −1

From here
v00 (t) a
mint∈[0,1] v(t) ≥ 2A4 ,
00 (t)
maxt∈[0,1] vv(t) 2a
≤A 2,

v (t) − 2A4 v(t) = 2A1 7

00 a
(2 − t)t,
a2

i.e. (H3) is hold.

2.1. Local existence of nontrivial solutions of homogeneous Cauchy prob-

lem (1), (2)
Let v(t) is fixed function which satisfies the hypothesis (H1) − (H3).

EJQTDE, 2007 No. 12, p. 6

 In this section we will prove that the homogeneous Cauchy problem (1), (2) has non-
trivial solution in the form
(
v(t)ω(r) f or r ≤ r1 , t ∈ [0, 1],
u(t, r) =
0 f or r ≥ r1 , t ∈ [0, 1].

Let us consider the integral equation

(?) ( R
r1 1 R r1  s4 v00 (t)

r τ 2 −Kτ +Q2 τ s2 −Ks+Q2 v(t)
u(t, s) − s2 f (u(t, s)) dsdτ, 0 ≤ r ≤ r1 , t ∈ [0, 1],
u(t, r) =
0 f or r ≥ r1 , t ∈ [0, 1],

where u(t, r) = v(t)ω(r).

Theorem 2.1. Let v(t) is fixed function which satisfies the hypothesis (H1)-(H3). Let
also p > 1, q ∈ [1, ∞) and γ ∈ (0, 1) are fixed and the positive constants A, a, b, B,
Q,K, α > β > 1 satisfy the conditions i1)-i6) and f ∈ C 1 (R1 ), f (0) = 0, a|u| ≤ f 0 (u) ≤
b|u|. Then the equation (?) has unique nontrivial solution u(t, r) = v(t)ω(r) for which
w ∈h C 2 [0,ir1 ], u(t, r) = ur (t, r) = urr (t, r) = h0 fori r ≥ r1 , u(t, r) ∈ C((0, 1]Ḣ 1 [0, r1 ]), for
r ∈ α1 , β1 and t ∈ [0, 1] u(t, r) ≥ A12 , for r ∈ 0, r1 and t ∈ [0, 1] |u(t, r)| ≤ AB 2
.
Proof. In [7, p. 303-309,theorem 3.1] is proved that the equation (?) has solution
u(t, r) in the form u(t, r) = v(t)ω(r) for which

u(t, r) ∈ C([0, 1] × [0, r1 ]);

u(t, r) = ur (t, r) = urr (t, r) = 0 f or r ≥ r1 and t ∈ [0, 1],
u(t, r) ∈ C((0, γ [0, r ]);
1]iḂp,q
h 1
f or r ∈ α1 , β1 and t ∈ [0, 1] u(t, r) ≥ A12 ;
h i
1
u(t, r) ≥ 0 f or t ∈ [0, 1] and r ∈ α , r1 ;
h i
2
f or r ∈ 0, r1 and t ∈ [0, 1] |u(t, r)| ≤ AB .

γ (M )-norm (γ ∈ (0, 1), p > 1, q ≥ 1) (see [3,

In [7] is used the following definition of the Ḃp,q
p.94, def. 2], [1])
Z 2  1q
||u||Ḃp,q
γ
(M ) =
−1−qγ
h ||∆h u||qLp (M ) dh ,
0
where
∆h u = u(x + h) − u(x).
Let t ∈ [0, 1] is fixed. Then

 ∂ 2
 
 ∂r u 2 =
L ([0,∞))
 R r1   2
R r1 1 s4 v00 (t)
= 0 r 2 −Kr+Q2 r s2 −Ks+Q2 v(t) u(t, s) − s2 f (u(t, s)) dsdτ dr ≤

EJQTDE, 2007 No. 12, p. 7

 s 4 1
here we use that from i5) we have that r1 < 1, s2 −Ks+Q 2 ≤ 1−K+Q2 for s ∈ [0, r1 ],
1 1
r 2 −Kr+Q2 ≤ 1−K+Q2 for r ∈ [0, r1 ](see [7, Remark, p.300])

r1  1 r1  1 v 00 (t) 2
Z Z 
2
≤ max |u(t, s)| + s |f (u(t, s))| dsdτ dr ≤
0 1 − K + Q2 r 1 − K + Q2 t∈[0,1] v(t)

here we use that f (0) = 0, |f (u)| ≤ 2b |u|2 ,

r1  1 r1  1 v 00 (t) 2b
2
Z Z 
2
≤ max |u(t, s)| + s |u(t, s)| dsdτ dr ≤
0 1 − K + Q2 r 1 − K + Q2 t∈[0,1] v(t) 2
2
here we use that |u(t, r)| ≤ AB for r ∈ [0, r1 ], t ∈ [0, 1],
r1  1 r1  1 v 00 (t) 2 2b 4  2
Z Z
≤ max + r 1 dsdτ dr ≤
0 1 − K + Q2 r 1 − K + Q2 t∈[0,1] v(t) AB 2 A2 B 2
v00 (t) 2a
here we use that maxt∈[0,1] v(t) ≤ A2 ,

R r1  R r1   2
1 1 2a 2
≤ 0 1−K+Q2 r 1−K+Q2 A2 AB
+ r12 2b A24B 2 dsdτ dr ≤
2
2r12 b
 
1 1 4a
≤ r13 1−K+Q 2 1−K+Q2 A3 B + A2 B 2 < ∞.

From here  ∂ 
 u <∞
 
∂r L2 ([0,∞))

for every fixed t ∈ (0, 1]. Therefore u(t, r) ∈ C((0, 1]Ḣ 1 ([0, ∞))). •
Let ũ is the solution from the theorem 2. 1, i.e ũ is the solution to the equation (?).
From proposition 2.1([7]) ũ satisfies the equation (1). Then ũ is solution to the Cauchy
problem (1), (2) with initial data
( R
r1 1 R r1  s4

r τ 2 −Kτ +Q2 τ s2 −Ks+Q2 v 00 (1)ω(s) − s2 f (v(1)ω(s)) dsdτ f or r ≤ r1 ,
u0 =
0 f or r ≥ r1 ,
( R
r1 1 R r1  s4

r τ 2 −Kτ +Q2 τ s2 −Ks+Q2 v 000 (1)ω(s) − s2 f 0 (u)v 0 (1)ω(s) dsdτ = 0 f or r ≤ r1 ,
u1 =
0 f or r ≥ r1 ,
u0 ∈ Ḣ 1 (R+ ), u1 ∈ L2 (R+ ), ũ ∈ C((0, 1]Ḣ 1 [0, r1 ]).

2.2. Uniformly continuity of the solution map for the homogeneous Cauchy
problem (1), (2)
Let v(t) is same function as in Theorem 2.1.
Theorem 2.2. Let p > 1, q ≥ 1 and γ ∈ (0, 1) are fixed and the positive constants
a, b, A, B, Q, K, 1 < β < α satisfy the conditions i1)-i6). Let f ∈ C 1 (R1 ), f (0) = 0,

EJQTDE, 2007 No. 12, p. 8

a|u| ≤ f 0 (u) ≤ b|u|. Then there exists t◦ ∈ [0, 1) for which there exists constant  > 0 such
that for every positive constant δ exist solutions u1 and u2 so that

E(1, u1 − u2 ) ≤ δ

and
E(t◦ , u1 − u2 ) ≥ .

Proof. Let us suppose that the solution map (u◦ , u1 , g) −→ u(t, r) is uniformly con-
tinuous.
Let
1 1 3  α2  a 2b 2
(4) 0<< − − .
β α 1 − αK + α2 Q2 2A6 α2 (1 − αL + α2 Q2 ) β 2 A2 B 2

Let also
u1 = ũ, u2 = 0.
Then there exists positive constant δ such that

E(1, u1 − u2 ) ≤ δ,

and
E(t, u1 − u2 ) ≤  f or ∀t ∈ [0, 1].
From here, for t ∈ [0, 1)
 2
∂
 ≥  ∂r ũ =
 
L2 ([0,∞))
 R r1   2
R r1 1 s4 v00 (t)
= 0 r −Kr+Q r
2 2 s2 −Ks+Q2 v(t) ũ(t, s) − s2 f (ũ(t, s)) dsdτ dr ≥
R β1  1 R r1  s4 v00 (t)
 2
≥ 1 r 2 −Kr+Q2 r s2 −Ks+Q2 v(t) ũ(t, s) − s2 f (ũ(t, s)) dsdτ dr ≥
α

h i
1 s4 v00 (t)
here we use that for s ∈ α , r1 and for t ∈ [0, 1] we have that s2 −Ks+Q2 v(t) ũ(t, s) −
1
s2 f (ũ(t, s)) ≥ 0(see [7, p. 305-306]) and r 2 −Kr+Q2 ≥ 0 for r ∈ [0, r1 ],

R β1  1 R β1  s4 v00 (t)
 2
≥ 1 r 2 −Kr+Q2 1 s2 −Ks+Q2 v(t) ũ(t, s) − s2 f (ũ(t, s)) dsdτ dr ≥
α α
1 1 2
R  1 R  s4 v00 (t)

≥ β
1 r 2 −Kr+Q2
β
1 s2 −Ks+Q2
mint∈[0,1] v(t) ũ(t, s) − s2 f (ũ(t, s)) dsdτ dr ≥
α α

v00 (t) a
from (H3) we have that mint∈[0,1] v(t) ≥ 2A4

1 1
1 s4 a 2
Z Z
β β
  
2
≥ ũ(t, s) − s f (ũ(t, s)) dsdτ dr ≥
1
α
r − Kr + Q2
2 1
α
s2 − Ks + Q2 2A4

EJQTDE, 2007 No. 12, p. 9

here we use that f (0) = 0, f (ũ) ≤ 2b ũ2
1 1
1 s4 a b 2
Z Z
β β
  
≥ ũ(t, s) − s2 ũ2 (t, s) dsdτ dr ≥
1
α
r − Kr + Q2
2 1
α
s2 2
− Ks + Q 2A 4 2
h i
1 1 1 4
here we use that ũ(t, s) ≥ A2
for t ∈ [0, 1] and s ∈ α, β , ũ2 (t, s) ≤ A2 B 2
for t ∈ [0, 1] and
s ∈ [0, r1 ]
1 1
1 s4 a 1 1 b 4  2
Z Z
β β
 
≥ − dsdτ dr ≥
1
α
r − Kr + Q2
2 1
α
s2 − Ks + Q2 2A4 A2 β 2 2 A2 B 2
h i
s2 1 1
here we use that s2 −Ks+Q2 is increase function for s ∈ [0, r1 ]. Therefore, for s ∈ α, β we
s4 1
have s2 −Ks+Q2 ≥ α2 (1−αK+α2 Q2 )
1 1
1 1 a 2b  2
Z Z
β β
 
≥ − dsdτ dr ≥
1
α
r 2 − Kr + Q2 1
α
α2 (1 − αK + α2 Q2 ) 2A6 β 2 A2 B 2
1
here we use that for r ∈ [0, r1 ] the function r 2 −Kr+Q2 is increase function. Therefore for
h i
1 1 1 α2
r∈ α, β we have r 2 −Kr+Q2 ≥ 1−αK+α2 Q2

1 1 3  α2  a 2b 2
≥ − −
β α 1 − αK + α2 Q2 2A6 α2 (1 − αL + α2 Q2 ) β 2 A2 B 2
which is contradiction with (4). •

3. Proof of Theorem 1.2.

3.1. Local existence of nontrivial solutions for nonhomogenious Cauchy
problem (1), (2)
Let v(t) is fixed function which satisfies the conditions (H1), (H2) and (H4), where
v 00 (t) a v 00 (t) 2a
(H4) min ≥ , max ≤ 2.
t∈[0,1] v(t) 4A4 t∈[0,1] v(t) A
For instance, the function
8A4
(t − 1)2 + a −1
v(t) = A3
,
a
satisfies the hypothesis (H1), (H2) and (H4).
Let us consider the equation
(?0 ) ( R
r1 1 R r1  s4 v00 (t)

2 f (u(t, s)) − s2 g(s) dsdτ, 0 ≤ r ≤ r ,
r τ −Kτ +Q τ
2 2 2 2
s −Ks+Q v(t)
u(t, s) − s 1
u(t, r) =
0 f or r ≥ r1 ,

EJQTDE, 2007 No. 12, p. 10

t ∈ [0, 1], where u(t, r) = v(t)ω(r).
Theorem 3.1. Let v(t) is fixed function which satisfies the hypothesis (H1), (H2),
(H4). Let also p > 1, q ∈ [1, ∞) and γ ∈ (0, 1) are fixed and the positive constants
A, a, b, B, Q,K, α > β > 1 satisfy the conditions i1)-i7) and f ∈ C 1 (R1 ), f (0) = 0,
a|u| ≤ f 0 (u) ≤ b|u|, g ∈ C(R+ ), g(r) ≥ 0 for ∀r ∈ R+ , g(r) = 0 for r ≥ r1 . Then
the equation (?0 ) has unique nontrivial solution u(t, r) = v(t)ω(r) for which w ∈ 2
h C [0,i r1 ],
1 1 1
u(t, r) = ur (t, r) = urr (t, r) = 0 for r ≥ r1 , u(t, r) ∈ C((0, 1]Ḣ [0, r1 ]), for r ∈ α , β and
h i
t ∈ [0, 1] u(t, r) ≥ A12 , for r ∈ 0, r1 and t ∈ [0, 1] |u(t, r)| ≤ AB
2
.
Proof. In [7, p. 313-316,theorem 4.1] is proved that the equation (?0 ) has solution
u(t, r) in the form u(t, r) = v(t)ω(r) for which

u(t, r) ∈ C([0, 1] × [0, r1 ]);

u(t, r) = ur (t, r) = urr (t, r) = 0 f or r ≥ r1 and t ∈ [0, 1],
u(t, r) ∈ C((0, γ [0, r ]);
1]iḂp.q
h 1
f or r ∈ α1 , β1 and t ∈ [0, 1] u(t, r) ≥ A12 ;
h i
1
u(t, r) ≥ 0 f or t ∈ [0, 1] and r ∈ α , r1 ;
h i
2
f or r ∈ 0, r1 and t ∈ [0, 1] |u(t, r)| ≤ AB .

Let t ∈ [0, 1] is fixed. Then

 ∂ 2
 
 ∂r u 2 =
L ([0,∞))
R r1  R r1   2
1 s4 v00 (t)
= 0 r 2 −Kr+Q2 r s2 −Ks+Q2 v(t) u(t, s) − s2 (f (u(t, s)) + g(s)) dsdτ dr ≤

s 4 1
here we use that from i5) we have that r1 < 1, s2 −Ks+Q 2 ≤ 1−K+Q2 for s ∈ [0, r1 ],
1 1
r 2 −Kr+Q2 ≤ 1−K+Q2 for r ∈ [0, r1 ](see [7, Remark, p.300])

r1  1 r1  1 v 00 (t) 2
Z Z 
2
≤ max |u(t, s)|+s (|f (u(t, s))|+g(s)) dsdτ dr ≤
0 1 − K + Q2 r 1 − K + Q2 t∈[0,1] v(t)

here we use that f (0) = 0, |f (u)| ≤ 2b |u|2 ,

r1  1 r1  1 v 00 (t) 2 b
2
Z Z 
2
≤ max |u(t, s)|+s ( |u(t, s)| +g(s)) dsdτ dr ≤
0 1 − K + Q2 r 1 − K + Q2 t∈[0,1] v(t) 2
2
here we use that |u(t, r)| ≤ AB for r ∈ [0, r1 ], t ∈ [0, 1], from i7) we have maxr∈[0,r1 ] g(r) ≤
 2
1 1 1
β − α A4

r1  1 r1  1 v 00 (t) 2 2b 4 2 1 1 2 1  2
Z Z 
≤ max +r +r − dsdτ dr ≤
0 1 − K + Q2 r 1 − K + Q2 t∈[0,1] v(t) AB 1 2 A2 B 2 1 β α A4

EJQTDE, 2007 No. 12, p. 11

 v00 (t) 2a
here we use that maxt∈[0,1] v(t) ≤ A2 ,

R r1  R r1   2  2
1 1 2a 2 2b 4 2 1 1 1
≤ 0 1−K+Q2 r 1−K+Q2 A2 AB + r1 2 A2 B 2 + r1 β − α A4 dsdτ dr ≤
2 2 2
2r1 b
  
1 1 4a 2 1 1 1
≤ r13 1−K+Q 2 1−K+Q2 A3 B + A2 B 2 + r1 β − α A4 < ∞.

From here  ∂ 
 u <∞
 
∂r L2 ([0,∞))

for every fixed t ∈ (0, 1]. Therefore u(t, r) ∈ C((0, 1]Ḣ 1 ([0, ∞))). •
Let ū is the solution from the theorem 3. 1. , i.e ū¯ is the solution to the equation (?0 ).
From proposition 2.3[7] we have that ū satisfies the equation (1). Then ū is solution to the
Cauchy problem (1), (2) with initial data
( R
r1 1 R r1  s4

¯0 = r τ 2 −Kτ +Q2 τ
00
s2 −Ks+Q2 v (1)ω(s) − s2 f (v(1)ω(s)) − s2 g(s) dsdτ f or r ≤ r1 ,
ū
0 f or r ≥ r1 ,
 Rr
1 R r1  s4

 r
 1
τ 2 −Kτ +Q2 τ s2 −Ks+Q2 v 000 (1)ω(s) − s2 f 0 (u)v 0 (1)ω(s) dsdτ ≡ 0
¯1 =
ū f or r ≤ r1 ,

0 f or r ≥ r1 ,


¯ 0 ∈ Ḣ 1 (R+ ), ū
ū ¯1 ∈ L2 (R+ ), ū
¯ ∈ C((0, 1]Ḣ 1 [0, r1 ]).

3.2. Uniformly continuity of the solution map for the nonhomogeneous

Cauchy problem (1), (2)
Let v(t) is same function as in Theorem 3.1.
Theorem 3.2. Let p > 1, q ≥ 1 and γ ∈ (0, 1) are fixed and the positive constants
a, b, A, B, Q, K, 1 < β < α satisfy the conditions i1)-i7). Let f ∈ C 1 (R1 ), f (0) = 0,
a|u| ≤ f 0 (u) ≤ b|u|, g ∈ C(R+ ), g(r) ≥ 0 for r ≥ 0, g(r) = 0 for r ≥ r1 . Then there exists
t◦ ∈ [0, 1) for which there exists constant  > 0 such that for every positive constants δ and
R exist solutions u1 , u2 of (1), (2) with right hands g = g1 , g = g2 of (1), so that

E(1, u1 − u2 ) ≤ δ, ||g1 ||L2 (R+ ) ≤ R, ||g2 ||L2 (R+ ) ≤ R, ||g1 − g2 ||L2 (R+ ) ≤ δ,

and
E(t◦ , u1 − u2 ) ≥ .

Proof. Let us suppose that the solution map (u◦ , u1 , g) −→ u(t, r) is uniformly con-
tinuous. Let
(5)
 1 1 3  α2  a 2b 2 1
 1 2 1 2
0<< − − −r 1 − .
β α 1 − αK + α2 Q2 4A6 α2 (1 − αK + α2 Q2 ) β 2 A2 B 2 β α A4

EJQTDE, 2007 No. 12, p. 12

Let also
u1 = ū, u2 = 0, g2 ≡ 0, g1 ≡ g,
where g is the function from theorem 3.1. Then there exist positive constants δ and R such
that
||g1 ||L2 (R+ ) ≤ R, ||g2 ||L2 (R+ ) ≤ R, ||g1 − g2 ||L2 (R+ ) ≤ δ,
E(1, u1 − u2 ) ≤ δ,
and
E(t, u1 − u2 ) ≤  f or ∀t ∈ [0, 1].
From here, for t ∈ [0, 1)
 2
∂
 ≥  ∂r ū =
 
L2 ([0,∞))
R r1  R r1   2
1 s4 v00 (t)
= 0 r 2 −Kr+Q2 r s2 −Ks+Q2 v(t)
ū(t, s) − s2 (f (ū(t, s)) + g(s)) dsdτ dr ≥
1 2
R  1 R r1 
s4 v00 (t)

≥ β
1 r 2 −Kr+Q2 r s2 −Ks+Q2 v(t) ū(t, s) − s2 (f (ū(t, s)) + g(s)) dsdτ dr ≥
α
h i
1
here we use that for r ∈ α , r1 we have

s4 v00 (t) 2 2
s2 −Ks+Q2 v(t) ū(t, s) − s f (ū(t, s)) − s g(s) ≥
2br12
a 2
≥ 4A6 α2 (1−αK+α 2 Q2 ) − A2 B 2 − r1 maxr∈[0,r1 ] g(r) ≥0
1
(see i5)) and for r ∈ [0, r1 ] we have r 2 −Kr+Q2 >0
R β1  1 R β1  s4 v00 (t)
 2
≥ 1 r 2 −Kr+Q2 1 s2 −Ks+Q2 v(t) ū(t, s) − s2 (f (ū(t, s)) + g(s)) dsdτ dr ≥
α α
1 1 2
R  1 R  s4 v00 (t)

≥ β
1 r 2 −Kr+Q2
β
1 s2 −Ks+Q2 mint∈[0,1] v(t) ū(t, s) − s2 (f (ū(t, s)) + g(s)) dsdτ dr ≥
α α

v00 (t) a
from (H4) we have that mint∈[0,1] v(t) ≥ 4A4
1 1
1 s4 a 2
Z Z
β β
  
2
≥ ū(t, s) − s (f (ū(t, s)) + g(s)) dsdτ dr ≥
1
α
r − Kr + Q2
2 1
α
s2 − Ks + Q2 4A4

here we use that f (0) = 0, f (ū) ≤ 2b ū2

1 1
1 s4 a 2b 2
2
Z Z
β β
  
2
≥ ū(t, s) − s ū (t, s) − s g(s) dsdτ dr ≥
1
α
r − Kr + Q2
2 1
α
s2 − Ks + Q2 4A4 2
h i
1 1 1 4
here we use that ū(t, s) ≥ A2 for t ∈ [0, 1] and s ∈ α, β , ū2 (t, s) ≤ A2 B 2 for t ∈ [0, 1] and
 2
1 1 1
s ∈ [0, r1 ], maxr∈[0,r1 ] g(r) ≤ β − α A4
,
1 1
1 s4 a 1 1 b 4 2 1 1 2 1  2
Z Z
β β
  
≥ − −r 1 − dsdτ dr ≥
1
α
r − Kr + Q2
2 1
α
s2 − Ks + Q2 4A4 A2 β 2 2 A2 B 2 β α A4

EJQTDE, 2007 No. 12, p. 13

 h i
s2 1 1
here we use that s2 −Ks+Q2 is increase function for s ∈ [0, r1 ]. Therefore, for s ∈ α, β we
s4 1
have s2 −Ks+Q2
≥ α2 (1−αK+α2 Q2 )

1 1
1 1 a 2b 2 1 1 2 1  2
Z Z
β β
  
≥ − −r − dsdτ dr ≥
1
α
r − Kr + Q2
2 1
α
α2 (1 − αK + α2 Q2 ) 4A6 β 2 A2 B 2 1 β α A4
1
here we use that for r ∈ [0, r1 ] the function r 2 −Kr+Q2 is increase function. Therefore for
h i
1 1 1 α2
r∈ α, β we have r 2 −Kr+Q2 ≥ 1−αK+α2 Q2

1 1 3  α2  a 2b 2 1
 1 2 1 2
≥ − − − r 1 −
β α 1 − αK + α2 Q2 4A6 α2 (1 − αK + α2 Q2 ) β 2 A2 B 2 β α A4

which is contradiction with (5). •

References
[1] Shatah, J., M. Struwe. Geometric wave equation. Courant lecture notes in mathematics
2(1998).

[2] Taylor, M. Partial differential equations III. Springer-Verlag, 1996.

[3] Runst, T., W. Sickel. Sobolev spaces of fractional order, Nemytskij operators and
nonlinear partial differential equations, 1996, New York.

[4] Rauch, J. The u5 − Klein - Gordon equation, Brezis. and Lions. eds., Pilman Research
Notes in Math. 53, pp. 335-364.

[5] Jörgens, K. Das Anfangswertproblem im Grossen für eine Klasse nicht-linear Wellen-
gleichungen, Math. Z., 77, 1961, pp. 295-308.

[6] John, F. Blow up of solutions of nonlinear wave equations in three space dimensions,
Manuscripta Math. 28, 1979, 235-268.

[7] Georgiev, S. Blow - up of the solutions of nonlinear wave equation in Reissner -

Nordström metric, Dynamics of PDE, Vol. 3, No. 4, 295-329, 2006.

(Received April 23, 2007)

EJQTDE, 2007 No. 12, p. 14