Electronic Journal of Differential Equations, Vol. 2009(2009), No. 89, pp. 1–12. ISSN: 1072-6691. URL: http://ejde.math.txstate.edu or http://ejde.math.unt.edu ftp ejde.math.txstate.edu EXISTENCE OF MULTIPLE POSITIVE SOLUTIONS FOR THIRD-ORDER BOUNDARY-VALUE PROBLEM ON THE HALF-LINE WITH DEPENDENCE ON THE FIRST ORDER DERIVATIVE SIHUA LIANG, JIHUI ZHANG Abstract. By using fixed-point theorem for operators on a cone, sufficient conditions for the existence of multiple positive solutions for a third-order boundary-value problem on the half-line are established. In the case of the pLaplace operator our results for p > 1 generalize previous known results. The interesting point lies in the fact that the nonlinear term is allowed to depend on the first order derivative u0 . 1. Introduction Boundary value problems on the half-line arise naturally in the study of radially symmetric solutions of nonlinear elliptic equations, see [3], and various physical phenomena [2, 5], such as an unsteady flow of gas through a semi-infinite porous media, the theory of drain flows, plasma physics, in determining the electrical potential in an isolated neutral atom. In recently years, the boundary-value problems on the half-line have received a great deal of attention in literature (see [1, 7, 10, 11, 15, 16] and references therein). However, in [6, 13, 14] the authors only studied multi-point boundary-value problems on the finite interval. They showed that there exist multiple positive solutions by using fixed-point theorems for operators on a cone. But so far, very few results are obtained for a third-order multi-point boundary-value problems on the half-line. To the author’s knowledge, there are no results about third order multi-point boundary-value problems, whose non- linear term does not depend on the first derivative u0 . The goal of this paper is to fill the gap in this area. In this paper, by using fixed-point theorem for operators on a cone, some sufficient conditions for the existence of multiple positive solutions for third-order boundary-value problem on the half-line are established, which are the complement of previously known results. In the case of the p-Laplace operator our results for some p > 1 generalize previous known results. 2000 Mathematics Subject Classification. 34B18, 34B40. Key words and phrases. Third-order boundary-value problems; positive solutions; half-line; fixed-point index theory. c 2009 Texas State University - San Marcos. Submitted March 26, 2009. Published July 27, 2009. 1 2 S. LIANG, J. ZHANG EJDE-2009/89 In this paper, we study the existence of multiple positive solutions for the following third-order boundary-value problem with dependence on the first order derivative on the half-line (ϕ(−u00 )(t))0 = a(t)f (t, u, u0 ), 0 < t < +∞, u(0) − βu0 (0) = 0, u0 (∞) = 0, (1.1) u00 (0) = 0, where ϕ : R → R is an increasing homeomorphism and positive homomorphism with ϕ(0) = 0 and f ∈ C([0, +∞)3 , [0, +∞)) and β ∈ (0, +∞). a(t) is a nonnegative measurable function defined in (0, +∞) and a(t) does not identically vanish on any subinterval of (0, +∞). A projection ϕ : R → R is called an increasing homeomorphism and positive homomorphism (see [8]), if the following conditions are satisfied: (i) if x ≤ y, then ϕ(x) ≤ ϕ(y), for all x, y ∈ R; (ii) ϕ is a continuous bijection and its inverse mapping is also continuous; (iii) ϕ(xy) = ϕ(x)ϕ(y), for all x, y ∈ [0, +∞). In above definition, we can replace the condition (iii) by the following stronger condition: (iv) ϕ(xy) = ϕ(x)ϕ(y), for all x, y ∈ R, where R = (−∞, +∞). Remark 1.1. (1) If conditions (i), (ii) and (iv) hold, then it implies that ϕ is homogenous generating a p-Laplace operator; i.e., ϕ(x) = |x|p−2 x, for some p > 1. (2) It is well known that a p-Laplacian operator is odd. However, the operator which we defined above is not necessary odd, see (5.2). We emphasize that the results of the papers [10, 12, 13, 14] cannot be applied if ϕ is defined as above. (3) The nonlinear term is allowed to depend on the first order derivative u0 which is the complement of previously known results [6, 7, 13, 14]. In this article, the following hypotheses are needed: (C1) f ∈ C([0, +∞)3 , [0, +∞)), f (t, 0, 0) 6≡ 0 on any subinterval of (0, +∞) and when u is bounded f (t, (1 + t)u, u0 ) is bounded on [0, +∞); (C2) a(t) satisfies the following relations: Z +∞ Z +∞ Z s Z s −1 −1 ϕ a(τ )dτ ds < +∞, sϕ a(τ )dτ ds < +∞. 0 0 0 0 The plan of the article is as follows. In Section 2 for the convenience of the reader we give some background and definitions. In Section 3 we present some lemmas in order to prove our main results. Section 4 is devoted to presenting and proving our main results. Some examples are presented in Section 5 to demonstrate the application of our main results. 2. Some definitions and fixed point theorems In this section, we provide background definitions from the cone theory in Banach spaces. Definition 2.1. Let (E, k · k) be a real Banach space. A nonempty, closed, convex set P ⊂ E is said to be a cone provided the following are satisfied: (a) if y ∈ P and λ ≥ 0, then λy ∈ P ; (b) if y ∈ P and −y ∈ P , then y = 0. EJDE-2009/89 EXISTENCE OF MULTIPLE POSITIVE SOLUTIONS 3 If P ⊂ E is a cone, we denote the order induced by P on E by ≤, that is, x ≤ y if and only if y − x ∈ P . Definition 2.2. A map α is said to be a nonnegative, continuous, concave functional on a cone P of a real Banach space E, if α : P → [0, ∞) is continuous, and α(tx + (1 − t)y) ≥ tα(x) + (1 − t)α(y) for all x, y ∈ P and t ∈ [0, 1]. Definition 2.3. An operator is called completely continuous if it is continuous and maps bounded sets into precompact sets. The following fixed point theorems are fundamental and important to the proofs of our main results. Theorem 2.4. [4] Let E be a Banach space and P ⊂ E be T a cone in E. Let r > 0 define Ωr = {x ∈ P : kxk < r}. Assume that T : P Ωr → P is completely continuous operator such that T x 6= x for x ∈ ∂Ωr . (i) If kT xk ≤ kxk for x ∈ ∂Ωr , then i(T, Ωr , P ) = 1. (ii) If kT xk ≥ kxk for x ∈ ∂Ωr , then i(T, Ωr , P ) = 0. Theorem 2.5. [6] Let K be a cone in a Banach space X. Let D be an open bounded set with Dk = D ∩ K 6= ∅ and Dk 6= K. Let T : Dk → K be a compact map such that x 6= T x for x ∈ ∂Dk . Then the following results hold: (1) If kT xk ≤ kxk for x ∈ ∂Dk , then ik (T, Dk ) = 1. (2) Suppose there is e ∈ K, e 6= 0 such that x 6= T x + λe for all x ∈ ∂Dk and all λ > 0, then ik (T, Dk ) = 0. (3) Let D1 be open in X such that D1 ⊂ Dk . If ik (T, Dk ) = 1 and ik (T, Dk1 ) = 0, then T has a fixed point in Dk \Dk1 . Then same result holds if ik (T, Dk ) = 0 and ik (T, Dk1 ) = 1. 3. Preliminaries and Lemmas Let E be the set defined as E = u ∈ C 1 [0, +∞) : |u(t)| < +∞, 0≤t<+∞ 1 + t sup lim u0 (t) = 0. t→+∞ Then E is a Banach space, equipped with the norm kuk = kuk1 + kuk2 , where 0 kuk1 = sup0≤t<+∞ |u(t)| 1+t < +∞, kuk2 = sup0≤t<+∞ |u (t)|. Also, define the cone K ⊂ E by K = u ∈ E : u(t) ≥ 0, t ∈ [0, +∞), u(0) − βu0 (0) = 0, u(t) is concave on [0, +∞). To prove the main results in this paper, we will employ several lemmas. Lemma 3.1. For any p ∈ C[0, +∞), the problem (ϕ(−u00 (t)))0 = p(t), 0 u(0) − βu (0) = 0, 0 0 < t < +∞, u (∞) = 0, 00 u (0) = 0 (3.1) (3.2) 4 S. LIANG, J. ZHANG EJDE-2009/89 has a unique solution Z +∞ Z +∞ Z s Z s −1 −1 p(τ )dτ ds sϕ p(τ )dτ ds + (t − s)ϕ u(t) = 0 0 0 t Z +∞ Z s p(τ )dτ ds. +β ϕ−1 (3.3) 0 0 Proof. Necessity. By taking the integral of the equation (3.1) on [0, t], we have Z t p(τ )dτ. (3.4) ϕ(−u00 (t)) − ϕ(−u00 (0)) = 0 By the boundary condition and ϕ(0) = 0, we have Z t 00 −1 p(τ )dτ . u (t) = −ϕ (3.5) 0 By taking the integral of (3.5) on [t, +∞), we obtain Z +∞ Z s u0 (∞) − u0 (t) = − ϕ−1 p(τ )dτ ds. t By the boundary condition u0 (∞) = 0, we obtain Z +∞ Z s u0 (t) = ϕ−1 p(τ )dτ ds. t (3.7) 0 By taking the integral of (3.7) on [t, +∞), we have Z +∞ Z +∞ Z τ u(t) = ϕ−1 p(η)dη dτ ds + u(0). t (3.6) 0 s (3.8) 0 Substituting u(0) = βu0 (0) and integrating by parts, we obtain Z +∞ Z +∞ Z s Z s −1 −1 u(t) = (t − s)ϕ p(τ )dτ ds + sϕ p(τ )dτ ds t 0 0 0 Z +∞ Z s +β ϕ−1 p(τ )dτ ds. 0 0 Sufficiency: Let u be as in (3.3). Taking the derivative of (3.3), it implies that Z +∞ Z s u0 (t) = ϕ−1 p(τ )dτ ds. t 0 Furthermore, we obtain 00 u (t) = −ϕ −1 Z ϕ(−u00 (t)) = Z t p(τ )dτ , 0 t p(τ )dτ, 0 taking the derivative of this expression yields (ϕ(−u00 (t)))0 = p(t). Routine calculation verifies that u satisfies the boundary value conditions, so that u given in (3.3) is a solution of (3.1)-(3.2). It is easy to see that the problem (ϕ(−u00 (t)))0 = 0, u(0) − βu0 (0) = 0, u0 (∞) = 0, u00 (0) = 0, has only the trivial solution. Thus u in (3.3) is the unique solution of (3.1)-(3.2). The proof is complete. EJDE-2009/89 EXISTENCE OF MULTIPLE POSITIVE SOLUTIONS 5 Lemma 3.2. For any p(t) ∈ C[0, +∞) and p(t) ≥ 0, the unique solution u of (3.1)-(3.2) satisfies u(t) ≥ 0, for t ∈ [0, +∞). Lemma 3.3. For any u ∈ K, it holds that βkuk2 ≤ kuk1 ≤ µkuk2 , where µ = max{β, 1}. Proof. Since u(t) is concave and nondecreasing, together with u0 (∞) = 0, we have kuk2 = u0 (0) and Z t+β u(t) 1 t 0 u (s)ds + βu0 (0) ≤ ≤ u0 (0) ≤ µkuk2 . 1+t 1+t 0 1+t On the other hand, u(0) |u(t)| ≥ = βu0 (0) = βkuk2 . 1+0 0≤t<+∞ 1 + t kuk1 = sup So we can obtain the desired result. Lemma 3.4. Let u ∈ K. Then mint∈[1/a,a] u(t) ≥ δ(t)kuk, where a > 1, δ(t) = 1 2 min{λ(t), β/a}, ( σ, t ≥ σ, λ(t) = t, t ≤ σ, and σ = inf{ξ ∈ [0, +∞) : supt∈[0,+∞) |u(t)| 1+t = u(ξ) 1+ξ }. Proof. From [7, Lemma 3.2], we know that min u(t) ≥ λ(t)kuk1 . 1 ,a] t∈[ a On the other hand, since u ∈ K, we have 1 β β min u(t) = u( ) ≥ u0 (0) = kuk2 . 1 a a a t∈[ a ,a] So (3.9) and (3.10) imply that the result of Lemma 3.4 holds. (3.9) (3.10) Remark 3.5. From the definition of λ(t), we know that 0 < δ(t) < 1, for t ∈ (0, 1). Define T : K → E by Z (T u)(t) = +∞ (t − s)ϕ −1 s Z t a(τ )f (τ, u(τ ), u0 (τ ))dτ ds 0 +∞ Z sϕ−1 + s Z 0 a(τ )f (τ, u(τ ), u0 (τ ))dτ ds 0 Z +∞ +β ϕ−1 Z 0 s a(τ )f (τ, u(τ ), u0 (τ ))dτ ds. 0 Obviously, we have (T u)(t) ≥ 0, for t ∈ (0, +∞), and Z +∞ Z s 0 −1 (T u) (t) = ϕ a(τ )f (τ, u(τ ), u0 (τ ))dτ ds ≥ 0. t 0 Furthermore, 00 (T u) (t) = −ϕ −1 Z 0 t a(s)f (s, u(s), u0 (s))ds ≤ 0, (3.11) 6 S. LIANG, J. ZHANG EJDE-2009/89 and (T u)(0) − β(T u)0 (0) = 0. This shows (T K) ⊂ K. To obtain the complete continuity of T , the following lemma is needed. Lemma 3.6 ([9]). Let W be a bounded subset of K. Then W is relatively compact in 0 E if { v(t) 1+t , v ∈ W } and {v (t), v ∈ W } are equicontinuous on any finite subinterval of [0, +∞) and for any ε > 0, there exists T = T (ε) > 0 such that v(t1 ) v(t2 ) < ε, |v 0 (t1 ) − v 0 (t2 )| < ε − 1 + t1 1 + t2 uniformly with respect to v ∈ W as t1 , t2 ≥ T . Lemma 3.7. Let (C1), (C2) hold. Then T : K → K is completely continuous. Proof. Firstly, it is easy to check that T : K → K is well defined. Let un → u in K, then from the definition of E, we can choose r0 such that supn∈N \{0} kun k < r0 . Let Ar0 = sup{f (t, (1 + t)u, v), (t, u, v) ∈ [0, +∞) × [0, r0 ]2 } and we have Z +∞ Z s ϕ−1 a(τ )|f (τ, un , u0n ) − f (τ, u, u0 )|dτ ds t 0 Z +∞ Z s ≤ 2ϕ−1 (Ar0 ) ϕ−1 a(τ )dτ ds. 0 0 Therefore, by the Lebesgue dominated convergence theorem, Z +∞ Z s 0 0 −1 |(T un ) (t) − (T u) (t)| = ϕ a(τ )(f (τ, un , u0n ) − f (τ, u, u0 ))dτ ds t 0 Z +∞ Z s ≤ ϕ−1 a(τ )|f (τ, un , u0n ) − f (τ, u, u0 )|dτ ds t → 0, 0 as n → +∞. From Lemma 3.3, we have kT un − T uk ≤ (1 + µ)kT un − T uk2 → 0, as n → +∞. Thus T is continuous. Let Ω be any bounded subset of K. Then there exists r > 0 such that kuk ≤ r for all u ∈ Ω and we have Z +∞ Z s −1 kT uk2 = ϕ a(τ )f (τ, u, u0 )dτ ds 0 0 Z +∞ Z s −1 −1 ≤ ϕ (Ar ) ϕ a(τ )dτ ds < +∞. 0 0 From Lemma 3.3, we have kT uk ≤ (1 + µ)kT uk2 < +∞. So T Ω is bounded. Moreover for any S ∈ (0, +∞) and t1 , t2 ∈ [0, S]. Without loss of generality, let t1 ≥ t2 . Then we have Z Z (T u)(t1 ) (T u)(t2 ) h +∞ −1 s ≤ − sϕ a(τ )f (τ, u, u0 )dτ ds 1 + t1 1 + t2 0 0 Z +∞ i 1 Z s 1 +β ϕ−1 a(τ )f (τ, u, u0 )dτ ds − 1 + t1 1 + t2 0 0 Z +∞ Z s t t2 1 + ϕ−1 a(τ )f (τ, u, u0 )dτ ds − 1 + t1 1 + t2 t2 0 EJDE-2009/89 EXISTENCE OF MULTIPLE POSITIVE SOLUTIONS Z +∞ t1 Z s 1 1 − 1 + t1 1 + t2 0 Z t1 Z s a(τ )f (τ, u, u0 )dτ ds sϕ−1 sϕ−1 + 7 a(τ )f (τ, u, u0 )dτ ds 1 1 + t2 t2 0 → 0, uniformly as t1 → t2 + and |(T u)0 (t2 ) − (T u)0 (t1 )| = Z t2 ϕ−1 s Z a(τ )f (τ, u(τ ), u0 (τ ))dτ ds → 0, 0 t1 uniformly as t1 → t2 . We obtain that T Ω is equicontinuous on any finite subinterval of [0, +∞). For any u ∈ Ω, we have Z Z s (T u)(t) 1 n +∞ = lim lim (t − s)ϕ−1 a(τ )f (τ, u(τ ), u0 (τ ))dτ ds t→+∞ t→+∞ 1 + t 1+t t 0 Z +∞ Z s + sϕ−1 a(τ )f (τ, u(τ ), u0 (τ ))dτ ds 0 0 Z +∞ Z s o +β ϕ−1 a(τ )f (τ, u(τ ), u0 (τ ))dτ ds 0 0 Z +∞ Z s ≤ lim ϕ−1 (Ar ) ϕ−1 a(τ )dτ ds t→+∞ t 0 =0 and lim |(T u)0 (t)| = lim ϕ−1 (Ar ) t→+∞ t→+∞ Z +∞ ϕ−1 Z t s a(τ )dτ ds = 0. 0 So T Ω is equiconvergent at infinity. By Lemma 3.6, T Ω is relatively compact. Therefore we know that T is compact. So T : K → K is completely continuous. The proof is complete. Let k > 1 be a fixed constant and choose a = k. Then define R Rk s 1 −1 δ β ϕ a(τ )dτ ds k 1/k 1/k 1 R , γ = δ( ) R k (1 + µ) +∞ ϕ−1 s a(τ )dτ ds 0 0 R R k s δ k1 β 1/k ϕ−1 1/k a(τ )dτ ds R , γ1 = R +∞ s (1 + µ) 0 ϕ−1 0 a(τ )dτ ds Kρ = {u ∈ K : kuk1 ≤ ρ}, u(t) < γρ} = {u ∈ K : γkuk1 ≤ min u(t) < γρ}. Ωρ = {u ∈ K : min 1 1 t∈[ k ,k] t∈[ k ,k] Lemma 3.8 ([6]). The set Ωρ has the following properties: (a) Ωρ is open relative to K. (b) Kγρ ⊂ Ωρ ⊂ Kρ . (c) u ∈ ∂Ωρ if and only if mint∈[ k1 ,k] u(t) = γρ. (d) u ∈ ∂Ωρ , then γρ ≤ u(t) ≤ ρ for t ∈ [ k1 , k]. 8 S. LIANG, J. ZHANG EJDE-2009/89 Now, we introduce the following notation. Let f (t, (1 + t)u, v) ρ fγρ = min : t ∈ [1/k, k], u ∈ [γρ, ρ], v ∈ [0, ρ/β] , ϕ(ρ) f (t, (1 + t)u, v) f0ρ = sup : t ∈ [0, +∞), u ∈ [0, ρ], v ∈ [0, ρ/β] , ϕ(ρ) f (t, (1 + t)u, v) : t ∈ [0, +∞) , f α = lim sup u→α ϕ(u) f (t, (1 + t)u, v) fα = lim min : t ∈ [1/k, k] (α := ∞ or 0+ ), u→α ϕ(u) Z +∞ Z s 1 a(τ )dτ ds, = (1 + µ) ϕ−1 m 0 0 Z k Z s 1 1 ϕ−1 a(τ )dτ ds. =δ β M k 1/k 1/k Remark 3.9. It is easy to see that 0 < m, M < ∞ and M γ = M γ1 δ( k1 ) = δ( k1 )m < m. Lemma 3.10. If f satisfies the condition f0ρ ≤ ϕ(m) and u 6= T u f or u ∈ ∂Kρ , (3.12) then ik (T, Kρ ) = 1. Proof. By (3.11) and (3.12), for u(t) ∈ ∂Kρ , we have kuk1 = sup0≤t<+∞ |u(t)| 1+t = ρ. Moreover, Lemma 3.3 implies that kuk2 ≤ β1 kuk1 = β1 ρ, Therefore, from definition of f0ρ we have f (t, u, u0 ) ≤ ϕ(ρ)ϕ(m) = ϕ(ρm). Therefore, kT uk = kT uk1 + kT uk2 ≤ (1 + µ)kT uk2 Z +∞ Z s = (1 + µ) ϕ−1 a(τ )f (τ, u(τ ), u0 (τ ))dτ ds 0 0 Z +∞ Z s ≤ ρm(1 + µ) ϕ−1 a(τ )dτ ds 0 0 ≤ ρ = kuk1 ≤ kuk. This implies that kT uk ≤ kuk for u(t) ∈ ∂Kρ . By Theorem 2.4 (1) we have ik (T, Kρ ) = 1. Lemma 3.11. If f satisfies the conditions ρ fγρ ≥ ϕ(M γ) and u 6= T u f or u ∈ ∂Ωρ , (3.13) then ik (T, Ωρ ) = 0. Proof. Let e(t) ≡ 1 for t ∈ [0, +∞). Then e ∈ ∂K1 , and we claim that u 6= T u + λe, u ∈ ∂Ωρ , λ > 0. If not, there exist u0 ∈ ∂Ωρ and λ0 > 0 such that u0 = T u0 + λ0 e. By (3.11) and (3.13) we have u0 = T u0 (t) + λ0 e EJDE-2009/89 EXISTENCE OF MULTIPLE POSITIVE SOLUTIONS 9 1 kT u0 k1 + λ0 k |T u0 | 1 sup + λ0 =δ k t∈[0,+∞) 1 + t 1 (T u)(0) ≥δ + λ0 k 1+0 Z +∞ Z s 1 =δ β ϕ−1 a(τ )f (τ, u(τ ), u0 (τ ))dτ ds + λ0 k 0 0 Z k Z s 1 −1 a(τ )f (τ, u(τ ), u0 (τ ))dτ ds + λ0 ϕ β ≥δ k 1/k 1/k Z k Z s 1 ≥δ a(τ )dτ ds + λ0 ϕ−1 M γρβ k 1/k 1/k ≥ γρ + λ0 . ≥δ This implies that γρ ≥ γρ + λ0 which is a contradiction. Hence by Theorem 2.4 (2), we have ik (T, Ωρ ) = 0. 4. Main results The main results in this articles are the following. Theorem 4.1. Assume that one of the following conditions holds: (C3) There exist ρ1 , ρ2 , ρ3 ∈ (0, ∞) with ρ1 < γρ2 and ρ2 < ρ3 such that f0ρ1 ≤ ϕ(m), ρ2 fγρ ≥ ϕ(M γ), 2 u 6= T u for u ∈ ∂Ωρ2 and f0ρ3 ≤ ϕ(m). (C4) There exist ρ1 , ρ2 , ρ3 ∈ (0, ∞) with ρ1 < ρ2 < γρ3 such that ρ1 fγρ ≥ ϕ(M γ), 1 f0ρ2 ≤ ϕ(m), ρ3 u 6= T u for u ∈ ∂Kρ2 and fγρ ≥ ϕ(M γ). 3 Then (1.1) has two positive solutions in K. Moreover if in (C3) f0ρ1 ≤ ϕ(m) is replaced by f0ρ1 < ϕ(m), then (1.1) has a third positive solution u3 ∈ Kρ1 . Proof. The proof is similar to the one in [6, Theorem 2.10]. We omit it here. As a special case of Theorem 4.1 we obtain the following result. Corollary 4.2. If there exists ρ > 0 such that one of the following conditions holds: ρ (C5) 0 ≤ f 0 < ϕ(m), fγρ ≥ ϕ(M γ), u 6= T u for u ∈ ∂Ωρ and 0 ≤ f ∞ < ϕ(m), ρ (C6) ϕ(M ) < f0 ≤ ∞, f0 ≥ ϕ(m), u 6= T u for u ∈ ∂Kρ and ϕ(M ) < f∞ ≤ ∞. Then (1.1) has two positive solutions in K. Proof. We show that (C5) implies (C3). It is easy to verify that 0 ≤ f 0 < ϕ(m) implies that there exist ρ1 ∈ (0, γρ) such that f0ρ1 < ϕ(m). Let a ∈ (f ∞ , ϕ(m)). Then there exists r > ρ such that supt∈[0,+∞) f (t, (1+t)u, v) ≤ aϕ(u) for u ∈ [r, ∞) since 0 ≤ f 0 < ϕ(m). Let r β = max sup f (t, (1 + t)u, v) : 0 ≤ u ≤ r, 0 ≤ v ≤ β t∈[0,+∞) and ρ3 > ϕ−1 β . ϕ(m) − a 10 S. LIANG, J. ZHANG EJDE-2009/89 Then sup f (t, (1 + t)u, v) ≤ aϕ(u) + β ≤ aϕ(ρ3 ) + β < ϕ(m)ϕ(ρ3 ) t∈[0,+∞) for u ∈ [0, ρ3 ]. This implies that f0ρ3 < ϕ(m) and (C3) holds. Similarly, (C6) implies (C4). By an argument similar to that of Theorem 4.1 we obtain the following result. Theorem 4.3. Assume that one of the following conditions holds: ρ2 (C7) There exist ρ1 , ρ2 ∈ (0, ∞) with ρ1 < γρ2 such that f0ρ1 ≤ ϕ(m) and fγρ ≥ 2 ϕ(M γ). ρ1 (C8) There exist ρ1 , ρ2 ∈ (0, ∞) with ρ1 < ρ2 such that fγρ ≥ ϕ(M γ) and 1 ρ2 f0 ≤ ϕ(m). Then (1.1) has a positive solution in K. As a special case of Theorem 4.3 we obtain the following result. Corollary 4.4. If there exists ρ > 0 such that one of the following conditions holds: ρ ≥ ϕ(M γ), u 6= T u for u ∈ ∂Ωρ and 0 ≤ f ∞ < ϕ(m), (C5) 0 ≤ f 0 < ϕ(m), fγρ ρ (C6) ϕ(M ) < f0 ≤ ∞, f0 ≥ ϕ(m), u 6= T u for u ∈ ∂Kρ and ϕ(M ) < f∞ ≤ ∞. Then (1.1) has two positive solutions in K. 5. Examples As an example we mention the boundary-value problem (ϕ(−u00 )(t))0 = a(t)f (t, u, u0 ), 0 < t < +∞, 0 u(0) − u (0) = 0, 0 u (∞) = 0, where ( ϕ(u) = and u (0) = 0, u3 1+u2 , 2 u 9 1+t 2 9 1+t We choose k = 2, β = 1, δ(t) = t/2 and so Z +∞ Z Z s −1 ϕ a(τ )dτ ds = 4, 0 0 u ≤ 0, u > 0, u ( 10−5 | sin t| + f (t, u, v) = 10−5 | sin t| + (5.1) 00 + + 1 v 100 1000 , 1 v 100 1000 , 2 ϕ 1 2 (5.2) −1 Z s u ≤ 2, u ≥ 2. a(τ )dτ ds = 2. 1 2 It is easy to see by calculating that µ = 1, γ = 1/64, γ1 = 1/16 and Z +∞ Z s 1 = (1 + µ) ϕ−1 a(τ )dτ ds = 8, m 0 0 Z 2 Z s 1 1 1 =δ β ϕ−1 a(τ )dτ ds = . 1 1 M 2 2 2 2 Thus m = 1/8, M = 2 and let ρ1 = 1/2, ρ2 = 128, ρ3 = 320. After some simple calculation we have 1 1 1 f (t, (1 + t)u, u0 ) ≤ 10−5 + + × 10−5 < = ϕ(mρ1 ) = ϕ(m)ϕ(ρ1 ), 512 2 256 EJDE-2009/89 EXISTENCE OF MULTIPLE POSITIVE SOLUTIONS 11 for all (t, u, u0 ) ∈ [0, +∞) × [0, 1/2] × [0, 12 ]. Therefore, f0ρ1 < ϕ(m). On the other hand, f (t, (1 + t)u, u0 ) ≥ 29 = 512 > 16 = ϕ(M γρ2 ) = ϕ(M )ϕ(γρ2 ), ρ2 for all (t, u, u0 ) ∈ [1/2, 2] × [2, 128] × [0, 128]. We have fγρ > ϕ(M γ). At last 2 f (t, (1 + t)u, u0 ) ≤ 10−5 + 29 + 320 × 10−5 < 513 < 1600 = ϕ(mρ3 ) = ϕ(m)ϕ(ρ3 ), for all (t, u, u0 ) ∈ [0, +∞) × [0, 320] × [0, 320]. Thus we have f0ρ3 < ϕ(m). Then the condition (C3) in Theorem 4.1 is satisfied. So boundary-value problem (5.1) has at least three positive solutions in K. Remark 5.1. From (5.2), we can see that ϕ is not odd, therefore the boundaryvalue problem with p-Laplacian operator [10, 12, 13, 14] do not apply to (5.2). So we generalize a p-Laplace operator for some p > 1 and the function ϕ which we defined above is more comprehensive and general than p-Laplace operator. Acknowledgments. The authors of this paper wish to thanks the referee for valuable suggestions regarding the original manuscript. This project is supported by grant NSFC(10871096), Foundation of Major Project of Science and Technology of Chinese Education Ministry, SRFDP of Higher Education, the project of Nanjing Normal University and Project of Graduate Education Innovation of Jiangsu Province (181200000214). References [1] R. P. Agarwal, D. O’Regan, Infinite Interval Problems for Differential, Difference and Integral Equations, Kluwer Academic, Dordrecht, 2001. [2] D. Aroson, M. G., Crandall, L.A. Peletier, Stabilization of solutions of a degenerate nonlinear diffusion problem, Nonlinear Anal. 6 (1982) 1001-1022. [3] J. V. Baxley, Existence and uniqueness of nonlinear boundary-value problems on infinite intervals, J. Math. Anal. Appl. 147 (1990) 122-133. [4] K. Deimling, Nonlinear Functional Analysis, Springer-Verlag, New York, 1985. [5] G. Iffland, Positive solutions of a problem Emden-Flower type with a type free boundary, SIAM J.Math. Anal. 18 (1987) 283-292. [6] K. Q. Lan, Multiple positive solutions of semilinear differential equations with singularities, J. London Math. Soc. 63 (2001) 690-704. [7] S. H. Liang, J. H. Zhang, The existence of countably many positive solutions for nonlinear singular m-point boundary-value problems on the half-line, J. Comput. Appl. Math., 222 (2008) 229-243. [8] B. F. Liu, J. H. Zhang, The existence of positive solutions for some nonlinear boundary-value problems with linear mixed boundary conditions, J. Math. Anal. Appl. 309(2005) 505-516. [9] Y. Sh. Liu, Boundary value probelm for second order differential equations on unbounded domain, Acta Anal. Funct. Appl. 4 (3) (2002) 211-216. (in Chinese) [10] H. Lian, H. H. Pan, W. G. Ge, Triple positive solutions for boundary-value problems on infinite intervals, Nonlinear Analysis 67 (2007) 2199-2207. [11] Y. S. Liu, Existence and unboundedness of positive solutions for singular boundary-value problems on half-line, Appl. Math. Comput. 1404(2003) 543-556. [12] J. L. Ren, W. G. Ge, B. X. Ren, Existence of positive solutions for quasi-linear boundaryvalue problems, Acta, Math. Appl. Sinica, Vol. 21, No. 3 (2005) 353-358. (in Chinese) [13] Y. Sang, H. Su, Positive solutions of nonlinear third-order m-point BVP for an increasing homeomorphism and homomorphism with sign-changing nonlinearity, J. Comput. Appl. Math., 225 (2009) 288-300. [14] Y. Y. Wang, C. Hou, Existence of multiple positive solutions for one-dimensional pLaplacian, J. Math. Anal. Appl. 315(2006) 144-153. [15] B. Q. Yan, Multiple unbounded solutions of boundary-value problems for second-order differential equations on the half-line, Nonlinear Anal. 51 (2002) 1031-1044. 12 S. LIANG, J. ZHANG EJDE-2009/89 [16] M. Zima, On positive solution of boundary-value problems on the half-line, J. Math. Anal. Appl. 259 (2001) 127-136. Sihua Liang Institute of Mathematics, School of Mathematical Sciences, Nanjing Normal University, 210097, Jiangsu, China College of Mathematics, Changchun Normal University, Changchun 130032, Jilin, China E-mail address: liangsihua@163.com Jihui Zhang Institute of Mathematics, School of Mathematical Sciences, Nanjing Normal University, 210097, Jiangsu, China E-mail address: jihuiz@jlonline.com