Electronic Journal of Differential Equations, Vol. 2015 (2015), No. 280, 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 POSITIVE SOLUTIONS FOR KIRCHHOFF PROBLEMS JIA-FENG LIAO, PENG ZHANG, XING-PING WU Abstract. We study problems for the Kirchhoff equation Z “ ” |∇u|2 dx ∆u = νu3 + Q(x)uq , in Ω, − a+b Ω u = 0, on ∂Ω, where Ω ⊂ R3 is a bounded domain, a, b ≥ 0 and a + b > 0, ν > 0, 3 < q ≤ 5 and Q(x) > 0 in Ω. By the mountain pass lemma, the existence of positive solutions is obtained. Particularly, we give a condition of Q to ensure the existence of solutions for the case of q = 5. 1. Introduction and main results In this article, we consider the Kirchhoff type problem Z − a+b |∇u|2 dx ∆u = νu3 + Q(x)uq , in Ω, Ω u = 0, (1.1) on ∂Ω, where Ω ⊂ R3 is a bounded domain, a, b ≥ 0 and a + b > 0, ν > 0, 3 < q ≤ 5 are four parameters. The coefficient function Q is a positive function in Ω. When a = 0, b > 0, problem (1.1) is called degenerate, and the case of a, b > 0 is called non-degenerate. When a ≥ 0 and b > 0, problem (1.1) is called the Kirchhoff type problem. Kirchhoff type problems R are often referred to as being nonlocal because of the presence of the term ( Ω |∇u|2 dx)∆u which implies that the equation in (1.1) is no longer a pointwise equation. The existence and multiplicity of solutions for the problem Z − a+b |∇u|2 dx ∆u = f (x, u), in Ω, (1.2) Ω u = 0, on ∂Ω, on a smooth bounded domain Ω ⊂ R3 and f : Ω × R → R a continuous function, has been extensively studied (see [1, 3],[7]-[23], [25]-[28],[30, 31]). 2010 Mathematics Subject Classification. 35B09, 35B33, 35J20. Key words and phrases. Kirchhoff type equation; resonance; positive solution; mountain pass lemma. c 2015 Texas State University. Submitted June 25, 2015. Published November 10, 2015. 1 2 J.-F. LIAO, P. ZHANG, X.-P. WU EJDE-2015/280 Particularly, Sun and Tang [26] considered the problem Z − a+b |∇u|2 dx ∆u = λu3 + g(u) − h(x), in Ω, Ω u = 0, (1.3) on ∂Ω, where h ∈ L2 (Ω) and g ∈ C(R, R) satisfies lim |t|→∞ g(t) = 0. t3 (1.4) Under a Landesman-Lazer type condition, by the minimax methods, they obtained the existence of solutions for problem (1.3). When a = 1 and b = 0, problem (1.1) reduces to the semilinear elliptic problem −∆u = νu3 + λuq , u = 0, in Ω, (1.5) on ∂Ω. Obviously when 3 < q < 5, problem (1.5) has a positive solution for all ν, λ > 0. While for q = 5, λ = 1, Brézis and Nirenberg [6] studied problem (1.5). By the variant of the mountain pass theorem of Ambrosetti and Rabinowitz without the (PS) condition, they obtained that there exists ν0 > 0 such that problem (1.5) has a positive solution for each ν ≥ ν0 . For u ∈ H01 (Ω), we define Z Z Z 2 ν Z b 1 a 2 2 4 |∇u| dx + |∇u| dx − |u| dx − Q(x)|u|q+1 dx, I(u) = 2 Ω 4 Ω 4 Ω q+1 Ω 1/2 R . where H01 (Ω) is a Sobolev space equipped with the norm kuk = Ω |∇u|2 dx Note that a function u is called a weak solution of (1.1) if u ∈ H01 (Ω) such that Z Z Z Z 3 2 (∇u, ∇ϕ)dx − ν u ϕdx − Q(x)uq ϕdx = 0, (1.6) a+b |∇u| dx Ω Ω Ω Ω for all ϕ ∈ H01 (Ω). We denote by ν1 is the first eigenfunction of the eigenvalue problem Z |∇u|2 dx ∆u = νu3 , x ∈ Ω, − Ω u = 0, x ∈ ∂Ω. From [23], we know that ν1 > 0. Let S be the best Sobolev constant, namely R |∇u|2 dx Ω S := inf (1.7) 1/3 . R u∈H01 (Ω)\{0} |u|6 dx Ω As well known, the function U (x) = (3ε2 )1/4 (ε2 + |x|2 ) 1/2 , x ∈ R3 , (1.8) is an extremal function for the minimum problem (1.7); that is, it is a positive solution of the problem −∆u = u5 , ∀x ∈ RN . Problem (1.1) with 3 < q ≤ 5 does not satisfy condition (1.4). It is natural to ask whether (1.1) has a positive solution. Using the mountain pass theorem, we study (1.1) and give a positive answer. It is worth pointing out that the result of the case EJDE-2015/280 KIRCHHOFF PROBLEMS 3 of q = 5 is more meaningful. Because of q = 5 is critical case and the coefficient of critical term is no longer constant. We give the suitable condition (A1) in the Theorem 1.3 below to ensure the existence of solutions to problem (1.1). Our main results are described as follows. 6 Theorem 1.1. Assume a, b > 0, 3 < q < 5 and Q ∈ L 5−q (Ω) is a positive function, then (1.1) possesses a positive solution u∗ for all ν > 0, and I(u∗ ) > 0. Remark 1.2. Obviously, Theorem 1.1 does not apply to (1.4). For all ν > 0, we obtain the existence of positive solutions for problem (1.1). For the degenerate case, that is a = 0, b > 0, we can also obtain that problem (1.1) possesses a positive solution for all 0 < ν < bν1 . Theorem 1.3. Assume a, b > 0, q = 5, Q ∈ C(Ω) is a positive function and satisfies the assumption (A1) There exists x0 ∈ Ω such that Q(x0 ) = QM = maxx∈Ω Q(x) and Q(x) − Q(x0 ) = o(|x − x0 |), as x → x0 . ∗ Then there exists ν > 0 such that (1.1) possesses a positive solution u∗ for all ν > ν ∗ , and I(u∗ ) > 0. Remark 1.4. This case is the critical exponent problem, and Theorem 1.3 does not apply to (1.4). When Q(x) ≡ 1, the Kirchhoff type problems with critical exponent have been considered by several papers, such as [1, 9, 12] [20]-[22], [27, 28, 30]. Particularly, problem (1.1) with Q(x) ≡ 1 was been considered in [21]. However, there exists a flaw in the proof of [21, Theorem 1.3] with the case θ = 4. To our best knowledge, problem (1.1) with Q(x) not constant has not been considered yet. When Q(x) is not constant, the analysis of the compactness becomes complicated, which results in much difficulty. It is worth pointing out that (A1) ensures the existence of solutions. Obviously, Theorem 1.3 extends the corresponding result of [21]. This article is organized as follows. In Section 2, we consider the case of 3 < q < 5 and prove Theorem 1.1 by the variational methods. We study the critical case of problem (1.1) with q = 5 and give the proof of Theorem 1.3 in Section 3. 2. The case 3 < q < 5 6 In this section, suppose that Q ∈ L 5−q (Ω) is a positive function and 3 < q < 5. We will prove Theorem 1.1 by the mountain pass theorem. Before proving Theorem 1.1, we give the following lemma. 6 Lemma 2.1. Assume a, b > 0, 3 < q < 5 and Q ∈ L 5−q (Ω) is a positive function, then the functional I satisfies the (PS)c condition for all ν > 0. Proof. Suppose that {un } is a (PS)c sequence of I, that is, I(un ) → c, I 0 (un ) → 0, H01 (Ω). (2.1) as n → +∞. We claim that {un } is bounded in In fact, from (2.1) one has 1 1 + c + o(1)kun k ≥ I(un ) − hI 0 (un ), un i 4 Z 1 a 1 2 − = kun k + Q(x)|un |q+1 dx 4 4 q+1 Ω 4 J.-F. LIAO, P. ZHANG, X.-P. WU ≥ EJDE-2015/280 a kun k2 . 4 Hence, we conclude that {un } is bounded in H01 (Ω). Going if necessary to a subsequence, still denoted by {un }, there exists u ∈ H01 (Ω) such that weakly in H01 (Ω), un * u, un → u, strongly in Ls (Ω), 1 ≤ s < 6, un (x) → u(x), (2.2) a.e. in Ω, as n → ∞. Now, we only need to prove that un → u as n → ∞ in H01 (Ω). As usually, letting wn = un − u, we need prove that kwn k → 0 as n → ∞. By the Vitali theorem (see [24, p.133]), we claim that Z Z lim Q(x)|un |q+1 dx = Q(x)|u|q+1 dx. (2.3) n→∞ Ω Ω Indeed, we only need to prove that { Ω Q(x)|un |p+1 dx, n ∈ N } is equi-absolutelycontinuous. Note that {un } is bounded in H01 (Ω), by the Sobolev embedding theorem, then exists a constant C > 0 such that |un |6 ≤ C < ∞. From the Hölder inequality, for every ε > 0, setting δ > 0, when E ⊂ Ω with meas E < δ, we have Z 5−q Z 6 6 q+1 q+1 < ε, Q(x)|un | dx ≤ |un |6 Q 5−q (x)dx R E E R 6 where the last inequality is from the absolutely-continuity of Ω Q 5−q (x)dx. Thus, our claim is proved. Moreover, one also has Z Z Z 2 2 |∇un | dx = |∇wn | dx + |∇u|2 dx + o(1), (2.4) Ω Ω Ω Z 2 |∇un |2 dx = kwn k4 + kuk4 + 2kwn k2 kuk2 + o(1). (2.5) Ω 0 Since I (un ) → 0, one obtains akun k2 + bkun k4 − ν Z |un |4 dx − Z Ω Q(x)|un |q+1 dx = o(1), Ω consequently, from (2.2)-(2.5), we deduce that 2 2 4 2 2 akwn k + akuk + bkwn k + 2bkwn k kuk + bkuk 4 − ν|u|44 − Z Q(x)|u|q+1 dx = o(1). Ω (2.6) From (2.1) it follows that lim hI 0 (un ), ui = akuk2 + bl2 kuk2 + bkuk4 − ν|u|44 − n→∞ Z Q(x)|u|q+1 dx = 0, (2.7) Ω where l = limn→∞ kwn k. According to (2.6) and (2.7), we have akwn k2 + bkwn k4 + bkwn k2 kuk2 = o(1), consequently, one has al2 + bl4 + bl2 kuk2 = 0. Thus l = 0; that is, un → u as n → ∞ in H01 (Ω). This completes the proof. EJDE-2015/280 KIRCHHOFF PROBLEMS 5 Proof of Theorem 1.1. The main idea is to construct a suitable geometry of mountain pass lemma (see[2]). Then obtain a critical point of I in H01 (Ω). We claim that I has the geometry of mountain pass lemma in H01 (Ω). Indeed, since Z Z a b ν 1 I(u) = kuk2 + kuk4 − |u|4 dx − Q(x)|u|q+1 dx, 2 4 4 Ω q+1 Ω then I(0) = 0, and for every u ∈ H01 (Ω)\{0} one has Z a 1 I(tu) I(tu) 2 = kuk , =− lim Q(x)|u|q+1 dx. lim t→+∞ tq+1 t→0+ t2 2 q+1 Ω R Since a > 0 and Ω Q(x)|u|q+1 dx > 0, then there exist R, α > 0 and e ∈ H01 (Ω) with kek > R such that I|∂BR ≥ α and I(e) < 0, where ∂BR = {u ∈ H01 (Ω) | kuk = R}. Thus, I satisfies the geometry of the mountain-pass lemma. Let c = inf max I(γ(t)), γ∈Γ t∈[0,1] where Γ = {γ ∈ C([0, 1], H01 (Ω)) : γ(0) = 0, γ(1) = e}. Then c ≥ α. According to Lemma 2.1, I satisfies the conditions of the mountain pass lemma. Applying the mountain-pass lema, there exists a sequence {un } ⊂ H01 (Ω), such that I(un ) → c and I 0 (un ) → 0 as n → ∞. Then c is a critical value of I and c > α > 0. Moreover, {un } ⊂ H01 (Ω) has a convergent subsequence, still denoted by {un }, we may assume that un → u∗ in H01 (Ω) as n → ∞. Thus I(u∗ ) = c > 0 and u∗ is a nonzero solution of (1.1). Since I(|u|) = I(u), by a result due to Brézis and Nirenberg [4, Theorem 10], we conclude that u∗ ≥ 0. By the strong maximum principle, one has u∗ > 0 in Ω. Therefore, u∗ is a positive solution of problem (1.1) with I(u∗ ) > 0. This completes the proof. 3. The case q=5 In this part, assume that Q ∈ C(Ω) is a positive function and satisfies (A1). We study the case of q = 5. This case is more delicate, because of the Sobolev embedding H01 (Ω) ,→ L6 (Ω) is not compact. Thus the functional I does not satisfy the (P S)c condition. When Q(x) is not constant, the analysis of (P S) sequences becomes complicated, which results in much difficulty. We will complete the proof of Theorem 1.3 by the mountain pass lemma. Now, we prove that I satisfies the local (P S)c condition. Lemma 3.1. Assume a, b > 0 and the positive function Q ∈ C(Ω) satisfies (A1), then I satisfies the (P S)c condition, where c ∈ (0, Λ) with p p abS 3 b3 S 6 aS b2 S 4 + 4aSQM b2 S 4 b2 S 4 + 4aSQM Λ= + + + . 4QM 24Q2M 6QM 24Q2M Proof. Suppose that {un } is a (P S)c sequence for c ∈ (0, Λ); that is, I(un ) → c, I 0 (un ) → 0, (3.1) as n → +∞. According to Lemma 2.1, we can easy obtain that {un } is bounded in H01 (Ω). Going if necessary to a subsequence, there exists u ∈ H01 (Ω) such that (2.2) holds. As usually, letting wn = un − u, we need prove that kwn k → 0 as 6 J.-F. LIAO, P. ZHANG, X.-P. WU EJDE-2015/280 n → ∞. We denote limn→∞ kwn k = l. As in Lemma 2.1, we have (2.4) and (2.5). By Brézis-Lieb’s Lemma [5], one has Z Z Z Q(x)|un |6 dx = Q(x)|wn |6 dx + Q(x)|u|6 dx + o(1). (3.2) Ω Ω Ω From (3.1) and (2.2), one obtains akun k2 + bkun k4 − ν Z Z |u|4 dx − Ω Q(x)|un |6 dx = o(1), Ω consequently, from (2.4)-(2.5) and (3.2) it follows that akuk2 + akwn k2 + bkuk4 + bkwn k4 + 2bkwn k2 kuk2 Z Z Z 6 6 − Q(x)|wn | dx − Q(x)|u| dx − ν |u|4 dx = o(1). Ω Ω (3.3) Ω From (3.1) it follows that 0 2 4 2 2 Z lim hI (un ), ui = akuk +bkuk +bl kuk − n→∞ 6 Z Q(x)|u| dx−ν Ω |u|4 dx = 0. (3.4) Ω On the one hand, from (3.4), we have Z Z b ν 1 a 2 4 4 |u| dx − Q(x)|u|6 dx I(u) = kuk + kuk − 2 4 4 Ω 6 Ω Z a 1 bl2 2 = kuk + Q(x)|u|6 dx − kuk2 4 12 Ω 4 bl2 ≥ − kuk2 . 4 On the other hand, from (3.3) and (3.4) it follows that Z akwn k2 + bkwn k4 + bkwn k2 kuk2 − Q(x)|wn |6 dx = o(1), (3.5) (3.6) Ω and b b 1 a I(un ) = I(u) + kwn k2 + kwn k4 + kwn k2 kuk2 − 2 4 2 6 Z Q(x)|wn |6 dx + o(1). (3.7) Ω From (A1) and (1.7), one has Z Z kwn k6 6 Q(x)|wn | dx ≤ QM |wn |6 dx ≤ QM , S3 Ω Ω consequently, it follows from (3.6) that al2 + bl4 + bl2 kuk2 ≤ QM l6 , S3 which implies that 1 h bS 3 l ≥ + 2 QM 2 p b2 S 6 + 4S 3 QM (a + bkuk2 ) i . QM (3.8) Thus, from (3.6)-(3.8), we obtain Z h i a b b 1 2 4 2 2 I(u) = lim I(un ) − kwn k − kwn k − kwn k kuk + Q(x)|wn |6 dx n→∞ 2 4 2 6 Ω a b b =c− l2 + l4 + l2 kuk2 3 12 3 EJDE-2015/280 KIRCHHOFF PROBLEMS h a bS 3 7 p b2 S 6 + 4S 3 QM (a + bkuk2 ) + 6 QM QM p 3 2 6 3 b S + 4S QM (a + bkuk2 ) 2 b bS + + 48 QM QM p b2 S 6 + 4S 3 QM (a + bkuk2 ) i bl2 bkuk2 bS 3 − + + kuk2 24 QM QM 4 p p abS 3 aS b2 S 4 + 4aSQM b3 S 6 b2 S 4 b2 S 4 + 4aSQM + ≤c− + + 4QM 24Q2M 6QM 24Q2M ≤c− bl2 kuk2 4 bl2 < − kuk2 , 4 which contradicts (3.5). Hence, l ≡ 0; that is, un → u in H01 (Ω) as n → ∞. Therefore, I satisfies the (P S)c condition for all c < Λ. This completes the proof. − Next, we estimate the level value of functional I and obtain the following lemma. Lemma 3.2. Assume that a, b > 0 and the positive function Q ∈ C(Ω) satisfies (A1). Then there exists u0 ∈ H01 (Ω), such that supt≥0 I(tu0 ) < Λ for all ν > ν ∗ , where Λ is defined by Lemma 3.1 and ν ∗ independent of u0 is a positive constant. Proof. Define a cut-off function η ∈ C0∞ (Ω) such that 0 ≤ η ≤ 1, |∇η| ≤ C1 . For some δ̃ > 0, we define ( 1, |x − x0 | ≤ 2δ̃ , η(x) = 0, |x − x0 | ≥ δ̃, where x0 is defined by (A1). Set uε = η(x)U (x − x0 ). As well known(see [6, 29]), one has kuε k2 = kUε k2 + O(ε) = S 3/2 + O(ε), |uε |66 = |Uε |66 3 + O(ε ) = S 3/2 (3.9) 3 + O(ε ), (3.10) and s 2 Z s usε dx ≤ C3 ε 2 , 1 ≤ s < 3, Ω Z s s C4 ε 2 | ln ε| ≤ usε dx ≤ C5 ε 2 | ln ε|, s = 3, Z Ω 6−s 6−s 2 C6 ε ≤ usε dx ≤ C7 ε 2 , 3 < s < 6. C2 ε ≤ (3.11) Ω Moreover, from [30], we have 9 kuε k4 = S 3 + O(ε), kuε k6 = S 2 + O(ε), kuε k8 = S 6 + O(ε), kuε k12 = S 9 + O(ε). (3.12) For all t ≥ 0, we define I(tuε ) by a b ν I(tuε ) = t2 kuε k2 + t4 kuε k4 − t4 2 4 4 t6 |uε | dx − 6 Ω Z 4 Z Ω Q(x)|uε |6 dx, 8 J.-F. LIAO, P. ZHANG, X.-P. WU EJDE-2015/280 then we have lim I(tuε ) = 0, uniformly for all 0 < ε < ε0 , t→+0 lim I(tuε ) = −∞, t→+∞ uniformly for all 0 < ε < ε0 , where ε0 > 0 is a small constant. Thus supt≥0 I(tuε ) attains for some tε > 0. Moreover, we can claim that there exist two constants t0 , T0 > 0, which independent of ε, such that t0 < tε < T0 . In fact, from limt→+0 I(tuε ) = 0 uniformly for all ε, we choose = I(tε4uε ) > 0, then there exists t0 > 0 such that |I(t0 uε )| = |I(t0 uε ) − I(0)| < . Then according to the monotonicity of I(tuε ) near t = 0, we have tε > t0 . Similarly, we can obtain that tε < T0 . Therefore, our claim is proved. Set I(tuε ) = Iε,1 (t) − νIε,2 (t), where Z a 2 b 4 t6 2 4 Q(x)u6ε dx, Iε,1 (t) = t kuε k + t kuε k − 2 4 6 Ω and Z t4 u4 dx. 4 Ω ε R 0 First, we estimate the value Iε,1 . Since Iε,1 (t) = atkuε k2 +bt3 kuε k4 −t5 Ω Q(x)u6ε dx, 0 letting Iε,1 (t) = 0; that is, Z akuε k2 + bt2 kuε k4 − t4 Q(x)u6ε dx = 0, (3.13) Iε,2 (t) = Ω one obtains q R b2 kuε k8 + 4akuε k2 Ω Q(x)u6ε dx R Tε2 = . 2 Ω Q(x)u6ε dx 0 0 Then Iε,1 (t) > 0 for all 0 < t < Tε and Iε,1 (t) < 0 for all t > Tε , so Iε,1 (t) attains + its maximum at Tε . From (A1), let ε → 0 , we claim that Z 1/3 1/3 6 Q(x)uε dx = QM |uε |26 + o(ε). (3.14) bkuε k4 + Ω In fact, for all ε > 0, it follows that Z Z Z 6 ≤ Q(x)u6ε dx − Q u dx |Q(x) − Q(x0 )|u6ε dx M ε Ω Ω Ω Z ≤ |Q(x) − Q(x0 )|u6ε dx. {x∈Ω: |x−x0 |≤δ̃} From (A1), for all η > 0, there exists δ > 0 such that |Q(x) − Q(x0 )| < η|x − x0 |, for all 0 < |x − x0 | < δ. When ε > 0 small enough, for δ > ε1/2 , it follows from (3.15) and (A1) that Z Z Q(x)u6ε dx − QM u6ε dx Ω Ω Z ≤ |Q(x) − Q(x0 )|u6ε dx {x∈Ω:|x−x0 |≤δ̃} Z η|x − x0 | < {x∈Ω:|x−x0 |≤δ} [ε2 (3ε2 )3/2 dx + |x − x0 |2 ]3 (3.15) EJDE-2015/280 KIRCHHOFF PROBLEMS 9 (3ε2 )3/2 dx + |x − x0 |2 ]3 {x∈Ω: δ<|x−x0 |≤δ̃} Z δ √ √ Z δ̃ ε3 ε3 r 2 3 = 27η dr + dr r 2 27 2 2 3 (ε + r2 )3 0 δ (ε + r ) δ̃ Z δε √ √ Z ε r2 r3 = 27ηε dr + 27 dr 2 3 δ (1 + r2 )3 0 (1 + r ) ε Z + [ε2 ≤ C8 ηε + C9 ε3 . Consequently, one has R R Q(x)u6ε dx − QM u6ε dx Ω Ω ≤ C8 η + C9 ε2 , ε which implies R R Q(x)u6ε dx − QM u6ε dx Ω Ω ≤ C8 η. lim sup ε ε→0+ Then from the arbitrariness of η, we obtain (3.14). Thus, from (3.10) and (3.14), one gets Z Q(x)u6ε dx = QM |uε |66 + o(ε) = QM S 3/2 + o(ε). (3.16) Ω Thus from (3.9),(3.12),(3.13) and (3.16), we have Iε,1 (t) ≤ Iε,1 (Tε ) a = Tε2 kuε k2 + 2 a = Tε2 kuε k2 + 3 b 2 T4 Tε kuε k4 − ε 4 6 b 2 Tε kuε k4 12 q Z Q(x)u6ε dx Ω abkuε k6 + akuε k2 R b2 kuε k8 + 4akuε k2 Ω Q(x)u6ε dx R = 6 Ω Q(x)u6ε dx R b3 kuε k12 + 2abkuε k6 Ω Q(x)u6ε dx + 2 R 24 Ω Q(x)u6ε dx q R b2 kuε k4 b2 kuε k8 + 4akuε k2 Ω Q(x)u6ε dx + 2 R 24 Ω Q(x)u6ε dx = abkuε k6 b3 kuε k12 R + 4 Ω Q(x)u6ε dx 24( Ω Q(x)u6ε dx)2 q R akuε k2 b2 kuε k8 + 4akuε k2 Ω Q(x)u6ε dx R + 6 Ω Q(x)u6ε dx q R b2 kuε k4 b2 kuε k8 + 4akuε k2 Ω Q(x)u6ε dx + 2 R 24 Ω Q(x)u6ε dx R 9 ab(S 2 + O(ε)) b3 (S 9 + O(ε)) = + 4(QM S 3/2 + o(ε)) 24(QM S 3/2 + o(ε))2 10 J.-F. LIAO, P. ZHANG, X.-P. WU p a(S 3/2 + O(ε)) b2 S 6 + 4aS 3 + O(ε) + 6(QM S 3/2 + o(ε)) p 2 6 b (S + O(ε)) b2 S 6 + 4aS 3 + O(ε) + 24(QM S 3/2 + o(ε))2 p abS 3 b3 S 6 aS b2 S 4 + 4aSQM = + + 4QM 24Q2M 6QM p 2 4 b S b2 S 4 + 4aSQM + + O(ε) 24Q2M = Λ + O(ε). EJDE-2015/280 (3.17) Second, we estimate the value of Iε,2 . From (3.11), since 0 < t0 < tε < T0 , one has Z Z t4 t4 u4ε dx ≥ 0 u4 dx ≥ C10 ε. (3.18) Iε,2 (tε ) = ε 4 Ω 4 Ω ε Thus, from (3.17) and (3.18), one gets I(tuε ) = Iε,1 (t) − νIε,2 (t) ≤ Iε,1 (tε ) − νIε,2 (tε ) Z t40 ≤ Iε,1 (Tε ) − ν u4ε dx 4 Ω ≤ Λ + O(ε) − C10 νε < Λ, provided that ν is large enough. Thus there exists ν ∗ > 0 such that I(tuε ) < Λ for all ν > ν ∗ . This completes the proof. Proof of Theorem 1.3. As in the proof of Theorem 1.1, we can obtain that I has the geometry of mountain pass lemma in H01 (Ω). According to Lemmas 3.1 and 3.2, it follows that I satisfies the conditions of the mountain pass lemma. Then as in the proof of Theorem 1.1, we obtain that (1.1) has a positive solution u∗ with I(u∗ ) > 0 as long as ν > ν ∗ . The proof is complete. Acknowledgments. This research was supported by the National Science Foundation of China (No. 11471267), by the Science and Technology Foundation of Guizhou Province (No. LKZS[2014]30, No. LH[2015]7001, No. LH[2015]7049). The authors would like to thank the anonymous referees and Professor G. Molica Bisci for their valuable suggestions. References [1] C. O. Alves, F. J. S. A. Corrêa, G. M. Figueiredo; On a class of nonlocal elliptic problems with critical growth, Differ. Equ. Appl. 2 (2010) 409–417. [2] A. Ambrosetti, P. H. Rabinowitz; Dual variational methods in critical point theory and applications, J. Funct. Anal., 14 (1973), 349–381. [3] G. Anello; A uniqueness result for a nonlocal equation of Kirchhoff type and some related open problems, J. Math. Anal. Appl., 373 (2011), 248–251. [4] H. Berestycki, I. Capuzzo-Dolcetta, L. Nirenberg; Variational methods for indefinite superlinear homogeneous elliptic problems, NoDEA Nonlinear Differential Equations Appl., 2 (1995), 553–572. [5] H. Brézis, E. Lieb; A relation between pointwise convergence of functions and convergence of functionals, Proc. Amer. Math. Soc. 88 (1983), 486–490. EJDE-2015/280 KIRCHHOFF PROBLEMS 11 [6] H. Brézis, L. Nirenberg; Positive solutions of nonlinear elliptic equations involving critical exponents, Comm. Pure Appl. Math., 36 (1983), 437–477. [7] C. Y. Chen, Y. C. Kuo, T. F. Wu; he Nehari manifold for a Kirchhoff type problem involving sign-changing weight functions, J. Differential Equations, 250 (2011), 1876–1908. [8] G. M. Figueiredo, G. Molica Bisci, R. Servadei; On a fractional Kirchhoff-type equation via Krasnoselskii’s genus, Asymptot. Anal., 94 (2015), 347–361. [9] X. M. He, W. M. Zou; Infnitely many solutions for Kirchhoff-type problems, Nonlinear Anal., 70 (2009), 1407–1414. [10] X. M. He, W. M. Zou; Existence and concentration behavior of positive solutions for a Kirchhoff equation in R3 , J. Differential Equations, 252 (2012), 1813–1834. [11] X. M. He, W. M. Zou; Ground states for nonlinear Kirchhoff equations with critical growth, Ann. Mat. Pura Appl., 193 (2014), 473–500. [12] C. Y. Lei, J. F. Liao, C. L. Tang; Multiple positive solutions for Kirchhoff type of problems with singularity and critical exponents, J. Math. Anal. Appl., 421 (2015), 521–538. [13] Y. H. Li, F. Y. Li, J. P. Shi; Existence of a positive solution to Kirchhoff type problems without compactness conditions, J. Differential Equations, 253 (2012), 2285–2294. [14] X. Liu, Y.J. Sun; Multiple positive solutions for Kirchhoff type problems with singularity, Commun. Pure Appl. Anal., 12 (2013), 721–733. [15] A. M. Mao, S. X. Luan; Sign-changing solutions of a class of nonlocal quasilinear elliptic boundary value problems, J. Math. Anal. Appl., 383 (2011), 239–243. [16] A. M. Mao, Z. T. Zhang; Sign-changing and multiple solutions of Kirchhoff type problems without the P.S. condition, Nonlinear Anal., 70 (2009), 1275–1287. [17] G. Molica Bisci, P. F. Pizzimenti; Sequences of weak solutions for non-local elliptic problems with Dirichlet boundary condition, Proc. Edinb. Math. Soc., 57 (2014), 779–809. [18] G. Molica Bisci, V. D. Rădulescu; Mountain Pass Solutions for Nonlocal Equations, Ann. Acad. Fenn. Math., 39 (2014), 579–592. [19] G. Molica Bisci, D. Repovš; On doubly nonlocal fractional elliptic equations, Atti Accad. Naz. Lincei Rend. Lincei Mat. Appl., 26 (2015), 161–176. [20] D. Naimen; The critical problem of Kirchhoff type elliptic equations in dimension four, J. Differential Equations, 257 (2014), 1168–1193. [21] D. Naimen; Positive solutions of Kirchhoff type elliptic equations involving a critical Sobolev exponent, NoDEA Nonlinear Differential Equations Appl., 21 (2014), 885–914. [22] D. Naimen; On the Brezis-Nirenberg problem with a Kirchhoff type perturbation, Adv. Nonlinear Stud., 15 (2015), 135–156. [23] K. Perera, Z. T. Zhang; Nontrivial solutions of Kirchhoff-type problems via the Yang index, J. Differential Equations 221 (2006), 246–255. [24] W. Rudin; Real and complex analysis, McGraw-Hill, New York, London etc. 1966. [25] J. J. Sun, C. L. Tang; Existence and multipicity of solutions for Kirchhoff type equations, Nonlinear Anal. 74 (2011), 1212–1222. [26] J. J. Sun, C. L. Tang; Resonance problems for Kirchhoff type equations, Discrete Contin. Dyn. Syst., 33 (2013), 2139–2154. [27] Y. J. Sun, X. Liu; Existence of positive solutions for Kirchhoff type problems with critical exponent, J. Partial Differ. Equ., 25 (2012), 187–198. [28] J. Wang, L.X. Tian, J. X. Xu, F. B. Zhang; Multiplicity and concentration of positive solutions for a Kirchhoff type problem with critical growth, J. Differential Equations, 253 (2012), 2314– 2351. [29] M. Willem; Minimax Theorems, Birthäuser, Boston, 1996. [30] Q. L. Xie, X. P. Wu, C. L. Tang; Existence and multiplicity of solutions for Kirchhoff type problem with critical exponent, Commun. Pure Appl. Anal., 12 (2013), 2773–2786. [31] Z. T. Zhang, K. Perera; Sign changing solutions of Kirchhoff type problems via invariant sets of descent flow, J. Math. Anal. Appl., 317 (2006), 456–463. Jia-Feng Liao School of Mathematics and Statistics, Southwest University, Chongqing 400715, China. School of Mathematics and Computational Science, Zunyi Normal College, Zunyi, China E-mail address: liaojiafeng@163.com 12 J.-F. LIAO, P. ZHANG, X.-P. WU EJDE-2015/280 Peng Zhang School of Mathematics and Computational Science, Zunyi Normal College, Zunyi, China E-mail address: gzzypd@sina.com Xing-Ping Wu (corresponding author) School of Mathematics and Statistics, Southwest University, Chongqing 400715, China E-mail address: wuxp@swu.edu.cn