ebook img

Global compactness for a class of quasi-linear elliptic problems PDF

0.24 MB·
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Global compactness for a class of quasi-linear elliptic problems

GLOBAL COMPACTNESS FOR A CLASS OF QUASI-LINEAR ELLIPTIC PROBLEMS CARLO MERCURI AND MARCO SQUASSINA Abstract. We prove a global compactness result for Palais-Smale sequences associated with a 2 class of quasi-linear elliptic equations on exterior domains. 1 0 2 n 1. Introduction and main result a J Let Ω be a smooth domain of RN with a boundedcomplement and N > p > m > 1. The main 0 goal of this paper is to obtain a global compactness result for the Palais-Smale sequences of the 2 energy functional associated with the following quasi-linear elliptic equation ] P (1.1) −div(L (Du))−div(M (u,Du))+M (u,Du)+V(x)|u|p−2u = g(u) in Ω, ξ ξ s A 1,p 1,m where u ∈ W (Ω)∩D (Ω), meant as the completion of the space D(Ω) of smooth functions . 0 0 h with compact support, with respect to the norm kuk = kuk +kuk , having set t W1,p(Ω)∩D1,m(Ω) p m a kuk := kuk and kuk := kDuk . We assume that V is a continuous function on Ω, m p W1,p(Ω) m Lm(Ω) lim V(x) = V and inf V(x) = V > 0. [ ∞ 0 |x|→∞ x∈Ω 2 As known, lack of compactness may occur due to the lack of compact embeddings for Sobolev v 6 spaces on Ω and since the limiting equation on RN 4 (1.2) −div(L (Du))−div(M (u,Du))+M (u,Du)+V |u|p−2u= g(u) in RN, 0 ξ ξ s ∞ 4 with u∈ W1,p(RN)∩D1,m(RN), is invariant by translations. A particular case of (1.1) is . 7 1 0 (1.3) −∆ u−div(a(u)|Du|m−2Du)+ a′(u)|Du|m +V(x)|u|p−2u =|u|σ−2u in Ω, 1 p m 1 where ∆ u := div(|Du|p−2Du), for a suitable function a ∈ C1(R;R+), or the even simpler case : p v where a is constant, namely i X (1.4) −∆ u−∆ u+V(x)|u|p−2u = |u|σ−2u in Ω. p m r a Since the pioneering work of Benci and Cerami [2] dealing with the case L(ξ) = |ξ|2/2 and M(s,ξ) ≡ 0, many papers have been written on this subject, see for instance the bibliography of [12]. Quite recently, in [12], the case L(ξ) = |ξ|p/p and M(s,ξ) ≡ 0 was investigated. The main point in the present contribution is the fact that we allow, under suitable assumptions, a quasi-linear term M(u,Du) depending on the unknown u itself. The typical tools exploited in [2,12], in addition to the point-wise convergence of the gradients, are some decomposition (splitting) results both for the energy functional and for the equation, along a given bounded Palais-Smale sequence (u ). To this regard, the explicit dependence on u in the term M(u,Du) n requires a rather careful analysis. In particular, we can handle it for ν|ξ|m ≤ M(s,ξ) ≤ C|ξ|m, p−1≤ m < p−1+p/N. 2000 Mathematics Subject Classification. 35D99, 35J62, 58E05, 35J70. Key words and phrases. Quasi-linear equations, global compactness of Palais-Smale sequences. Supported by Miur project: “Variational and Topological Methods in the Study of Nonlinear Phenomena”. 1 2 C.MERCURIANDM.SQUASSINA The restriction on m, together with the sign condition (1.9) provides, thanks to the presence of L, the needed a priori regularity on the weak limit of (u ), see Theorems 3.2 and 3.4. n Besides the aforementioned motivations, which are of mathematical interest, it is worth pointing out that in recent years, some works have been devoted to quasi-linear operators with double homogeneity, which arise from several problems of Mathematical Physics. For instance, the reaction diffusion problem u = −div(D(u)Du)+ℓ(x,u), where D(u) = d |Du|p−2+d |Du|m−2, t p m d > 0 and d > 0, admitting a rather wide range of applications in biophysics [10], plasma p m physics [16] and in the study of chemical reactions [1]. In this framework, u typically describes a concentration and div(D(u)Du) corresponds to the diffusion with a coefficient D(u), whereas ℓ(x,u) plays the roˇle of reaction and relates to source and loss processes. We refer the interested readerto[5]andtothereferencetherein. Furthermore,amodelforelementary particles proposed by Derrick [9] yields to the study of standing wave solutions ψ(x,t) = u(x)eiωt of the following nonlinear Schro¨dinger equation iψ +∆ ψ−b(x)ψ+∆ ψ−V(x)|ψ|p−2ψ+|ψ|σ−2ψ = 0 in RN, t 2 p for which we refer the reader e.g. to [3]. In order to state the first main result, assume N > p > m ≥ 2 and (1.5) p−1 ≤ m < p−1+p/N, p < σ < p∗, and consider the C2 functions L : RN → R and M : R×RN → R such that both the functions ξ 7→ L(ξ) and ξ 7→ M(s,ξ) are strictly convex and (1.6) ν|ξ|p ≤ |L(ξ)| ≤ C|ξ|p, |L (ξ)| ≤ C|ξ|p−1, |L (ξ)| ≤ C|ξ|p−2, ξ ξξ for all ξ ∈ RN. Furthermore, we assume (1.7) ν|ξ|m ≤ M(s,ξ)| ≤ C|ξ|m, |M (s,ξ)| ≤ C|ξ|m, |M (s,ξ)| ≤ C|ξ|m−1, s ξ (1.8) |M (s,ξ)| ≤ C|ξ|m, |M (s,ξ)| ≤ C|ξ|m−1, |M (s,ξ)| ≤ C|ξ|m−2, ss sξ ξξ for all (s,ξ) ∈ R×RN and that the sign condition (cf. [14]) (1.9) M (s,ξ)s ≥ 0, s holds for all (s,ξ) ∈ R×RN. Also, G : R → R is a C2 function with G′(s) := g(s) and (1.10) |G′(s)| ≤ C|s|σ−1, |G′′(s)|≤ C|s|σ−2, for all s ∈ R. We define (1.11) j(s,ξ) := L(ξ)+M(s,ξ)−G(s), 1,p 1,m and on W (Ω)∩D (Ω) with kuk = kuk +kuk the functional 0 0 W1,p(Ω)∩D1,m(Ω) p m |u|p φ(u) := j(u,Du)+ V(x) . Z Z p Ω Ω Finally, on W1,p(RN)∩D1,m(RN) with kukW1,p(RN)∩D1,m(RN) = kukp +kukm we define |u|p φ (u) := j(u,Du)+ V . ∞ ∞ ZRN ZRN p See Section 2 for some properties of the functionals φ and φ . ∞ The first main global compactness type result is the following GLOBAL COMPACTNESS FOR QUASI-LINEAR PROBLEMS 3 Theorem 1.1. Assume that (1.5)-(1.11) hold and let (u ) ⊂ W1,p(Ω)∩D1,m(Ω) be a bounded n 0 0 sequence such that φ(u ) → c φ′(u )→ 0 in (W1,p(Ω)∩D1,m(Ω))∗ n n 0 0 Then, up to a subsequence, there exists a weak solution v ∈ W1,p(Ω)∩D1,m(Ω) of 0 0 0 −div(L (Du))−div(M (u,Du))+M (u,Du)+V(x)|u|p−2u= g(u) in Ω, ξ ξ s a finite sequence {v ,...,v } ⊂ W1,p(RN)∩D1,m(RN) of weak solutions of 1 k −div(L (Du))−div(M (u,Du))+M (u,Du)+V |u|p−2u = g(u) in RN ξ ξ s ∞ and k sequences (yi)⊂ RN satisfying n |yi|→ ∞, |yi −yj|→ ∞, i6= j, as n → ∞, n n n k kun−v0− vi((·−yni)kW1,p(RN)∩D1,m(RN) → 0, as n → ∞, Xi=1 k k ku kp → kv kp, ku km → kv km, as n → ∞, n p i p n m i m Xi=0 Xi=0 as well as k φ(v )+ φ (v )= c. 0 ∞ i Xi=1 Let us now come to a statement for the cases 1 < m ≤ 2 or 1 < p ≤ 2. Let us define |L (ξ +h)−L (ξ)| L(ξ,h) := ξ ξ , if 1 < p < 2, |h|p−1 |G′(s+t)−G′(s)| G(s,t) := , if 1 < σ < 2, |t|σ−1 |M (s,ξ +h)−M (s,ξ)| M(s,ξ,h) := ξ ξ , if 1 < m < 2. |h|m−1 If either p < 2, σ < 2 or m < 2, we shall weaken the twice differentiability assumptions, by requiring L ∈ C1(RN \{0}), G′ ∈ C1(R\{0}), M ∈ C1(R×(RN \{0})), M ∈ C0(R×RN) ξ ξ sξ and M ∈ C0(R×RN). Moreover we assume the same growth conditions for L,M,G and their ss derivatives, replacing only the growth assumptions for L ,M ,G′′ by the following hypotheses: ξξ ξξ (1.12) sup L(ξ,h) < ∞, h6=0,ξ∈RN (1.13) sup G(s,t) < ∞, t6=0,s∈R (1.14) sup M(s,ξ,h) < ∞. h6=0,(s,ξ)∈R×RN Conditions (1.12)-(1.13), in some more concrete situations, follow immediately by homogeneity of L and G′ (see, for instance, [12, Lemma 3.1]). Similarly, (1.14) is satisfied for instance when ξ M is of the form M(s,ξ) = a(s)µ(ξ), being a : R → R+ a bounded function and µ : RN → R+ a C1 strictly convex function such that µ is homogeneous of degree m−1. ξ Theorem 1.2. Under the additional assumptions (1.12)-(1.14) in the sub-quadratic cases, the assertion of Theorem 1.1 holds true. As a consequence of the above results we have the following compactness criterion. 4 C.MERCURIANDM.SQUASSINA Corollary 1.3. Assume (2.1) below for some δ > 0 and µ > p. Under the hypotheses of Theorem 1.1 or 1.2, if c <c∗, then (u ) is relatively compact in W1,p(Ω)∩D1,m(Ω) where n 0 0 p c∗ := min δ,µ−pV min{ν,V∞} σ−p , ∞ (cid:26)µ µp (cid:27)(cid:20) C S (cid:21) g p,σ and S and C are constants such that S kukσ ≥ kukσ and |g(s)| ≤ C |s|σ−1. p,σ g p,σ p Lσ(RN) g Remark 1.4. It would be interesting to get a global compactness result in the case L = 0 and p = m, namely for the model case 1 (1.15) −div(a(u)|Du|m−2Du)+ a′(u)|Du|m +V(x)|u|m−2u= |u|σ−2u in Ω. m Notice that, even assuminga′ bounded,a′(u)|Du|m ismerely inL1(Ω)forW1,m(Ω)distributional 0 solutions. Ingeneral, inthis setting, thesplitting propertiesof theequation arehardtoformulate in a reasonable fashion. Remark 1.5. The restriction of between m and p in assumption (1.5) is no longer needed in the case where M is independent of the first variable s, namely M ≡ 0. s Remark 1.6. We prove the above theorems under the a-priori boundedness assumption of (u ). n This occurs in a quite large class of problems, as Proposition 2.2 shows. Remark 1.7. With no additional effort, we could cover the case where an additional term W(x)|u|m−2u appears in (1.1) and the functional framework turns into W1,p(Ω)∩W1,m(Ω). 0 0 In the spirit of [11], we also get the following Corollary 1.8. Let N > p ≥ m > 1 and assume that ξ 7→ L(ξ) is p-homogeneous, ξ 7→ M(ξ) is m-homogeneous, L(ξ) ≥ p|ξ|p, M(ξ) ≥ m|ξ|m and set L(Du) M(Du) V(x) (1.16) S := inf + + |u|p, Ω kukLσ(Ω)=1ZΩ p m p |Du|p |u|p S := inf + , RN kukLσ(RN)=1ZRN p p with V(x)→ 1 as |x|→ ∞. Assume furthermore that σ−p σ−p (1.17) SΩ < σ SRN. σ−m (cid:16) (cid:17) Then (1.16) admits a minimizer. Remark 1.9. We point out that, some conditions guaranteeing the nonexistence of nontrivial solutions inthestar-shapedcase Ω = RN can beprovided. For thesake of simplicity, assumethat L is p-homogeneous and that ξ 7→ M(s,ξ) is m-homogeneous. Then, in view of [13, Theorem 3], that holds for C1 solutions by virtue of the results of [8], we have that (1.1) admits no nontrivial C1 solution well behaved at infinity, namely satisfying condition (19) of [13], provided that there exists a number a ∈ R+ such that a.e. in RN and for all (s,ξ) ∈ R×RN (N −p(a+1))L(ξ)+(N −m(a+1))M(s,ξ)+(asg(s)−NG(s)) (N −ap)V(x)+x·DV(x) + |s|p−aM (s,ξ)s ≥ 0, s p holding, for instance, if there exists 0≤ a ≤ N−p such that p asg(s)−NG(s) ≥0, (N −ap)V(x)+x·DV(x) ≥ 0, M (s,ξ)s ≤0, s GLOBAL COMPACTNESS FOR QUASI-LINEAR PROBLEMS 5 fora.e. x ∈ RN andforall (s,ξ) ∈ R×RN. Also, in themoreparticular case whereg(s) = |s|σ−2s and V(x) = V > 0, then the above conditions simply rephrase into ∞ σ ≥ p∗, M (s,ξ)s ≤ 0, s for every (s,ξ) ∈ R×RN. In fact, in (1.9), we consider the opposite assumption on M . s 2. Some preliminary facts It is a standard fact that, under condition (1.6) and (1.10), the functionals u7→ L(Du), u7→ V(x)|u|p, u7→ G(u) Z Z Z Ω Ω Ω are C1 on W1,p(Ω)∩D1,m(Ω). Analogously, although M depends explicitly on s, the functional 0 0 M :W1,p(Ω)∩D1,m(Ω)→ R, M(u)= M(u,Du), 0 0 Z Ω 1,p 1,m admits, thanks to condition (1.5), directional derivatives along any v ∈ W (Ω)∩D (Ω) and 0 0 M′(u)(v) = M (u,Du)·Dv+ M (u,Du)v, ξ s Z Z Ω Ω as it can be easily verified observing that p ≤ p ≤ p∗ and that, for u∈ W1,p(Ω)∩D1,m(Ω), by p−m 0 0 Young’s inequality, for some constant C it holds |M (u,Du)·Dv|≤ C|Du|m+C|Dv|m ∈L1(Ω), ξ p |Ms(u,Du)v| ≤ C|Du|p+C|v|p−m ∈ L1(Ω). Furthermore, if u → u in W1,p(Ω)∩D1,m(Ω) as k → ∞ then M′(u )→ M′(u) in the dual space k 0 0 k (W1,p(Ω)∩D1,m(Ω))∗, as k → ∞. Indeed, for kvk ≤ 1, we have 0 0 W1,p(Ω)∩D1,m(Ω) 0 0 |M′(u )(v)−M′(u)(v)| k ≤ |M (u ,Du )−M (u,Du)||Dv|+ |M (u ,Du )−M (u,Du)||v| ξ k k ξ s k k s Z Z Ω Ω ≤ kMξ(uk,Duk)−Mξ(u,Du)kLm′kDvkLm +kMs(uk,Duk)−Ms(u,Du)kLp/mkvkLp/(p−m) ≤ kMξ(uk,Duk)−Mξ(u,Du)kLm′ +kMs(uk,Duk)−Ms(u,Du)kLp/m. Thisyieldsthedesiredconvergence, using(1.7)andtheDominatedConvergenceTheorem. Notice thatthesameargumentcarriedoutbeforeapplieseithertointegralsdefinedonΩoronRN.Hence the following proposition is proved. Proposition 2.1. In the hypotheses of Theorems 1.1 and 1.2, the functionals φ and φ are C1. ∞ In addition to the assumptions on L,M and g,G set in the introduction, assume now that there exist positive numbers δ > 0 and µ > p such that (2.1) µM(s,ξ)−M (s,ξ)s−M (s,ξ)·ξ ≥ δ|ξ|m, µL(ξ)−L (ξ)·ξ ≥ δ|ξ|p, sg(s)−µG(s) ≥ 0, s ξ ξ for any s ∈ R and all ξ ∈ RN. This hypothesis is rather well established in the framework of quasi-linear problems (cf. [14]) and it allows an arbitrary Palais-Smale sequence (u ) to be n 1,p 1,m bounded in W (Ω)∩D (Ω), as shown in the following 0 0 6 C.MERCURIANDM.SQUASSINA Proposition 2.2. Let j be as in (1.11) and assume that (1.5) holds. Let (u ) ⊂ W1,p(Ω) ∩ n 0 D1,m(Ω) be a sequence such that 0 φ(u ) → c φ′(u )→ 0 in (W1,p(Ω)∩D1,m(Ω))∗ n n 0 0 Then, if condition (2.1) holds, (u ) is bounded in W1,p(Ω)∩D1,m(Ω). n 0 0 Proof. Let(w ) ⊂ (W1,p(Ω)∩D1,m(Ω))∗ withw → 0asn → ∞besuchthatφ′(u )(v) =hw ,vi, n 0 0 n n n 1,p 1,m for every v ∈ W (Ω)∩D (Ω). Whence, by choosing v = u , it follows 0 0 n j (u ,Du )·Du + j (u ,Du )u + V(x)|u |p = hw ,u i. ξ n n n s n n n n n n Z Z Z Ω Ω Ω Combining this equation with µφ(u )= µc+o(1) as n → ∞, namely n µ µj(u ,Du )+ V(x)|u |p = µc+o(1), n n n Z p Z Ω Ω recalling the definition of j, and using condition (2.1), yields µ−p V(x)|u |p+δ |Du |p+δ |Du |m ≤ µc+kw kku k +o(1), p ZΩ n ZΩ n ZΩ n n n W01,p(Ω)∩D01,m(Ω) as n → ∞, implying, due to V ≥ V that 0 ku kp +ku km ≤ C +Cku k +Cku k +o(1), n W1,p(Ω) n D1,m(Ω) n W1,p(Ω) n D1,m(Ω) as n → ∞. The assertion then follows immediately. (cid:3) From now on we shall always assume to handle bounded Palais-Smale sequences, keeping in mind that condition (2.1) can guarantee the boundedness of such sequences. Proposition 2.3. Let j be as in (1.11) and assume that 1 < m < p < N and p < σ < p∗. Let (u ) ⊂ W1,p(Ω)∩D1,m(Ω) bounded be such that n 0 0 φ(u ) → c φ′(u ) → 0 in (W1,p(Ω)∩D1,m(Ω))∗. n n 0 0 Then, up to a subsequence, (u ) converges weakly to some u in W1,p(Ω)∩D1,m(Ω), u (x) → u(x) n 0 0 n and Du (x) → Du(x) for a.e. x ∈ Ω. n Proof. Itis sufficient to justify that Du (x) → Du(x) for a.e. x ∈ Ω. Given an arbitrary bounded n subdomainω ⊂ ω ⊂ Ω of Ω, from the fact that φ′(u ) → 0 in (W1,p(Ω)∩D1,m(Ω))∗, we can write n 0 0 a(u ,Du )·Dv = hw ,vi+hf ,vi+ vdµ , for all v ∈ D(ω), n n n n n Z Z ω ω where (w ) ⊂ (W1,p(Ω)∩D1,m(Ω))∗ is vanishing, and hence in particular w ∈ W−1,p′(ω), with n 0 0 n w → 0 in W−1,p′(ω) as n → ∞ and we have set n a (x,s,ξ) := L (ξ)+M (s,ξ), for all (s,ξ) ∈ R×RN, n ξ ξ f := −V(x)|u |p−2u +g(u ) ∈ W−1,p′(ω), n ∈N, n n n n µ := −M (u ,Du ) ∈ L1(ω), n ∈ N. n s n n Due to the strict convexity assumptions on the maps ξ 7→ L(ξ) and ξ 7→ M(s,ξ) and the growth conditions on L ,M ,M and g, all the assumptions of [6, Theorem 1] are fulfilled. Precisely, ξ ξ s |a (x,s,ξ)| ≤ |L (ξ)|+|M (s,ξ)| ≤ C|ξ|p−1+C|ξ|m−1 ≤ C +C|ξ|p−1, n ξ ξ GLOBAL COMPACTNESS FOR QUASI-LINEAR PROBLEMS 7 for a.e. x ∈ ω and all (s,ξ) ∈ R×RN, and f → f, f := −V(x)|u|p−2u+g(u), strongly in W−1,p′(ω), n µ ⇀ µ, weakly* in M(ω), since supkM (u ,Du )k < +∞. n s n n L1(ω) n∈N Then,itfollows thatDu (x) → Du(x)fora.e. x ∈ ω. Finally, asimpleCantordiagonal argument n allows to recover the convergence over the whole domain Ω. (cid:3) Next we prove a regularity result for the solutions of equation (1.1). Proposition 2.4. Let j be as in (1.11) and assume (1.5) and (1.9). Let u ∈W1,p(Ω)∩D1,m(Ω) 0 0 be a solution of (1.1). Then u ∈ Lq(Ω), u∈ L∞(Ω) and lim u(x) = 0. q\≥p |x|→∞ Proof. Let k,i ∈ N. Then, setting v (x):= (u (x))i with u (x) := min{u+(x),k}, it follows that k,i k k 1,p 1,m v ∈ W (Ω)∩D (Ω) can be used as a test function in (1.1), yielding k,i 0 0 L (Du)·Dv + M (u,Du)·Dv ξ k,i ξ k,i Z Z Ω Ω + M (u,Du)v + V(x)|u|p−2uv = g(u)v . s k,i k,i k,i Z Z Z Ω Ω Ω Taking into account that Dv is equal to iui−1Duχ , by convexity and positivity of the k,i {0<u<k} map ξ 7→ M(s,ξ) we deduce that M (u,Du)·Dv ≥ 0. Moreover, by the sign condition (1.9) it ξ k,i follows M (u,Du)v ≥ 0 a.e. in Ω. Then, we reach s k,i i(u )i−1L (Du )·Du + V(x)|u|p−2u(u (x))i ≤ g(u)(u (x))i, k ξ k k k k Z Z Z Ω Ω Ω yielding in turn, by (1.10), that for all k,i ≥ 1 (2.2) νi (u )i−1|Du |p ≤ C (u+(x))σ−1+i. k k Z Z Ω Ω If uˆ := min{u−(x),k}, a similar inequality k (2.3) νi (uˆ )i−1|Duˆ |p ≤ C (u−(x))σ−1+i, k k Z Z Ω Ω can be obtained by using vˆ := −(uˆ )i as a test function in (1.1), observing that by (1.9), k,i k M (u,Du)vˆ = −M (u,Du)χ (−u)i ≥ 0, s k,i s {−k<u<0} M (u,Du)·Dv = i(−u)i−1χ M (u,Du)·Du ≥0. ξ k,i {−k<u<0} ξ Once (2.2)-(2.3) are reached, the assertion follows exactly as in [15, Lemma 2, (a) and (b)]. (cid:3) We now recall the following version of [7, Lemma 4.2] which turns out to be a rather useful tool in order to establish convergences in our setting. Roughly speaking, one needs some kind of sub-criticality in the growth conditions. Lemma 2.5. Let Ω ⊂ RN and h : Ω×R×RN be a Carath´eodory function, p,m > 1, µ ≥ 1, p ≤ σ ≤ p∗ and assume that, for every ε > 0 there exist a ∈ Lµ(Ω) such that ε (2.4) |h(x,s,ξ)| ≤ a (x)+ε|s|σ/µ +ε|ξ|p/µ +ε|ξ|m/µ, ε a.e. in Ω and for all (s,ξ) ∈ R×RN. Assume that u → u a.e. in Ω, Du → Du a.e. in Ω and n n (u ) is bounded in W1,p(Ω), (u ) is bounded in D1,m(Ω). n 0 n 0 8 C.MERCURIANDM.SQUASSINA Then h(x,u ,Du ) converges to h(x,u,Du) in Lµ(Ω). n n Proof. The proof follows as in [7, Lemma 4.2] and we shall sketch it here for self-containedness. 1,p 1,m By Fatou’s Lemma, it immediately holds that u∈ W (Ω)∩D (Ω). Furthermore, there exists 0 0 a positive constant C such that |h(x,s ,ξ )−h(x,s ,ξ )|µ ≤ C(a (x))µ+Cεµ|s |σ +Cεµ|s |σ 1 1 2 2 ε 1 2 +Cεµ|ξ |m+Cεµ|ξ |m+Cεµ|ξ |p+Cεµ|ξ |p, 1 2 1 2 a.e. in Ω and for all (s ,ξ ) ∈ R× RN and (s ,ξ ) ∈ R× RN. Then, taking into account the 1 1 2 2 boundednessof (Du ) in Lp(Ω)∩Lm(Ω) and of (u ) in Lσ(Ω) by interpolation being p ≤ σ ≤ p∗, n n the assertion follows by applying Fatou’s Lemma to the sequence of functions ψ : Ω → [0,+∞] n ψ (x) :=−|h(x,u ,Du )−h(x,u,Du)|µ +C(a (x))µ +Cεµ|u |σ +Cεµ|u|σ n n n ε n +Cεµ|Du |m+Cεµ|Du|m+Cεµ|Du |p+Cεµ|Du|p, n n and, finally, exploiting the arbitrariness of ε. (cid:3) 3. Proof of the result 3.1. Energy splitting. The next result allows to perform an energy splitting for the functional 1,p 1,m J(u) = j(u,Du), u∈ W (Ω)∩D (Ω), Z 0 0 Ω 1,p 1,m along a bounded Palais-Smale sequence (u ) ⊂ W (Ω)∩D (Ω). The result is in the spirit of n 0 0 the classical Brezis-Lieb Lemma [4]. Lemma 3.1. Let the integrand j be as in (1.11) and p−1 ≤m < p−1+p/N, p ≤ σ ≤ p∗. Let (u ) ⊂ W1,p(Ω)∩D1,m(Ω) with u ⇀ u, u → u a.e. in Ω and Du → Du a.e. in Ω. Then n 0 0 n n n (3.1) lim j(u −u,Du −Du)−j(u ,Du )+j(u,Du) = 0. n n n n n→∞Z Ω Proof. We shall apply Lemma 2.5 to the function h(x,s,ξ) := j(s−u(x),ξ −Du(x))−j(s,ξ), for a.e. x ∈ Ω and all (s,ξ) ∈ R×RN. Given x ∈ Ω, s ∈ R and ξ ∈ RN, consider the C1 map ϕ : [0,1] → R defined by setting ϕ(t) := j(s−tu(x),ξ −tDu(x)), for all t ∈ [0,1]. Then, for some τ ∈[0,1] depending upon x ∈ Ω, s∈ R and ξ ∈ RN, it holds h(x,s,ξ) = ϕ(1)−ϕ(0) = ϕ′(τ) = −j (s−τu(x),ξ −τDu(x))u(x)−j (s−τu(x),ξ −τDu(x))·Du(x) s ξ = −L (ξ−τDu(x))·Du(x) ξ −M (s−τu(x),ξ −τDu(x))u(x) s −M (s−τu(x),ξ −τDu(x))·Du(x)+G′(s−τu(x))u(x). ξ GLOBAL COMPACTNESS FOR QUASI-LINEAR PROBLEMS 9 Hence, for a.e. x ∈ Ω and all (s,ξ) ∈ R×RN, it follows that |h(x,s,ξ)| ≤ |L (ξ−τDu(x))||Du(x)|+|M (s−τu(x),ξ −τDu(x))||u(x)| ξ s +|M (s−τu(x),ξ −τDu(x))||Du(x)|+|G′(s−τu(x))||u(x)| ξ ≤ C(|ξ|p−1+|Du(x)|p−1)|Du(x)|+C(|ξ|m+|Du(x)|m)|u(x)| +C(|ξ|m−1+|Du(x)|m−1)|Du(x)|+C(|s|σ−1+|u(x)|σ−1)|u(x)| ≤ ε|ξ|p +C |Du(x)|p +ε|ξ|p +C |Du(x)|p +C |u(x)|p/(p−m) ε ε ε +ε|ξ|m+C |Du(x)|m +ε|s|σ +C |u(x)|σ ε ε = a (x)+ε|s|σ +ε|ξ|p +ε|ξ|m, ε where a :Ω → R is defined a.e. by ε a (x) := C |Du(x)|p +C |Du(x)|m +C |u(x)|p/(p−m) +C |u(x)|σ. ε ε ε ε ε Notice that, as p−1 ≤ m < p−1+p/N it holds p ≤ p/(p−m) ≤ p∗, yielding u ∈ Lp/(p−m)(Ω) and in turn, a ∈ L1(Ω). The assertion follows directly by Lemma 2.5 with µ = 1. (cid:3) ε We have the following splitting result Theorem 3.2. Let the integrand j be as in (1.11) and p−1 ≤m ≤ p−1+p/N, p < σ < p∗. Assume that (u ) ⊂ W1,p(Ω)∩D1,m(Ω) is a bounded Palais-Smale sequence for φ at the level n 0 0 c∈ R weakly convergent to some u ∈W1,p(Ω)∩D1,m(Ω). Then 0 0 |u −u|p |u|p n lim j(u −u,Du −Du)+ V = c− j(u,Du)− V(x) , n n ∞ n→∞(cid:16)ZΩ ZΩ p (cid:17) ZΩ ZΩ p namely lim φ (u −u) = c−φ(u), ∞ n n→∞ being u and u regarded as elements of W1,p(RN)∩D1,m(RN) after extension to zero out of Ω. n Proof. In light of Proposition 2.3, up to a subsequence, (u ) converges weakly to some function n 1,p 1,m u in W (Ω)∩D (Ω), u (x) → u(x) and Du (x) → Du(x) for a.e. x ∈ Ω. Also, recalling that 0 0 n n by assumption V(x) → V as |x| → ∞, we have [4,17] ∞ (3.2) lim V(x)|u −u|p−V |u −u|p = 0, n ∞ n n→∞Z Ω (3.3) lim V(x)|u −u|p−V(x)|u |p+V(x)|u|p = 0. n n n→∞Z Ω Therefore, by virtue of Lemma 3.1, we conclude that |u −u|p n lim φ (u −u)= lim j(u −u,Du −Du)+ V ∞ n n n ∞ n→∞ n→∞(cid:16)ZΩ ZΩ p (cid:17) |u −u|p n = lim j(u −u,Du −Du)+ V(x) n n n→∞(cid:16)ZΩ ZΩ p (cid:17) |u |p |u|p n = lim j(u ,Du )+ V(x) − j(u,Du)− V(x) n n n→∞(cid:16)ZΩ ZΩ p (cid:17) ZΩ ZΩ p = lim φ(u )−φ(u) = c−φ(u), n n→∞ concluding the proof. (cid:3) 10 C.MERCURIANDM.SQUASSINA Remark 3.3. In order to shed some light on the restriction (1.5) of m, it is readily seen that it is a sufficient condition for the following local compactness property to hold. Assume that ω is a smooth domain of Rn with finite measure. Then, if (u ) is a bounded sequence in W1,p(ω), there h 0 exists a subsequence (u ) such that hk Υ(x,u ,Du ) converges strongly to some Υ in W−1,p′(ω), hk hk 0 whereΥ(x,s,ξ) = g(s)−M (s,ξ)−V(x)|s|p−2s. Infact, takingintoaccount thegrowth condition s on g and M , this can be proved observing that, for every ε> 0, there exists C such that s ε N(p−1)+p |Υ(x,s,ξ)| ≤ Cε+ε|s| N−p +ε|ξ|p−1+p/N, for a.e. x ∈ ω and all (s,ξ) ∈ R×RN. 3.2. Equation splitting I (super-quadratic case). We shall assume that m,p ≥ 2 and that conditions (1.7)-(1.8) hold. The following Theorem 3.4 and the forthcoming Theorem 3.5 (see next subsection) are in the spirit of the Brezis-Lieb Lemma [4], in a dual framework. For the particular case |ξ|p M(s,ξ) = 0 and L(ξ) = , p we refer the reader to [12]. Theorem 3.4. Assume that (1.5)-(1.11) hold and that p−1 ≤m < p−1+p/N, p < σ < p∗. Assume that (u ) ⊂ W1,p(Ω)∩D1,m(Ω) is such that u ⇀ u, u → u a.e. in Ω, Du → Du a.e. n 0 0 n n n in Ω and there is (w ) in the dual space (W1,p(Ω)∩D1,m(Ω))∗ such that w → 0 as n → ∞ and, n 0 0 n for all v ∈ W1,p(Ω)∩D1,m(Ω), 0 0 (3.4) j (u ,Du )·Dv+ j (u ,Du )v+ V(x)|u |p−2u v = hw ,vi. ξ n n s n n n n n Z Z Z Ω Ω Ω Then φ′(u) = 0. Moreover, there exists a sequence (ξ ) that goes to zero in (W1,p(Ω)∩D1,m(Ω))∗, n 0 0 such that (3.5) hξ ,vi := j (u −u,Du −Du)v+ j (u −u,Du −Du)·Dv n s n n ξ n n Z Z Ω Ω − j (u ,Du )v− j (u ,Du )·Dv+ j (u,Du)v + j (u,Du)·Dv, s n n ξ n n s ξ Z Z Z Z Ω Ω Ω Ω for all v ∈ W1,p(Ω)∩D1,m(Ω). 0 0 Furthermore, there exists a sequence (ζ ) in (W1,p(Ω)∩D1,m(Ω))∗ such that n 0 0 j (u −u,Du −Du)·Dv+ j (u −u,Du −Du)v+ V |u −u|p−2(u −u)v = hζ ,vi ξ n n s n n ∞ n n n Z Z Z Ω Ω Ω for all v ∈ W1,p(Ω)∩D1,m(Ω) and ζ → 0 as n → ∞, namely φ′ (u −u) → 0 as n → ∞. 0 0 n ∞ n Proof. Fixed some v ∈ W1,p(Ω)∩D1,m(Ω), let us define for a.e. x ∈ Ω and all (s,ξ) ∈ R×RN, 0 0 f (x,s,ξ) := j (s−u(x),ξ −Du(x))v(x) v s +j (s−u(x),ξ −Du(x))·Dv(x)−j (s,ξ)v(x)−j (s,ξ)·Dv(x). ξ s ξ In order to prove 3.5 we are going to show that (3.6) lim sup f (x,u ,Du )−f (x,u,Du) = 0. v n n v n→∞kvkW01,p(Ω)∩D01,m(Ω)≤1(cid:12)(cid:12)(cid:12)ZΩ (cid:12)(cid:12)(cid:12)

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.