
Citation: Francesca G. Alessio, Piero Montecchiari. Gradient Lagrangian systems and semilinear PDE[J]. Mathematics in Engineering, 2021, 3(6): 1-28. doi: 10.3934/mine.2021044
[1] | Juan-Carlos Felipe-Navarro, Tomás Sanz-Perela . Semilinear integro-differential equations, Ⅱ: one-dimensional and saddle-shaped solutions to the Allen-Cahn equation. Mathematics in Engineering, 2021, 3(5): 1-36. doi: 10.3934/mine.2021037 |
[2] | Elena Beretta, M. Cristina Cerutti, Luca Ratti . Lipschitz stable determination of small conductivity inclusions in a semilinear equation from boundary data. Mathematics in Engineering, 2021, 3(1): 1-10. doi: 10.3934/mine.2021003 |
[3] | Luigi Montoro, Berardino Sciunzi . Qualitative properties of solutions to the Dirichlet problem for a Laplace equation involving the Hardy potential with possibly boundary singularity. Mathematics in Engineering, 2023, 5(1): 1-16. doi: 10.3934/mine.2023017 |
[4] | Hitoshi Ishii . The vanishing discount problem for monotone systems of Hamilton-Jacobi equations: a counterexample to the full convergence. Mathematics in Engineering, 2023, 5(4): 1-10. doi: 10.3934/mine.2023072 |
[5] | Filippo Gazzola, Gianmarco Sperone . Remarks on radial symmetry and monotonicity for solutions of semilinear higher order elliptic equations. Mathematics in Engineering, 2022, 4(5): 1-24. doi: 10.3934/mine.2022040 |
[6] | Huyuan Chen, Laurent Véron . Weak solutions of semilinear elliptic equations with Leray-Hardy potentials and measure data. Mathematics in Engineering, 2019, 1(3): 391-418. doi: 10.3934/mine.2019.3.391 |
[7] | Menita Carozza, Luca Esposito, Lorenzo Lamberti . Quasiconvex bulk and surface energies with subquadratic growth. Mathematics in Engineering, 2025, 7(3): 228-263. doi: 10.3934/mine.2025011 |
[8] | Alberto Farina . Some results about semilinear elliptic problems on half-spaces. Mathematics in Engineering, 2020, 2(4): 709-721. doi: 10.3934/mine.2020033 |
[9] | Daniele Castorina, Giovanni Catino, Carlo Mantegazza . A triviality result for semilinear parabolic equations. Mathematics in Engineering, 2022, 4(1): 1-15. doi: 10.3934/mine.2022002 |
[10] | Luca Placidi, Emilio Barchiesi, Francesco dell'Isola, Valerii Maksimov, Anil Misra, Nasrin Rezaei, Angelo Scrofani, Dmitry Timofeev . On a hemi-variational formulation for a 2D elasto-plastic-damage strain gradient solid with granular microstructure. Mathematics in Engineering, 2023, 5(1): 1-24. doi: 10.3934/mine.2023021 |
Dedicated to Alberto Farina on the occasion of his 50th birthday.
We met Alberto, when we were all younger, in 1998 in Paris. Since then we have had the privilege of enjoying his friendship and sharing with him some math. It is a real pleasure to contribute to this volume to celebrate Alberto's fiftieth birthday.
At that time Alberto's research was focused on the study of Liouville type theorems and related symmetry properties for entire solutions of Ginzburg Landau systems or Allen Cahn type equations in RN(see [25,26]). These studies led him to the proof of the celebrated Gibbons conjecture in [27] (see also [28]), proving that
Theorem 1.1. [27] Let N>1 and u∈C2(R2) be a bounded solution to −Δu+u(u2−1)=0 such that limx1→±∞u(x)=±1 uniformly with respect to (x2,…,xN)∈RN−1. Then u(x)=q(x1) on RN where q∈C2(R) is a solution of the ordinary differential equation −¨q+q(q2−1)=0 on R and limt→±∞q(t)=±1.
In [27], Theorem 1.1 has been proved to hold for a wide class of double wells bistable potentials F(u) modeled on that of Ginzburg Landau. The same results had already been obtained in [40] in the cases N=2,3 and then for each N>1, independently of [27], in [16].
This kind of qualitative analysis of "rigidity" results for semilinear elliptic equations remained until now one of the main research themes developed by Alberto. Among others, we refer to the papers [29,30,31,32,33,34,35,36,37,38] where different aspects of this line of study have been deeply investigated for various models, including the Allen Cahn type equations (in relation with the study of the De Giorgi conjecture [24]), Nonlinear Schrödinger equations, quasilinear elliptic equations and Gross–Pitaevskii systems.
These symmetries are generally lost and multiple differently shaped solutions appear when one considers non autonomous or higher dimensional problems, see e.g., [1,2,4,5,6,7,8,9,10,12,23,47,52,53,54,55]. In Section 2 we describe some of these multiplicity results for Allen Cahn models. In particular we emphasize how the Lagrangian structure of the problem allows to obtain infinitely many solutions viewed as orbits of an infinite dimensional Newtonian system connecting disjoint well separated subsets of a fixed sublevel of the related potential. An application of the same method, as done in [11], is described for the Nonlinear Schroedinger equation in Section 3 and, related to this problem, Section 4 is devoted to characterize metric separation properties of sublevels sets of the "potential" functional V(u)=12‖∇u‖2H1(RN)−1p+1‖u‖p+1Lp+1(RN), for different values of the exponent p+1∈(2,2∗N) (see Propositions 4.1 and 4.2).
Consider the Allen Cahn type equation
−Δu(x)+Fu(x,u)=0,x∈RN | (2.1) |
where the potential F∈C2(RN×R) is 1-periodic in the space variables xi, i=1,…,N, and it is of double well type in the variable u, i.e.,
F(x,u)>F(x,±1)=0, Fu,u(x,±1)>0 for x∈RN,u∈R∖{−1,1}. | (2.2) |
Look at the problem of finding solutions u∈C2(Rn) of (2.1) satisfying
limx1→±1u(x)=±1 uniformly with respect to (x2,…,xn)∈RN−1. | (2.3) |
Rabinowitz and Stredulinsky in [52,53] proved that problem (2.1)–(2.3) generically admits a rich variety of geometrically different solutions connected to each other. They first showed by variational methods the existence of minimal solutions of (2.1)–(2.3) which are 1-periodic with respect to the variables x2,…,xN. Moreover, the set of such minimal solutions, M0, is an ordered set in the sense that if u0,v0∈M0 then either u0>v0 or u0=v0 or u0<v0 on RN. Hence an alternative occurs, either
i) for any y∈(−1,1) there is x0∈RN and u∈M0 such that u(x0)=y (the graphs of the minimal solutions foliate RN×(−1,1)); or
ii) M0 has a gap pair: there exist u0<v0∈M0 such that no member of M0 lies between u0 and v0.
Note that (i) is always satisfied in the autonomous case, as for F(u)=14(u2−1)2 like in Theorem 1.1, since the problem is invariant by translation. On the other hand, as proved by Rabinowitz and Stredulinski, the case (ii) occurs generically for periodic potentials F satisfying (2.2) (see [54]).
If case (ii) occurs other solutions appear. If u0<v0∈M0 is a gap pair, they proved the existence of another ordered family M1 of minimal solutions of (2.1)–(2.3), between u0 and v0, which are asymptotic as x2→±∞ to the elements of the gap and are 1-periodic in the remaining variable x3,…,xN (see Figure 1).
If the ordered set M1 has itself a gap pair, u1<v1, the argument can be repeated to find ordered classes of minimal solutions connecting u1 and v1 and the reasoning can be iterated to obtain higher dimensional connections. When gaps are present in M0, variational gluing techniques were moreover employed to construct multitransition type solutions of (2.1)–(2.3), solutions which oscillate with respect to the variable x2, in a prescribed fashion, between the elements of the gap, being periodic in the variables x3,…,xN (see the monograph [54] for an exposition of such kind of results in connection with the Moser-Bangert Theory [14,48]).
Unlike the rigid character of Allen Cahn's autonomous equation, as described in Theorem 1.1, these studies show how the solutions set of (2.1)–(2.3) may be complex when the potential explicitly depends on the spatial variables.
The contrast between these results and the Gibbons conjecture is even more explicit considering potentials F depending (1-periodically) only on the space variable x1. In this case any solution q∈C2(R) of the heteroclinic one dimensional problem
−¨q+Fq(t,q)=0,q(±∞)=±1, | (2.4) |
generates a solutions u(x)=q(x1) of (2.1)–(2.3). The set M0 corresponds here to the set of minimal solutions of (2.4), i.e., the set
K={q∈Γ∣V(q)=c≡infΓV} |
where
Γ={q∈W1,2loc(R)∣q(±∞)=±1} and V(q)=∫R12|˙q|2+F(t,q)dt. |
The minimal set K is ordered and if it has a gap pair q0<q1 the above results give the existence of u∈C2(RN) solution to −Δu+Fu(x1,u)=0 on RN satisfying (2.2) and being asymptotic, as x2→±∞, to q0 and q1 (not depending on the others variable x3,…,xN). Then, even if the nonlinearity depends only on the variable x1, differently from the autonomous case, problem (2.1)–(2.3) has solutions which do not come from the one dimensional problem (2.4).
This result was originally obtained with L. Jeanjean in [4] (see also [5]), pointing out how Theorem 1.1 is a particular feature of the autonomous Allen-Cahn stationary equation. In [4] a factorized case was considered, F(t,u)=a(t)W(u) with a positive and periodic and W a double well potential, but the arguments does not change for general potential F(x1,u) satisfying (2.2). It is worth mentioning that in [4], the ordered structure of the minimal set K was not used and the above gap condition, (ii), was formulated in terms of a separation property of the minimal set K:
(∗) there exists K0⊂K such that, setting Kj={q(⋅−j)|q∈K0} for j∈Z, there result
(i) K0 is compact with respect to the W1,2(R,RN) metric,
(ii) there exists ρ0>0 such that if i≠j then distL2(R)(Ki,Kj)≥ρ0,
(iii) K=∪j∈ZKj.
Thanks to the ordering property of K and its invariance with respect to integer translations, the existence of a gap pair in K is in fact equivalent to the property (∗). In [4] are given examples of singularly perturbed cases verifying (∗) and, moreover, it has been proved that (∗) does not hold if and only if K is a continuum with respect to the W1,2(R) metric, homeomorphic to R. It should be interesting to understand, in relation with the Gibbons conjecture, what happens when (∗) is not satisfied (see [46] for a related study).
The work in [4] was motivated by the paper of S. Alama, L. Bronsard and C. Gui, [1], where they study a vectorial version of the Allen Cahn model. More precisely for potentials F∈C2(R2) satisfying
(F1) F((±1,0))=0, F(ξ)>0 for every ξ∈R2∖{(±1,0)}, D2F((±1,0)) are definite positive and there exists R>0 such that ∇F(ξ)ξ≥0 for |ξ|>R;
(F2) F(−ξ1,ξ2)=F(ξ1,ξ2) for any (ξ1,ξ2)∈R2,
they look for the existence of solutions u∈C2(R2,R2) to the problem
−Δu(x,y)+∇F(u(x,y))=0,(x,y)∈R2, | (2.5) |
limx→±∞u(x,y)=(±1,0)uniformly w.r.t. y∈R. | (2.6) |
Again, in their study, an important role is played by the structure of the set of one dimensional minimal solutions to (2.5) and (2.6), Ks, the set of minima of the action V(q)=∫R12|˙q|2+F(q)dx on the class of symmetric functions
Γs={q∈W1,2loc(R,R2)∣q(±∞)=(±1,0) and q1(−x)=−q1(x)}. | (2.7) |
In [1] it is proved that Ks≠∅ and if
(∗∗) Ks={q1,…,qk} with k≥2 and qi≠qj when i≠j,
then there exists a solution u∈C2(R2,R2) of (2.5)–(2.6) which is asymptotic as y→±∞ to a pair of different elements in Ks, showing the failing of Gibbons-like results for autonomous systems.
The symmetry assumption (F2) is used in [1] to gain compactness in the problem working with the symmetric functions in Γs. That assumption was later dropped in [55], by M. Schatzman, showing how it is possible to overcome the lack of compactness assuming a nondegeneracy property of the one dimensional heteroclinic connections. Precisely, letting
S={q∈W1,2loc(R,R2)|q(±∞)=(±1,0)}, and K={q∈S∣V(q)=c=infSV} |
in [55] it is assumed that
(∗∗∗) there exists z−≠z+∈K such that
(i) K decomposes in the disjoint union of the sets of the translated of z− and z+:
K=C(z−)∪C(z+) where C(z±)={z±(⋅−s)∣s∈R}; |
(ii) the operators A±:H2(R,R2)⊂L2(R,R2)→L2(R,R2), A±h=−¨h+D2W(z±)h, are such that Ker(A±)=span{˙z±}.
The non degeneracy condition (∗∗∗)-(ii) allows to gain compactness in the directions orthogonal to C(z±), and in [55] this is used to obtain the existence of a solution u∈C2(R2,R2) of (2.5)–(2.6) such that limy→±∞u(x,y)=z±(x−s±) for certain values s±∈R.
The results in [1,4,55] have been obtained by variational methods. Differently from [1], where a limit procedure throught domains in R2 bounded in the y direction was employed, in [4] the two dimensional solutions in M1 were found as minima of a global functional.
It is useful for what follows to give a brief description of the variational framework used in [4]. Recalling the assumption (∗), solutions of (2.1)–(2.3), which are not "one dimensional", have been searched as minima of the (renormalized) functional
φ(u)=∫R[∫R12|∇u(x,y)|2+F(x,u(x,y))dx−c]dy |
on the set
Γ2,j={u∈H1loc(R2)|u(⋅,y)∈Γ for a.e. y∈R,limy→−∞distL2(R)(u(⋅,y),K0)=limy→+∞distL2(R)(u(⋅,y),Kj)=0}, j∈Z∖{0}. |
Using Fubini Theorem, since F depends only on the variable x, the functional φ can be written in the more enlighting form
φ(u)=∫R12‖∂yu(⋅,y)‖2L2(R)+(V(u(⋅,y))−c)dy | (2.8) |
which reveals the Lagrangian structure of the problem. The cyclic variable y plays the role of an evolution variable and the elements u∈Γ2,j can be seen as test trajectories, y↦u(⋅,y)∈Γ, connecting two different minimal sets, K0 and Kj, of the "potential" V(u(⋅,y))−c.
This point of view led in [4] to the use of minimization techniques in the spirit of earlier papers on the heteroclinic problem for Hamiltonian systems and PDE (see [3,50,51]) to show that at least for two different values of j∈Z∖{0}, φ attains the minimum in Γ2,j. After [52], we know that the ordered structure of K implies that the values of j have to be equal to ±1 and the solutions found in [4] are in fact elements of the set M1 described above.
Note that, both for a single equation with nonlinearities F depending only on the variable x1 as in [4] and for the case of autonomous systems as in [1,55], the set M1 of minimal double heteroclinic connections is invariant with respect to y translations. Hence M1 has not "gap-like" properties and one cannot hope to use the arguments in [52,53] to obtain further solutions (like higher dimensional connections or multitransition type solutions) of this kind of problems.
Pursuing the study of the case F=F(x1,u), a different method, based on Hamiltonian identities due the conservation of the energy conjugated to the cyclic variable y, was then developed in [6] showing that, when (∗) holds, (2.1)–(2.3) has other (infinitely many) two dimensional solutions, heteroclinic or homoclinic in the variable y to (locally minimal) solutions of (2.4) not belonging to K. The method was refined and adapted in [8] to show the existence of infinitely many other solutions to (2.1)–(2.3), called brake orbit type solutions, which have a periodic behaviour in the variable y (see also [7] where the case of a potential F=F(x1,u) depending in an almost periodic way on x1 is studied). The same approach was used by F. Alessio in [2] to obtain analogous multiplicity results for the problem (2.5)–(2.6), relative to the Allen Cahn type systems considered in [1]. In [12] we finally relaxed the symmetric assumption (F2) obtaining the same kind of solutions in the asymmetric setting considered by M. Schatzman in [55].
We give here below a brief illustration of these "energy constrained" methods describing with some details the main ideas in the setting studied in [2].
The fist step in [2] is to weaken the finiteness assumption (∗∗) on the minimal set Ks={q∈Γs∣V(q)=c≡infΓsV} made in [1]. Indeed, in analogy with the assumption (∗) in [4], in [2] a weak separation property on Ks is considered:
(∗c) Ks=K+∪K− with K±≠∅ and distL2(R,R2)(K+,K−)>0. |
In Theorem 1.1 of [2], a first strenghtening of the result in [1] is given showing that if (F1) (F2) and (∗c) are satisfied then the problem (2.5)–(2.6) admits a bidimensional solution U∈C2(R2,R2) satisfying
limy→±∞distL2(R,R2)(U(⋅,y),K±)=0. | (2.9) |
As in [4] this solution is obtained as a minimizer of the renormalized functional
φ(u)=∫R12‖∂yu(⋅,y)‖2L2(R,R2)+(V(u(⋅,y))−c)dy. |
Note that φ is well defined and not negative on the space
H={u∈H1loc(R2,R2)|u(⋅,y)∈Γs for almost every y∈R}. |
Since φ(q)=0 for all q∈Ks, the elements of Ks are global minimizers of φ on H. The bidimensional solution satisfying (2.9) is obtained by minimizing φ on the space
Hc={u∈H∣lim infy→±∞distL2(R,R2)(u(⋅,y),K±)=0}. |
Roughly, the element in H can be seen as trajectories y∈R↦u(⋅,y)∈Γs (continuous with respect to the L2(R,R2) metric) and the bidimensional solution U as a minimal trajectory in Hc satisfying the infinite dimensional gradient system
∂2yu(⋅,y)=−∂2xu(⋅,y)+∇F(u(⋅,y))⏟V′(u(⋅,y)),u∈H. | (2.10) |
The one dimensional action V plays the role of a potential in (2.10) and its minimal sets, K±, are constituted by equilibria of the equation. Then U is an heteroclinic connection between the two sets.
The new step is now to observe (as done analogously in [6,7,8]) that since (2.10) is autonomous, an energy is conserved along its trajectories. More precisely, by Lemma 4.3.10 in [2], if u∈H solves (2.5) on R×(y1,y2) then
Eu(y)=12‖∂yu(⋅,y)‖2L2(R,R2)−V(u(⋅,y)) |
is constant on (y1,y2) (see [41] for more general identities of this kind). In particular EU(y)=−c for every y∈R and a different way to characterize the solution U is as a solution with energy equal to −c connecting as y→±∞ the disjoint subsets K± of the level set {q∈Γ|V(q)≤c}.
The idea is to look for new solutions by varying the value of the energy and looking for solutions connecting disjoint subsets of the correponding sublevel of the potential. Indeed if λ>0 is small enough, by (∗c) we obtain that for any b∈(c,c+λ),
(∗b) {q∈Γs|V(q)≤b}=Vb−∪Vb+ with Vb±≠∅ and distL2(Vb−,Vb+)>0. |
The second problem considered in [2] is to show that when (∗b) holds, there are solutions u∈H to (2.5) with Energy Eu=−b and "connecting" the sets Vb− and Vb+. The basic remark in the argument is that if u is such a solution then
V(u(⋅,y))=−Eu(y)+12‖∂yu(⋅,y)‖2L2(R,R2)≥b for every y∈R |
and the condition V(u(⋅,y))≥b becomes a natural constraint in such a problem. Then, in [2] these solutions are searched by minimizing the new renormalized functional
φb(u)=∫R12‖∂yu(⋅,y)‖2L2(R)2+(V(u(⋅,y))−b)dy |
on the space
Hb={u∈H|lim infy→±∞distL2(u(⋅,y),Vb±)=0 and V(u(⋅,y))≥b for a.e. y∈R}. |
Note that, differently for the minimum problem for φc on Hc, when b>c, minimizing φb on Hb is a constrained problem. The advantage to require V(u(⋅,y))≥b for y∈R is that this allows to say that φb is well defined and positive on Hb. On the other hand, if u is a minimum of φb on Hb we have that u solves the equations in (2.5) on a strip R×(y1,y2) only when the constraint V(u(⋅,y))≥b is satified with a strict inequality on (y1,y2).
In [2] it is shown the existence of a minimizing sequences of φb on Hb weakly convergent to a function ub∈H for which there exists a transition interval (σb,τb)⊂R such that
(1) if [y1,y2]⊂(σb,τb) then infy∈[y1,y2]V(ub(⋅,y))>b,
(2) lim infy→σ+bdistL2(R,R2)(ub(⋅,y),Vb−)=lim infy→τ−bdistL2(R,R2)(ub(⋅,y),Vb+)=0.
By (1), the minimality property of ub guarantees that ub satisfies the equations in (2.5) on R×(σb,τb). By regularity, using (2), one recovers that V(ub(⋅,y))→b whenever y→σ+b or y→τ−b. Again the minimality property of ub gives that Eub=−b for all y∈(σb,τb) and so that
‖∂ub(⋅,y)‖L2(R,R2)=2(V(u(⋅,y))+Eub)→0 as y→σ+b or y→τ−b. | (2.11) |
If σb=−∞ and τb=+∞, ub is actually a solution of (2.5)–(2.6) connecting by (2) the sets Vb±. When σb∈R or τb∈R, it is possible to extend ub|R×[σb,τb] on R2 by reflection in the variable y with respect to σb or τb and then by periodic continuation. By (2.11) one again obtains a solution vb of (2.5)–(2.6) connecting Vb± with energy equal to −b.
The behaviour of these solutions can be characterized more precisely depending on the boundedness properties of the transition interval (σb,τb).
When σb=−∞ (resp. τb=+∞), in [2] it is shown that the α-limit (resp. ω-limit) of ub is a set K−(b)⊂Vb− (resp. K+(b)⊂Vb+) of {critical points} of V at level b and in particular b is a critical value of V.
Then, if σb=−∞ and τb=+∞ we have that vb=ub is a solution of heteroclinic type connecting K±(b) as y→±∞.
If σb=−∞ and τb∈R, vb is equal to ub on R×(−∞,τb] and symmetric in the variable y with respect to τb. It defined a homoclinic type solution of (2.5)–(2.6) which emanates from K−(b) as y→−∞, reachs Vb+ for y=τb and then symmetrically returns to K−(b) as y→+∞. Analogously, when σb∈R and τb=+∞, vb is a solution to (2.5)–(2.6) homoclinic to K+(b) and symmetric with respect to y=σb.
Finally, if both σb,τb∈R, vb(x,y) is equal to ub on R×[σb,τb] being symmetric in y with respect to both σb and τb, and so periodic, with period Tb=2(τb−σb). We say that vb is a brake orbit type solution of (2.5)–(2.6) defining a periodic trajectory of period Tb which oscillates back and forth in H along a simple curve connecting the two turning points vb(⋅,σb)∈Vb− and vb(⋅,τb)∈Vb+.
By the above observations, when b is a regular value of V, the corresponding solution has to be of brake orbit type. As in Lemma 4.2.9 in [8], as a consequence of the Sard Smale Theorem, it can be proved that the set of regular values of V is open and dense in [c,c+λ). In this sense we can say that generically for b∈[c,c+λ) the corresponding solution is periodic in y, of the brake orbit type. If otherwise b∈[c,c+λ) is a critical value of V, then the corresponding connecting orbit may exhibit homoclinic or heteroclinic behaviour.
We refer to the paper with A. Zuniga [13] where the above described "constrained energy method" is applied to obtain analogous classes of connecting type solutions in the simpler context of ODE gradient systems of the form
¨q=∇W(q),q∈RN, |
including applications to classical cases as double well, Duffing or pendulum like potentials.
Other more recent papers obtained multiplicity results regarding periodic solutions for Allen Cahn systems or different type of equations with related but different methods, based again on the Lagrangian structure of the problem. We quote the nice papers by G. Fusco, G. F. Gronchi, M. Novaga [23], A. Cesaroni and M. Cirant [21], A. Monteil, F. Santambrogio [47] and the reference therein. In particular in [47] a general study on the geodesic problem in metric space is done. As an application the result in [1] is recovered and the heteroclinic connection between two isolated minima q± of V is obtained as a minimal geodesic relative to the Maupertius Jacobi distance
dJ(q−,q+)=infγ∈Γ(z−,z+)∫[0,1]√2(V(γ(t))−c)‖˙γ(t)‖L2(R,R2)dt. | (2.12) |
Here Γ(z−,z+) is the set of path γ:[0,1]→Γs which are piecewise locally absolutely continuous with respect to the L2(R,R2) metric (or, by density, the W1,2(R,R2) metric) with γ(0)=z− and γ(1)=z+. For the use of the Maupertius principle in the study of the periodic or the heteroclinic problem for gradient ODE systems we refer to the paper by V. Benci [15] and the one by P. Sternberg and A. Zuniga [57].
To end this section we only briefly mention for the interested readers the papers [9] and [10] where the problem of finding solutions depending on more than two variables was studied, with different methods, both for the problem with F=F(x1,u) considered in [4] or the system case considered in [1].
In this Section we describe some applications of the above illustrated methods to non linear Schrödinger equations. Even if the geometry of the problem is in this case very different with respect to the one relative to the Allen Cahn models, as done in [11] the constrained energy method applies also in this context to find and characterize families of connecting orbit type solutions, at least for a restricted range of values of the exponent p.
In [11] we studied equations of the form
−Δv(x,y)+v(x,y)=f(v(x,y)),(x,y)∈RN×R | (3.1) |
where N≥1 and the nonlinearity f is modeled on the power case f(v)=|v|p−1v with p+1 subcritical and greater than 2. This kind of equations is widely used in physical models, in particular in the study of standing waves solutions of the corresponding nonlinear Schrödinger type equations (see [18]).
The problem of finding positive solutions v∈H1(RN+1) of (E) has been deeply studied and the existence of radial solutions was first obtained by W. A. Strauss in [58]. Nearly optimal existence results of least energy solutions for (E) was given by H. Berestycki and P. L. Lions in [18] (see also [19]). In the pure power case, uniqueness and non degeneracy properties of solutions of (E) in H1(RN+1) was derived by M. K. Kwong in [44] and by J. Serrin and M. Tang for more general nonlinearity in [56].
In the seminal work by N. Dancer, [22], it is studied a new class of entire solutions of (E) in the pure power case. The starting idea is to see the ground state solution u0(x) of (E) in RN (a mountain pass type solution) as a solution on RN+1, constant with respect to the y variable. In the pure power case, when the ground state is nondegenerate, Dancer used bifurcation and continuation arguments to get the existence of a continuous branch of entire positive solutions of (E) in RN+1 bifurcating from the cylindric type solution u0. These solutions are periodic in the variable y and decay to zero as |x|→+∞. See Figure 2.
Analogous periodic solutions were found and used, again in the pure power case, as prescribed asymptotes in the constructions of multiple ends solutions by A. Malchiodi in [45]. These periodic solutions have large periods and, differently from the Dancer result, emanate via the Implicit Function Theorem, from the mountain pass solutions of (E) in RN+1. See Figure 3.
Different qualitative properties of positive solutions v(x,y) of (E) which decay to zero as |x|→+∞, such as radial symmetry with respect to the variable x, have been described by C. Gui, A. Malchiodi and H. Xu in [42] and by A. Farina, A. Malchiodi and M. Rizzi in [36]. In these studies, based on moving plane techniques, is again involved the use of Hamiltonian identities, such as a conservation of energy along the cylindrical variable y.
Indeed, as in the preceding section for the Allen Cahn models, prescribing the decay properties of a solution only with respect to the variable x∈RN, naturally gives to y the role of an evolution variable. The periodic solutions described above belong to the space
H=L2loc(R,H1(RN))∩H1loc(R,L2(RN)) |
and in analogy with (2.10)verify (at least in a weak sense) the Newtonian evolution equation
∂2yv(⋅,y)=−Δxv(⋅,y)+f(v(⋅,y))⏟V′(v(⋅,y)),y∈R. | (3.2) |
In (3.2), V′ is the gradient of the Euler functional relative to the problem in RN,
V(u)=∫RN12|∇u|2+12|u|2−F(u)dx,u∈H1(RN), |
where F(s)=∫s0f(t)dt. We name layered solution any v∈H which solves (E).
The Eq (3.2) has a Lagrangian structure and the functional U=−V plays the role of the energy potential. In this connection, any u∈H1(RN) which solves (E) is an equilibrium of (3.2) and the Dancer's or Malchiodi's solutions are periodic orbits of the system. As for the Allen Cahn models, if v is a layered solution then the Energy function
y↦Ev(y)=12‖∂yv(⋅,y)‖2L2(RN)−V(v(⋅,y)) |
is constant.
In [11], we applied the constrained energy method to find connecting orbit type layered solution of (E) with prescribed energy. This was done under slightly general assumptions on the non linearity f. It was assumed that
(f1) f∈C1(R),
(f2) there exists C>0 and p∈(1,1+4N) such that |f(t)|≤C(1+|t|p) for any t∈R,
(f3) there exists μ>2 such that 0<μF(t)≤f(t)t for any t≠0, where F(t)=∫t0f(s)ds,
(f4) f(t)t<f′(t)t2 for any t≠0.
By (f1)−(f4) the above defined functional V∈C1(H1(RN)) and satisfies the geometrical assumptions of the mountain pass Theorem. Considering the mountain pass level
c=infγ∈Γsupt∈[0,1]V(γ(t)), |
where
Γ={γ∈C([0,1],H1(RN))|γ(0)=0,V(γ(1))<0}, |
we have that c>0 is an asymptotical critical level for V. As it is well known c is the lowest positive critical level of V (see [43]). Using the definition of the mountain pass level, the assumptions on f are sufficient to show that for any b∈[0,c) the sublevel set {u∈H1(RN)∣V≤b} splits into the disjoint union of two connected nonempty subsets:
{u∈H1(RN)∣V(u)≤b}=Vb−∪Vb+ with Vb−∩Vb+=∅, | (3.3) |
where Vb− denotes the component which contains 0. In [11] we adapt the constrained energy method described in the previous section to obtain for any b∈[0,c), layered solution vb of (E) with energy Evb=−b "connecting" the sets Vb±. Since, as remarked above, c is the lowest positive critical level of V, any b∈(0,c) is a regular value of V and as discussed in the previous section, the corresponding solution vb is a {brake orbit type} periodic layered solution of (E). Theorem 1.1 in [11] precisely states the following
Theorem 3.1. [11] If f satisfies (f1)−(f4) then for any b∈[0,c) the equation (E) has a solution vb∈C2(RN+1) with energy Evb=−b and such that
(i) vb>0 on RN+1,
(ii) vb(x,y)=vb(|x|,y)→0 as |x|→+∞, uniformly w.r.t. y∈R,
(iii) ∂rvb(x,y)<0 for r=|x|>0 and y∈R.
Moreover, if b>0,
(iv) there exists Tb>0 such that vb is periodic of period 2Tb in the variable y and symmetric with respect to y=0 and y=Tb.
(v) ∂yvb(x,y)>0 on RN×(0,Tb), vb(⋅,0)∈Vb−, vb(⋅,Tb)∈Vb+.
Finally, if b=0,
(vi) v0∈H1(RN+1) is radially symmetric and ∂rv0<0 for r=|(x,y)|>0,
(vii) v0(⋅,0)∈V0+ and v0 is a mountain pass point of the Euler funcional relative to (E) on H1(RN+1).
Theorem 3.1 provides for any b∈[0,c) a positive layered solution vb to (E) with energy −b. With respect to the variable x it is radially symmetric and decay to 0 as |x|→+∞ uniformly with respect to y∈R. When b>0 the solution vb is a periodic solution of period 2Tb which is symmetric with respect to y=0 and y=Tb and has the beaviour of a brake orbit type solution. Its trajectory y→vb(⋅,y)∈H1(RN) oscillates back and forth along a simple curve connecting the two turning points vb(⋅,0)∈Vb− and vb(⋅,Tb)∈Vb+. These solutions are clearly related to the Dancer's and Malchiodi's solutions. When b=0 the solution v0 belongs to H1(RN+1) and defines a homoclinic type layered solution to 0∈H1(RN). By (Vii), v0 is a mountain pass point of the Action functional to (E) on H1(RN+1). On the other hand the mountain pass point of V in H1(RN) is an equilibrium of (3.2) at energy −c. The Energy diagram in Figure 4 wish to summarize these considerations.
The application of the constrained energy method in this case is much more complicated than that for the Allen Cahn models, due to strong lack of compactness and weak semicontinuity of the problem. This comes from the competition among the terms ‖u‖2H1(RN) and ∫RNF(u) which enter in the potential V(u) with different sign and explains why we assume in (f2) that p<1+4/N.
The exponent p=1+4/N=2G,N−1, where 2G,N has different critical properties due to the Gagliardo Niremberg inequalities (see [39,49]), see (4.7) and the related observations in the next section. We recall that 2G,N is critical for the minimum problem inf{V(u)|u∈H1(RN),‖u‖L2(RN)=1} and with respect to the property of orbital stability of the solutions of (E) in H1(RN). The ground state solution of (E) in H1(RN) is stable if and only if p+1∈(2,2G,N), see [17,20,59].
When p+1<2G,N the functional V exhibits the coerciveness and semicontinuous properties sufficient to recover some of the basic properties used applying the constrained energy method for the Allen Cahn model. In particular note that the exponent 2G,N is critical with respect to the separation of the sets Vb± with respect to the L2(RN) metric (see proposition 1 in the next section). Assuming p+1<2G,N leads to a situation like the one in (∗b) for the Allen Cahn models where an L2 separation between the sets Vb± is used.
We think however that the condition p+1<2G,N is only technical and that a refined analisys can lead to apply the constrained energy method to cover all the subcritical cases 2<p+1<2∗N+1 (where as usual we denote the critical Sobolev exponent 2∗n=2n/(n−2) for n>2 and 2∗n=+∞ for n=1,2). This will be the object of future studies. In the next section, as a fist step in this program, we characterize the metric separation properties of the sets Vb± in relation to the value of the exponent p+1∈(2,2∗N).
A basic point in applying the constrained energy method to look for layered solutions of (E) connecting the sets Vb±, is to well understand how and with respect to which metrics the sets Vb± are separated, depending on the values of the exponent p. The present more technical section is devoted to study this problem. This will be done relatively to the L2(RN) distance and a Jacobi type metric (see (2.12) for the Allen Cahn model), which are naturally related with the constrained energy approach. In fact we show that while distL2(RN)(Vb−,Vb+)>0 if and only if p+1≤2G,N we have that the two sets have positive Jacobi distance dJ(Vb−,Vb+) (see (4.19)) if and only if p+1≤2∗N+1. For the sake of simplicity we will do this study in the power case F(u)=1p+1|u|p+1.
We assume N≥1 and 1<p<2∗N−1 when N>1 while p>1 if N=1. We work on the space of radially symmetric functions
X=H1r(RN)={u∈H1(RN)∣u(x)=u(|x|) on RN} |
endowed with its standard norm and scalar product
‖u‖=(∫RN|∇u|2+u2dx)12,⟨u,v⟩=∫RN(∇u∇v+uv)dx. |
For q≥1 we denote with ‖u‖q the usual norm of the function u in the space Lq(RN). We recall that by the Strauss Lemma (see [58]) X is compactly embedded in Lq(RN) for q∈(2,2∗N).
We study the functional
V(u)=12‖u‖2−1p+1‖u‖p+1p+1,u∈X. |
As it is well known V∈C1(X,R) with
V′(u)v=⟨u,v⟩−∫RN|u(x)|p−1u(x)v(x)dx. |
Thanks to the compact embedding of X in Lp+1(RN) and the weakly semicontinuity of the L2-norm, the functional V is weakly semicontinuous on X. In fact, we have
Lemma 4.1. Let un→u and vn→v weakly in X. Then
(i) limn→+∞∫RN|un(x)|p+1dx=∫RN|u(x)|p+1dx
(ii) limn→+∞∫RN|un(x)|p−1un(x)vn(x)dx=∫RN|u(x)|p−1u(x)v(x)dx
In particular V(u)≤lim infn→+∞V(un) and V′(u)h=limn→+∞V′(un)h for any h∈X.
As it is well known, the functional V satisfies the geometrical hypotheses of the mountain pass theorem. Indeed, since p+1<2∗N, by the Sobolev immersion Theorem we have
infu∈H1(RN)∖{0}‖u‖‖u‖p+1>0, |
and so
V(u)=12‖u‖2+o(‖u‖2) and V′(u)u=‖u‖2+o(‖u‖2) as ‖u‖→0. | (4.1) |
In particular there exists ρ>0 such that
V(u)≥14‖u‖2 and V′(u)u≥12‖u‖2 if ‖u‖≤ρ. | (4.2) |
Moreover, if u∈X∖{0} then
V(su)=s22‖u‖2−sp+1p+1‖u‖p+1p+1→−∞ as s→+∞. |
Defining
Γ={γ∈C([0,1],X): γ(0)=0, γ(1)≠0 and V(γ(1))≤0} |
and c=infγ∈Γmaxs∈[0,1]V(γ(s)), by the mountain Pass Theorem we infer that c≥14ρ2 and that there exists a Palais Smale sequence for V at level c. Since
(p+1)V(u)−V′(u)u=p−12‖u‖2 | (4.3) |
the Palais Smale sequences of V are bounded in X. Another plain consequence of (4.3) is that
if V′(u)u=0 then V(u)=p−12(p+1)‖u‖2 | (4.4) |
The existence of a critical point of V at level c is then easily deduced.
Theorem 4.1. There exists u0∈X∖{0} such that V′(u0)=0 and V(u0)=c.
Proof. By the mountain Pass Theorem and (4.3) there exists a bounded sequence {un}⊂X such that V(un)→c and V′(un)→0. Up to subsequence, un→u0 weakly in X. By Lemma 4.1 we deduce that V′(u0)=0 and V(u0)≤c. To show that V(u0)=c observe that V′(un)un→0 and ‖un‖p+1→‖u0‖p+1 imply ‖un‖2→‖u0‖p+1p+1. Since V′(u0)u0=0 we also have ‖u0‖2=‖u0‖p+1p+1. Then ‖un‖2→‖u0‖2 and so V(un)→12‖u0‖2−1p+1‖u0‖p+1p+1=V(u0). Hence V(u0)=c.
We characterise here some properties of the sublevel sets
Vb={u∈X/V(u)≤b} |
for level values b∈[0,c).
We will make use of the Gagliardo Nirenberg inequalities (see [39,49]) which in particular imply
‖u‖p+1≤KG‖u‖θN,p2‖∇u‖1−θN,p2 for any u∈X, | (4.5) |
where KG>0 and
θN,p=1−N2p−1p+1. | (4.6) |
By (4.5) the functional V has good coercivity properties on the bounded sets of L2(RN) when p+1<2G,N where we set
2G,N≡2+4N. | (4.7) |
Indeed by (4.6), if p+1<2G,N then
(1−θN,p)(p+1)=N2(p−1)<2. | (4.8) |
Then, by (4.5) and (4.8) we obtain that if ‖∇un‖2→+∞ and ‖un‖2≤C<+∞ then
V(un)=12‖un‖2−1p+1‖u‖p+1p+1≥12‖∇un‖22+12‖u‖22−1p+1Kp+1GCθN,p(p+1)‖∇un‖(1−θN,p)(p+1)2→+∞. |
It is useful to describe the behaviour of V along the rays in X starting from the origin.
Lemma 4.2. For any u∈X∖{0} setting tu=(‖u‖2‖u‖p+1p+1)1p−1 we have V′(tuu)u=0 and
V′(tu)u>0 for t∈(0,tu) and V′(tu)u<0 for t∈(tu,+∞). | (4.9) |
Moreover, for u∈X∖{0} we have
V(tuu)=(12−1p+1)(‖u‖‖u‖p+1)2(p+1)p−1≥(12−1p+1)(1KG)2(p+1)p−1[(‖u‖2‖∇u‖2)N+(‖∇u‖2‖u‖2)2(p+1)p−1−N]. | (4.10) |
Proof. The first part of the lemma is obtained since V(tu)=12t2‖u‖2−tp+1p+1‖u‖p+1p+1 and so ddtV(tu)=t(‖u‖2−tp−1‖u‖p+1p+1) for any t≥0. To derive the equality in (4.10) it is enough to perform the computation:
V(tuu)=12(‖u‖2‖u‖p+1p+1)2p−1‖u‖2−1p+1(‖u‖2‖u‖p+1p+1)p+1p−1‖u‖p+1p+1=(‖u‖2‖u‖p+1p+1)2p−1(12−1p+1)‖u‖2=(12−1p+1)(‖u‖‖u‖p+1)2(p+1)p−1 |
Finally the inequality in (4.10) follows by (4.5)–(4.6), which imply
(‖u‖‖u‖p+1)2(p+1)p−1≥(1KG)2(p+1)p−1(‖u‖22+‖∇u‖22‖u‖2θN,p2‖∇u‖2(1−θN,p)2)p+1p−1=(1KG)2(p+1)p−1[(‖u‖2‖∇u‖2)2(1−θN,p)+(‖∇u‖2‖u‖2)2θN,p]p+1p−1≥(1KG)2(p+1)p−1[(‖u‖2‖∇u‖2)2(1−θN,p)p+1p−1+(‖∇u‖2‖u‖2)2θN,pp+1p−1]=(1KG)2(p+1)p−1[(‖u‖2‖∇u‖2)N+(‖∇u‖‖u‖2)2(p+1)p−1−N], |
and the Lemma is proved.
Remark 4.1 By continuity and Lemma 4.2, for any u∈X∖{0} and b∈[0,c), there exist unique αu,b∈(0,tu) and ωu,b∈(tu,+∞) such that V(αu,bu)=V(ωu,bu)=b. Note that αu,b is strictly increasing and ωu,b strictly decreasing for b∈[0,c). By definition V(ωu,0u)=ω2u,0(12‖u‖2−1p+1ωp−1u,0‖u‖p+1p+1)=0. Hence
ωu,0=(p+12‖u‖2‖u‖p+1p+1)1p−1=(p+12)1p−1tu, |
and so for any b∈[0,c) we have
0=αu,0<αu,b<tu<ωu,b<ωu,0=(p+12)1p−1tu | (4.11) |
Hence, nothing that ddsV(su)=V′(su)u, for any b∈[0,c), if u∈X∖{0}, there results that if t∈(0,αu,b]⊂(0,tu) then V′(tu)u>0 and V(tu)≤V(αu,bu)=b. Analogously, if t∈[ωu,b,+∞) then V′(tu)u<0 and V(tu)≤b. Moreover, if t∈(αu,b,ωu,b), then V(tu)>b.
Therefore, if u∈X∖{0} and V(u)≤b, we have either αu,b≥1 and hence V′(u)u>0, or ωu,b≤1 from which V′(u)u<0. $
We consider the sublevel sets of V
Vb={u∈X/V(u)≤b}. |
By definition of the mountain pass level the set Vb is not path connected for any b∈[0,c). Given b∈[0,c), recalling Remark 4.1, we denote
Vb−={tu|u∈X∖{0},t∈[0,αu,b]} and Vb+={tu|u∈X∖{0},t∈[ωu,b,+∞)} |
and observe that, by Remark 4.1,
Vb=Vb−∪Vb+ with Vb−∩Vb+=∅ and Vb±≠∅. |
Note for all b∈[0,c) we have u∈Vb− if 1∈(0,αu,b], that is when αu,b≥1. Analogously, if ωu,b≤1 then u∈Vb+. Hence, by Lemma 4.2 and Remark 4.1, the two sets are characterised as follows
Vb−={u∈X∖{0}/αu,b≥1}∪{0} and Vb+={u∈X∖{0}/ωu,b≤1}. |
Moreover, again by Remark 4.1, if b∈(0,c) then
u∈Vb−∖{0} if and only if V(u)≤b and V′(u)u>0. | (4.12) |
while
u∈Vb+ if and only if V(u)≤b and V′(u)u<0. | (4.13) |
We consider now the Nehari Manifold associated with V
N={u∈X∖{0}∣V′(u)u=0}={tuu∣u∈X∖{0}}. |
Remark 4.2. Note that, by Lemma 4.2, we have
V(tuu)=maxt≥0V(tu)∀u∈X∖{0} |
Hence, by definition, for any u∈N we get V(u)=maxt≥0V(tu). By the definition of the mountain pass level c, this implies that V(u)≥c for any u∈N. Moreover, since any nonzero critical point of V belongs to N, by Theorem 4.1, there exists u0∈N such that V(u0)=c. Therefore
infu∈NV(u)=minu∈NV(u)=c | (4.14) |
from which in particular N∩(Vb−∪Vb+)=∅ for any b∈[0,c). $
Remark 4.3. As a complementary property to Remark 4.2 note that for any b∈[0,c) the Nehari Manifold separates Vb− from Vb+ in the sense that given any path γ∈C([0,1],X) such that γ(0)∈Vb− and γ(1)∈Vb+, there exists sN∈(0,1) such that γ(sN)∈N.
Indeed, if γ(0)≠0, by (4.12) and (4.13), V′(γ(0))γ(0)>0 and V′(γ(1))γ(1)<0, so that the existence of sN follows by the continuity of the function s↦V′(γ(s))γ(s) on [0,1]. Otherwise, if γ(0)=0, since γ is not constant, there exists s∈(0,1) such that 0<‖γ(s)‖<ρ and so, by (4.2), V′(γ(s))γ(s)≥12‖γ(s)‖>0. Then, the existence of sN follows again by continuity since V′(γ(1))γ(1)<0.
Lemma 4.3. If b∈[0,c) then Vb− and Vb+ are weakly closed in X.
Proof. Let A≠B∈{Vb−,Vb+}. To prove the proposition we consider a sequence (un)⊂A such that un→u0 weakly in X and we show that u0∈A. Since (un)⊂Vb we know that V(un)≤b. Hence by Lemma 4.1 V(u0)≤b. Then, if we assume by contradiction that u0∉A we have u0∈B. Consider the path γn(s)=u0+s(un−u0), s∈[0,1]. Since γn(0)=u0∈B and γn(1)=un∈A, by Remark 4.3, for any n∈N we find sn∈(0,1) such that γn(sn)∈N and so V(γn(sn))≥c by (4.14). Since (un) is bounded in X there exists C>0 such that maxs∈[0,1]‖γn(s)‖p+1≤C for any n∈N. Then for any s∈[0,1] and n∈N, we have
ddsV(γn(s))=V′(γn(s))(un−u0)≥s‖un−u0‖2+⟨u0,un−u0⟩−Cp‖un−u0‖p+1≥⟨u0,un−u0⟩−Cp‖un−u0‖p+1. |
Integrating on [sn,1] we get
b−c≥V(un)−V(γn(sn))≥(1−sn)(⟨u0,un−u0⟩−Cp‖un−u0‖p+1). |
Since the weak convergence in X implies the convergence in Lp+1(RN) we have ⟨u0,un−u0⟩−Cp‖un−u0‖p+1→0 and the above inequality contradicts b<c. Hence the Lemma follows.
Remark 4.4. Note that, by (4.3), if b∈[0,c) and u∈Vb−, since V′(u)u≥0, then
‖u‖2≤2(p+1)p−1V(u)≤2(p+1)p−1b. |
In particular we obtain that Vb− is bounded in X. Then, by Lemma 4.3, Vb− is weakly compact in X. In particular, if (un)⊂Vb− is such that un→u0 with respect to the L2(RN) metric then u0∈Vb− and so Vb− is also closed with respect to the L2(RN) metric.
By the Gagliardo Niremberg inequality (4.5), we obtain
Lemma 4.4. We have distL2(RN)(N,0)>0 if and only if p+1≤2G,N.
Proof. Let u∈X∖{0} and, for ε>0, denote uε(x)=u(xε), x∈RN. By Lemma 4.2, tuεuε∈N for tuε=(‖uε‖2‖uε‖p+1p+1)1p−1. Since
‖∇uε‖22=εN−2‖∇u‖22,‖uε‖22=εN‖u‖22,‖uε‖p+1p+1=εN‖u‖p+1p+1 |
we obtain
‖tuεuε‖2=‖u‖2tuεεN2=‖u‖2(‖∇u‖22‖u‖p+1p+1+ε2‖u‖22‖u‖p+1p+1)1p−1εN2−2p−1 | (4.15) |
When p>2G,N−1=1+4N we have N2−2p−1>0 and so ‖tuεuε‖2→0 as ε→0. This shows that the set N accumulates the point 0 with respect to the L2 metric when p+1>2G,N.
It remains to show that if p+1≤2G,N then distL2(RN)(N,0)>0. By contradiction assume that there exists (un)⊂N such that ‖un‖2→0 as n→+∞. By (4.2) we have that ‖u‖≥ρ for any u∈N and so, in particular, ‖un‖≥ρ for any n∈N. Then, since ‖un‖2→0, we have ‖∇un‖2≥ρ+o(1) as n→+∞. Hence, since by (4.5) we have
0=V′(un)un≥‖∇un‖22+‖un‖22−Kp+1G‖un‖(p+1)θN,p2‖∇un‖(p+1)(1−θN,p)2=‖∇un‖22(1+‖un‖22‖∇un‖22−Kp+1G‖un‖(p+1)θN,p2‖∇un‖2−(p+1)(1−θN,p)2) |
and since p≤1+4N implies 2≥(p+1)(1−θN,p) (see (4.8)) we obtain
0=V′(un)un≥‖∇un‖22(1+o(1))≥ρ2+o(1), as n→+∞, |
a contradiction.
Then we can give our first separation result.
Proposition 4.1. For any b∈[0,c) we have
distL2(RN)(Vb−,Vb+)>0 if and only if p+1≤2G,N. |
Proof. Let p+1>2G,N. By Remark 4.1, if b∈[0,c) and uε(x)=u(xε), as in the proof of Lemma 4.4, we obtain αuε,buε∈Vb−, ωuε,buε∈Vb+ and 0≤αuε,b<tuε<ωuε,b≤p+12tuε. In particular, ‖αuε,buε‖2≤‖ωuε,buε‖2≤p+12‖tuεuε‖2→0 as ε→0+, as proved in (4.15). So, if p+1>2G,N, then distL2(RN)(Vb−,Vb+)=0 for any b∈[0,c).
Let now p+1≤2G,N and b∈[0,c) and let us show that distL2(RN)(Vb−,Vb+)>0. Arguing by contradiction, assume there exist (un)⊂Vb− and (vn)⊂Vb+ such that ‖vn−un‖2→0. Since Vb− is weakly compact (see Remark 4.4) there exists ˉu∈Vb− such that, up to subsequences, un→ˉu weakly in X.
We claim that along the same subsequence also (vn) is bounded in X. If the claim is proved then the Proposition follows. Indeed since ‖vn−un‖2→0 we have vn→ˉu weakly in L2(RN) and the boundedness of (vn) in X would imply, by uniqueness, vn→ˉu weakly in X. Then, by Lemma 4.3, ˉu∈Vb+, contradicting that ˉu∈Vb− and concluding the proof.
To prove that (vn) is bounded in X we argue in an indirect way assuming that, along a subsequence, still denoted vn, ‖vn‖→+∞. Since V(vn)=12‖vn‖2−1p+1‖vn‖p+1p+1≤b we derive ‖vn‖p+1→+∞ too. Since (un) is bounded in X, and hence in Lp+1, we obtain that there exists ˉn∈N such that if n≥ˉn then
‖vn−un‖p+1≥‖vn‖p+1−‖un‖p+1≥12‖vn‖p+1. |
Analogously, since ‖vn‖→+∞ and (un) is bounded in X we can assume that ˉn is so large that if n≥ˉn then
‖vn−un‖≤‖vn‖+‖un‖≤2‖vn‖. |
Using (4.5), for n≥ˉn, it follows that
(12‖vn‖p+1)p+1≤‖vn−un‖p+1p+1≤Kp+1G‖vn−un‖θN,p(p+1)2‖∇vn−∇un‖(1−θN,p)(p+1)2≤Kp+1G‖vn−un‖θN,p(p+1)2(2‖∇vn‖2)(1−θN,p)(p+1). |
Recalling that ‖vn−un‖2→0 and since (1−θN,p)(p+1)≤2 when p+1≤2G,N (see (4.7) and (4.8)), the above inequality gives that for n large enought
‖vn‖p+1p+1≤(2KG)p+1‖vn−un‖θN,p(p+1)2(2‖∇vn‖2)(1−θN,p)(p+1)≤(2KG)p+1‖vn−un‖θN,p(p+1)24‖vn‖2≤12‖vn‖2. |
Hence, for such values of n, V′(vn)vn=‖vn‖2−‖vn‖p+1p+1≥12‖vn‖2>0, in contradiction by (4.13) with the assumption vn∈Vb+. This proves that (vn) is bounded in X and, as explained above, concludes the proof of the Proposition. To study the cases p+1>2G,N it is useful to observe that
Lemma 4.5. If p+1∈(2G,N,2∗N) and (un)⊂N is such that ‖un‖2→0 then ‖∇un‖2→+∞.
Proof. By Lemma 4.4.4 the Nehari manifold N accumulates the origin in X when p+1∈(2G,N,2∗N). By (4.10) we have that if u∈N then
V(u)=V(tuu)=12(12−1p+1)(‖u‖‖u‖p+1)2(p+1)p−1≥12(12−1p+1)ν[(‖u‖2‖∇u‖2)N+(‖∇u‖2‖u‖2)2(p+1)p−1−N]. |
In particular if (un)∈N is such that ‖un‖2→0, since by (4.2) ‖un‖≥ρ, and 2(p+1)p−1>N when p+1<2∗N, we obtain
V(un)≥12(12−1p+1)ν(‖∇un‖2‖un‖2)2(p+1)p−1−N→+∞ as n→+∞. |
Since un∈N we have ‖∇un‖22=‖un‖p+1p+1 and so
V(un)=(12−1p+1)‖∇un‖22→+∞ |
and the Lemma follows.
Note that if u∈N, since ‖∇u‖22=‖u‖p+1p+1, (4.5) implies
‖∇u‖22=‖u‖p+1p+1≤Kp+1G‖u‖(p+1)θN,p2‖∇u‖(p+1)(1−θN,p)2 |
and so
‖u‖(p+1)θN,p(p+1)(1−θN,p)−22≥(1KG)(p+1)(p+1)(1−θN,p)−21‖∇u‖2 |
Hence, setting
σN,p=(p+1)(1−θN,p)−2(p+1)θN,p, |
we get
‖u‖2‖∇u‖2≥(1KG)1θN,p‖∇u‖1−σN,p2 | (4.16) |
Note that
σN,p<1 if p+1∈(2G,N,2∗N+1) and σN,p=1 for p+1=2∗N+1. | (4.17) |
Lemma 4.4 states that when p+1>2G,N there are sequences (un) in N such that ‖un‖2→0 and, by Lemma 4.5, ‖∇un‖2→+∞. By (4.16), we get that in such cases
limn→+∞‖un‖2‖∇un‖2=+∞ if p+1<2∗N+1 andlim infn→+∞‖un‖2‖∇un‖2≥(1KG)1θN,p if p+1=2∗N+1. |
We have furthermore
Lemma 4.6. Let p+1∈(2G,N,2∗N+1]. If u0∈N and u∈X are such that
V′(u)u≥0 and ‖u−u0‖2≤(14KG)1θN,p1‖u0‖σN,p |
then
V(u)≥(12)p+1V(u0). |
Proof. Consider first the case ‖u‖≥‖u0‖. Since u0∈N and V′(u)u≥0 we have ‖u0‖2=‖u0‖p+1p+1 and ‖u‖2≥‖u‖p+1p+1. Hence V(u0)=(12−1p+1)‖u0‖2 and V(u)≥(12−1p+1)‖u‖2, from which
V(u)−V(u0)≥(12−1p+1)(‖u‖2−‖u0‖2)≥0. |
Let now assume ‖u‖<‖u0‖. By (4.5) we have
‖u−u0‖p+1≤KG‖u−u0‖θN,p2‖∇u−∇u0‖1−θN,p2. | (4.18) |
Since ‖u‖<‖u0‖, we have ‖∇u−∇u0‖2≤‖u−u0‖≤2‖u0‖. Thus, since ‖u−u0‖2≤(14KG)1θN,p1‖u0‖σN,p, by (4.18) we obtain
‖u−u0‖p+1≤KG‖u−u0‖θN,p2‖∇u−∇u0‖1−θN,p2<KG‖u−u0‖θN,p221−θN,p‖u0‖1−θN,p<KG((14KG)1θN,p1‖u0‖σN,p)θN,p21−θN,p‖u0‖1−θN,p<12‖u0‖2p+11‖u0‖θN,pσN,p‖u0‖(p+1)(1−θN,p)−2p+1==12‖u0‖2p+11‖u0‖θN,pσN,p‖u0‖θN,pσN,p=12‖u0‖2p+1. |
Hence, recalling that ‖u0‖p+1=‖u0‖2/(p+1) being u0∈N, we get
‖u‖p+1p+1≥(‖u0‖p+1−‖u−u0‖p+1)p+1≥(1−1‖u0‖2/(p+1)‖u−u0‖p+1)p+1‖u0‖p+1p+1≥(12)p+1‖u0‖p+1p+1. |
But V′(u)u≥0 implies ‖u‖2≥‖u‖p+1p+1 and moreover V(u0)=(12−1p+1)‖u0‖p+1p+1 since u0∈N. Then
V(u)=12‖u‖2−1p+1‖u‖p+1p+1≥(12−1p+1)‖u‖p+1p+1≥(12)p+1(12−1p+1)‖u0‖p+1p+1=(12)p+1V(u0), |
and the Lemma follows.
We have seen in Corollary 2.1 that Vb− and Vb+ have zero L2(RN) distance when p+1>2G,N. Lemma 4.6 allows us to show, as a final result, that the two sets are instead well separated when 2G,N<p+1≤2∗N+1 with respect to the Jacobi distance
dJ(Vb−,Vb+)=infΓb∫10‖˙γ(t)‖2√2(V(γ(t))−b)dt | (4.19) |
where
Γb={γ∈AC([0,1],X)∣γ(0)∈Vb−,γ(1)∈Vb+, γ((0,1))⊂X∖Vb}. |
Proposition 4.2. If b∈[0,c) then
dJ(Vb−,Vb+)>0 if and only if p+1≤2∗N+1. |
Proof. Consider first the case p+1∈(2∗N+1,2∗N), and so N>1. As in Proposition 4.1 we show that dJ(Vb−,Vb+)=0 by using dilation. For u∈X∖{0} and ε>0 we set uε(x)=u(xε). By Lemma 4.2, tuεuε∈N for tuε=(‖uε‖2‖uε‖p+1p+1)1p−1 and by Remark 4.1, setting
γε(t)=(αuε,b+t(ωuε,b−αuε,b))uε,t∈[0,1], |
we have γε∈Γb with, by (4.10) in Lemma 4.2,
V(γε(t))≤V(tuεuε)=(12−1p+1)(‖uε‖‖uε‖p+1)2(p+1)p−1 for any t∈[0,1]. | (4.20) |
By (4.11) we have
0≤ωuε,b−αuε,b≤p+12tuε |
and so
dJ(Vb−,Vb+)≤∫10‖˙γε(t)‖2√2(V(γε(t))−b)dt≤√(1−2p+1)(‖uε‖‖uε‖p+1)2(p+1)p−1∫10‖˙γε(t)‖2dt=√1−2p+1(‖uε‖‖uε‖p+1)p+1p−1(ωuε,b−αuε,b)‖uε‖2≤√1−2p+1(‖uε‖‖uε‖p+1)p+1p−1p+12tuε‖uε‖2. | (4.21) |
Since ‖∇uε‖2=εN−22‖∇u‖2, ‖uε‖2=εN2‖u‖2, ‖uε‖p+1=εNp+1‖u‖p+1 for ε∈(0,1) we obtain
(‖uε‖‖uε‖p+1)(p+1)p−1≤(‖u‖‖u‖p+1)p+1p−1ε12(p−1)((N−2)(p+1)−2N). |
As in (4.15) we have furthermore that
tuε‖uε‖2=‖u‖2(‖∇u‖22‖u‖p+1p+1+ε2‖u‖22‖u‖p+1p+1)1p−1ε12(p−1)(N(p−1)−4) |
Hence, by (4.21) we obtain that for all u∈X∖{0} there exists a constant C>0 such that for ε∈(0,1) we get
dJ(Vb−,Vb+)≤Cε12(p−1)((N−2)(p+1)−2N+N(p−1)−4). | (4.22) |
Since N>1 and p+1>2∗N+1 we have
(N−2)(p+1)−2N+N(p−1)−4=2(N−1)(p+1−2∗N+1)>0, |
and (4.22) gives
dJ(Vb−,Vb+)=0. |
We now consider the case p+1≤2∗N+1 and prove that dJ(Vb−,Vb+)>0 for any b∈[0,c). We first observe that, for u0 as given in Theorem 4.1, we have
dJ(Vb−,Vb+)≤p+12‖u0‖2√2(c−b) for any b∈[0,c). | (4.23) |
Indeed, for b∈[0,c) let αu0,b and ωu0,b be given by Remark 4.1 and define
γ(t)=[αu0,b+t(ωu0,b−αu0,b)]u0,t∈[0,1]. |
Then γ∈Γb and, since tu0=1, (4.11) gives ‖˙γ(t)‖2=(ωu0,b−αu0,b)‖u0‖2≤p+12‖u0‖2. Hence, since V(γ(t))≤V(u0)=c for any t∈[0,1], (4.23) follows.
Let now (γn)⊂Γb be such that
∫10‖˙γn(t)‖2√2(V(γn(t))−b)dt→dJ(Vb−,Vb+) as n→+∞. |
By Remark 4.3 and by continuity, for n∈N, there is sn∈(0,1) such that γn(sn)∈N and V′(γn(t))γn(t)>0 for any t∈[0,sn).
We analyse first the case in which γn(sn) is not bounded in X assuming that along a subsequence, still denoted (γn(sn)), we have
‖γn(sn)‖→+∞ | (4.24) |
By (4.24) we have in particular
V(γn(sn))=(12−1p+1)‖γn(sn)‖2→+∞ as n→+∞. | (4.25) |
By (4.16) we know that ‖γn(sn)‖2≥(1KG)1θN,p1‖∇γn(sn)‖σN,p2 and by Lemma 4.6, if t∈[0,sn) is such that ‖γn(t)−γn(sn)‖2≤(14KG)1θN,p1‖γn(sn)‖σN,p then
V(γn(t))≥(12)p+1V(γn(sn)). |
By (4.25) there exists ˉn∈N such that (12)p+1V(γn(sn))>b≥V(γn(0)) for any n≥ˉn and so ‖γn(0)−γn(sn)‖>(14KG)1θN,p1‖γn(sn)‖σN,p. Then, for any n≥ˉn there is ˉsn∈(0,sn) such that
(i) ‖γn(t)−γn(sn)‖2≤(14KG)1θN,p1‖γn(sn)‖σN,p for any t∈(ˉsn,sn),
(ii) ‖γn(ˉsn)−γn(sn)‖2=(14KG)1θN,p1‖γn(sn)‖σN,p,
(iii) V(γn(t))≥(12)p+1V(γn(sn)) for any t∈(ˉsn,sn).
Since V(γn(sn))→+∞, by (iii) we can assume that for n≥ˉn we have V(γn(t))−b≥(12)p+2V(γn(sn)) for any t∈(ˉsn,sn). Hence, using (i) and (ii) we obtain
∫snˉsn‖˙γn(t)‖2√2(V(γn(t))−b)dt≥√(12)p+1V(γn(sn))∫snˉsn‖˙γn(t)‖2≥√(12)p+1V(γn(sn))‖γn(sn)−γn(ˉsn)‖2=√(12)p+1(12−1p+1)‖γn(sn)‖(14KG)1θN,p1‖γn(sn)‖σN,p. |
If p+1<2∗N+1 by (4.17) we have σN,p<1 and so the above inequality would imply
dJ(Vb−,Vb+)=limn→+∞∫10‖˙γn(t)‖2√2(V(γn(t))−b)dt=+∞, |
in contradiction with (4.23). So the case (4.24) cannot occur when p+1<2∗N+1. If p+1=2∗N+1 by (4.17) we have σN,p=1 and the above inequality gives
∫snˉsn‖˙γn(t)‖2√2(V(γn(t))−b)dt≥√(12)p+1(12−1p+1)(14KG)1θN,p. |
Since dJ(Vb−,Vb+)=limn→+∞∫10‖˙γn(t)‖2√2(V(γn(t))−b)dt, this proves that the Jacobi distance between Vb− and Vb+ is positive in this case.
To conclude the proof of the Proposition we have finally to consider the case in which (γn(sn)) is bounded in X. We claim that in this case for any β∈[0,c) there exists dβ>0 such that
distL2(RN)(γn(sn),Vβ−)≥dβ for any n∈N. | (4.26) |
Indeed, by Remark 4.4, Vβ− is weakly compact in X. Then for any n∈N there exists vn∈Vβ− such that distL2(RN)(γn(sn),Vβ−)≥‖γn(sn)−vn‖L2(RN)>0 and (4.26) will follows if we show that lim infn→+∞‖γn(sn)−vn‖L2(RN)>0. To this aim assume by contradiction that lim infn→+∞‖γn(sn)−vn‖L2(RN)=0. By the weak compactness of Vβ− and the boundedness of (γn(tn)), there exist v0∈Vβ− and u0∈X such that, along common subsequences still denoted (γn(sn)) and (vn), we have ‖γn(sn)−vn‖2→0, vn→v0 and γn(sn)→u0 weakly in X and strongly in Lp+1(RN). Since ‖γn(sn)−vn‖2→0, by weak semicontinuity of the norm we recover ‖u0−v0‖L2(RN)=0. Hence u0=v0∈Vb−. Since V(γn(sn))≥c>b and V(u0)≤b we then have that γn(sn) does not converge strongly in X to u0 and so ‖u0‖<lim infn→+∞‖γn(sn)‖. In particular, since by the strong convergence in Lp+1(RN) we have ‖γn(sn)‖p+1p+1→‖u0‖p+1p+1, we obtain
V′(u0)u0=‖u0‖2−‖u0‖p+1p+1<lim infn→+∞‖γn(sn)‖2−limn→+∞‖γn(sn)‖p+1p+1=lim infn→+∞(‖γn(sn)‖2−‖γn(sn)‖p+1p+1)=lim infn→+∞V′(γn(sn))γn(sn)=0 |
which says that V′(u0)u0<0 in contradiction with u0∈Vb−, by (4.13). Then (4.26) follows.
In particular (4.26) gives that for β=12(b+c) there exists dβ>0 such that
distL2(RN)(γn(sn),V(b+c)/2−)≥dβ>0 for any n∈N. | (4.27) |
Since V(γn(0))=b and V(γn(sn))≥c, by (4.27) and by continuity there exists ˉsn∈(0,sn) such that
(iv) V(γn(t))≥12(b+c) for any t∈(ˉsn,sn), (v) ‖γn(ˉsn)−γn(sn)‖2≥dβ.
Using (iv) and (v), we derive
dJ(Vβ−,Vβ+)+o(1)≥∫snˉsn‖˙γn(t)‖2√2(V(γn(t))−b)dt≥≥√c−b∫snˉsn‖˙γn(t)‖2≥≥√c−b‖γn(sn)−γn(ˉsn)‖2≥√c−bdβ>0 |
and the proposition is proved.
The authors declare they do not have a conflict of interest.
[1] | S. Alama, L. Bronsard, C. Gui, Stationary layered solutions in R.2 for an Allen-Cahn system with multiple well potential, Calc. Var., 5 (1997), 359-390. |
[2] | F. Alessio, Brake orbits type solutions for a system of Allen-Cahn type equations, Indiana U. Math. J., 62 (2013), 1535-1564. |
[3] | F. Alessio, M. L. Bertotti, P. Montecchiari, Multibump solutions to possibly degenerate equilibria for almost periodic Lagrangian systems, Z. Angew. Math. Phys., 50 (1999), 860-891. |
[4] | F. Alessio, L. Jeanjean, P. Montecchiari, Stationary layered solutions in R.2 for a class of non autonomous Allen-Cahn equations, Calc. Var., 11 (2000), 177-202. |
[5] | F. Alessio, L. Jeanjean, P. Montecchiari, Existence of infinitely many stationary layered solutions in R.2 for a class of periodic Allen Cahn Equations, Commun. Part. Diff. Eq., 27 (2002), 1537-1574. |
[6] | F. Alessio, P. Montecchiari, Entire solutions in R.2 for a class of Allen-Cahn equations, ESAIM: COCV, 11 (2005), 633-672. |
[7] | F. Alessio, P. Montecchiari, Multiplicity of entire solutions for a class of almost periodic AllenCahn type equations, Adv. Nonlinear Stud., 5 (2005), 515-549. |
[8] | F. Alessio, P. Montecchiari, Brake orbits type solutions to some class of semilinear elliptic equations, Calc. Var., 30 (2007), 51-83. |
[9] | F. Alessio, P. Montecchiari, Layered solutions with multiple asymptotes for non autonomous Allen-Cahn equations in R.3, Calc. Var., 46 (2013), 591-622. |
[10] | F. Alessio, P. Montecchiari, Multiplicity of layered solutions for Allen-Cahn systems with symmetric double well potential, J. Differ. Equations, 257 (2014), 4572-4599. |
[11] | F. Alessio, P. Montecchiari, An energy constrained method for the existence of layered type solutions of NLS equations, Ann. Inst. H. Poincaré Anal. NonLinéaire, 31 (2014), 725-749. |
[12] | F. Alessio, P. Montecchiari, Brake orbit solutions for semilinear elliptic systems with asymmetric double well potential, J. Fix. Point Theory A., 19 (2017), 691-717. |
[13] | F. Alessio, P. Montecchiari, A. Zuniga, Prescribed energy connecting orbits for gradient systems, Discrete Cont. Dyn. A, 39 (2019), 4895-4928. |
[14] | V. Bargert, On minimal laminations on the torus, Ann. Inst. H. Poincaré Anal. NonLinéaire, 6 (1989), 95-138. |
[15] | V. Benci, Closed geodesics for the Jacobi metric and periodic solutions of prescribed energy of natural Hamiltonian systems, Ann. Inst. H. Poincaré Anal. NonLinéaire, 1 (1984), 401-412. |
[16] | H. Berestycki, F. Hamel, R. Monneau, One-dimensional symmetry for some bounded entire solutions of some elliptic equations, Duke Math. J., 103 (2000), 375-396. |
[17] | H. Berestycki, T. Cazenave, Instabilité des états stationnaires dans les équations de Schrödinger et de Klein-Gordon non linéaires, C. R. Acad. Sci. Paris, 293 (1981), 489-492. |
[18] | H. Berestycki, P. L. Lions, Nonlinear scalar field equations, I. Existence of a ground state, Arch. Ration. Mech. Anal., 82 (1983), 313-345. |
[19] | H. Berestycki, T. Gallouët, O. Kavian, Equations de Champs scalaires euclidiens non lineaires dans le plan, C. R. Acad. Sci Paris, 297 (1983), 307-310. |
[20] | T. Cazenave, P. L. Lions, Orbital stability of standing waves for some nonlinear Schrödinger equations, Commun. Math. Phys., 85 (1982), 549-561. |
[21] | A. Cesaroni, M. Cirant, Brake orbits and heteroclinic connections for first order mean field games, Preprint, 2019. |
[22] | E. N. Dancer, New solutions of equations on R.n, Ann. Sc. Norm. Super. Pisa Cl. Sci., 30 (2001), 535-563. |
[23] | G. Fusco, G. F. Gronchi, M. Novaga, Existence of periodic orbits near heteroclinic connections, Minimax Theory and its Applications, 4 (2019), 113-149. |
[24] | E. De Giorgi, Convergence problems for functionals and operators, In: Proc. Int. Meeting on Recent Methods in Nonlinear Analysis, Rome, 1978. |
[25] | A. Farina, Finite-energy solutions, quantization effects and Liouville-type results for a variant of the Ginzburg-Landau systems in RK, Differ. Integral Equ., 11 (1998), 875-893. |
[26] | A. Farina, Some remarks on a conjecture of De Giorgi, Calc. Var., 8 (1999), 233-245. |
[27] | A. Farina, Symmetry for solutions of semilinear elliptic equation in R.N and related conjectures, Papers in memory of E. De Giorgi, Ric. Mat., 48 (1999), 129-154. |
[28] | A. Farina, Simmetria delle soluzioni di equazioni ellittiche semilineari in R.N, Atti Accad. Naz. Lincei Cl. Sci. Fis. Mat. Natur. Rend. Lincei Mat. Appl., 10 (1999), 255-265. |
[29] | A. Farina, Monotonicity and one-dimensional symmetry for the solutions of Δu+f(u)=0 in R.N with possibly discontinuous nonlinearity, Adv. Math. Sci. Appl., 11 (2001), 811-834. |
[30] | A. Farina, Rigidity and one-dimensional symmetry for semilinear elliptic equations in the whole of R.N and in half spaces, Adv. Math. Sci. Appl., 13 (2003), 65-82. |
[31] | A. Farina, B. Sciunzi, E. Valdinoci, Bernstein and De Giorgi type problems: New results via a geometric approach, Ann. Sc. Norm. Super. Pisa Cl. Sci., 7 (2008), 741-791. |
[32] | A. Farina, E. Valdinoci, Geometry of quasiminimal phase transitions, Calc. Var., 33 (2008), 1-35. |
[33] | A. Farina, E. Valdinoci, The state of the art of a conjecture of De Giorgi and related problems, In: Recent progress in Reaction-Diffusion systems and viscosity solutions, Hackensack: World Scientific Publishers, 2009, 74-96. |
[34] | A. Farina, E. Valdinoci, Rigidity results for elliptic PDEs with uniform limits: An abstract framework with applications, Indiana U. Math. J., 60 (2011), 121-141. |
[35] | A. Farina, E. Valdinoci, Some results on minimizers and stable solutions of a variational problem, Ergod. Theor. Dyn. Syst., 32 (2012), 1302-1312. |
[36] | A. Farina, A. Malchiodi, M. Rizzi, Symmetry properties of some solutions to some semilinear elliptic equations, Ann. Sc. Norm. Super. Pisa Cl. Sci., XVI (2016), 1209-1234. |
[37] | A. Farina, N. Soave, Monotonicity and 1-dimensional symmetry for solutions of an elliptic system arising in Bose-Einstein condensation, Arch. Ration. Mech. Anal., 213 (2014), 287-326. |
[38] | A. Farina, B. Sciunzi, N. Soave, Monotonicity and rigidity of solutions to some elliptic systems with uniform limits, Commun. Contemp. Math., 22 (2020), 1950044. |
[39] | E. Gagliardo, Proprietà di alcune classi di funzioni di più variabili, Ric. Mat., 7 (1958), 102-137. |
[40] | N. Ghoussoub, C. Gui, On a conjecture of De Giorgi and some related problems, Math. Ann., 311 (1998), 481-491. |
[41] | C. Gui, Hamiltonian identities for elliptic partial differential equations, J. Funct. Anal., 254 (2008), 904-933. |
[42] | C. Gui, A. Malchiodi, H. Xu, Axial symmetry of some steady state solutions to nonlinear Schroedinger Equations, P. Am. Math. Soc., 139 (2011), 1023-1032. |
[43] | L. Jeanjean, K. Tanaka, A remark on least energy solutions in R.N, P. Am. Math. Soc., 131 (2003), 2399-2408. |
[44] | M. K. Kwong, Uniqueness of positive solutions of Δu−u+u.p in R.n, Arch. Ration. Mech. Anal., 105 (1989), 243-266. |
[45] | A. Malchiodi, Some new entire solutions of semilinear elliptic equations on R.n, Adv. Math., 221 (2009), 1843-1909. |
[46] | H. Matano, P. H. Rabinowitz, On the necessity of gaps, J. Eur. Math. Soc., 8 (2006), 355-373. |
[47] | A. Monteil, F. Santambrogio, Metric methods for heteroclinic connections in infinite dimensional spaces, Indiana U. Math. J., 69 (2020), 1445-1503. |
[48] | J. Moser, Minimal solutions of variational problem on a torus, Ann. Inst. H. Poincaré Anal. NonLinéaire, 3 (1986), 229-272. |
[49] | L. Nirenberg, On Elliptic partial differential equations, Ann. Sc. Norm. Super. Pisa Cl. Sci., 13 (1959), 116-162. |
[50] | P. H. Rabinowitz, Heteroclinic for reversible Hamiltonian system, Ergod. Theor. Dyn. Syst., 14 (1994), 817-829. |
[51] | P. H. Rabinowitz, Solutions of heteroclinic type for some classes of semilinear elliptic partial differential equations, J. Math. Sci. Univ. Tokio, 1 (1994), 525-550. |
[52] | P. H. Rabinowitz, E. Stredulinsky, Mixed states for an Allen-Cahn type equation, Commun. Pure Appl. Math., 56 (2003), 1078-1134. |
[53] | P. H. Rabinowitz, E. Stredulinsky, Mixed states for an Allen-Cahn type equation, II, Calc. Var., 21 (2004), 157-207. |
[54] | P. H. Rabinowitz, E. Stredulinsky, Extensions of Moser-Bangert theory: Locally minimal solutions, Boston: Birkhauser, 2011. |
[55] | M. Schatzman, Asymmetric heteroclinic double layers, ESAIM: COCV, 8 (2002), 965-1005. |
[56] | J. Serrin, M. Tang, Uniqueness of ground states for quasilinear elliptic equations, Indiana U. Math. J., 49 (2000), 897-923. |
[57] | P. Sternberg, A. Zuniga, On the heteroclinic connection problem for multi-well potentials with several global minima, J. Differ. Equations, 261 (2016), 3987-4007. |
[58] | W. A. Strauss, Existence of solitary waves in higher dimensions, Commun. Math. Phys., 55 (1977), 149-162. |
[59] | M. I. Weinstein, Nonlinear Schrödinger equations and sharp interpolation estimates, Commun. Math. Phys., 87 (1983), 567-576. |
1. | Serena Dipierro, Luca Lombardini, Partial differential equations from theory to applications: Dedicated to Alberto Farina, on the occasion of his 50th birthday, 2023, 5, 2640-3501, 1, 10.3934/mine.2023050 | |
2. | Claudianor O. Alves, Renan J. S. Isneri, Piero Montecchiari, Existence of heteroclinic and saddle-type solutions for a class of quasilinear problems in whole ℝ2, 2024, 26, 0219-1997, 10.1142/S0219199722500614 |