Dual variational methods for a nonlinear Helmholtz system

Abstract: This paper considers a pair of coupled nonlinear Helmholtz equations �egin{align*} -Delta u - mu u = a(x) left( |u|^frac{p}{2} + b(x) |v|^frac{p}{2} ight)|u|^{frac{p}{2} - 2}u, end{align*} �egin{align*} -Delta v - u v = a(x) left( |v|^frac{p}{2} + b(x) |u|^frac{p}{2} ight)|v|^{frac{p}{2} - 2}v end{align*} on

m a t h b b R^{N}

where

f r a c 2 (N + 1) N - 1 < p < 2^{a} s t

. The existence of nontrivial strong solutions in

W^{2, p} (m a t h b b R^{N})

is established using dual variational methods. The focus lies on necessary and sufficient conditions on the parameters deciding whether or not both components of such solutions are nontrivial.

Review text

The authors consider existence and non-existence of nontrivial solutions \((u,v) \in W^{2,p} (\mathbb{R}^N) \times W^{2,p} (\mathbb{R}^N)\) of the coupled nonlinear equations \[ \begin{aligned} -\Delta u - \mu u, &= a(x)\left( |u|^{\frac{p}{2}} + b(x) |v|^{\frac{p}{2}} \right) |u|^{\frac{p}{2}-2}u\\ -\Delta v - \nu v &= a(x)\left( |v|^{\frac{p}{2}} + b(x) |u|^{\frac{p}{2}} \right) |v|^{\frac{p}{2}-2}v ,\end{aligned}\tag{1} \] on \(\mathbb{R}^N\), \(N\geq 2\), where \(\frac{2(N+1)}{N-1} < p < 2^\ast\) (the critical Sobolev exponent), \(a,b \in L^\infty (\mathbb{R}^N)\) are nonnegative \(\mathbb{Z}^N\)-periodic coefficients, assumed to satisfy the bounds \(0 \leq b(x) \leq p-1\) and \(a(x)\geq a_0 \geq 0\) almost everywhere, and, most importantly, \(\mu,\nu>0\), i.e., we are in the much less well understood case of Helmholtz equations than the Schrödinger case \(\mu,\nu<0\). As the title suggests, the approach taken involves introducing and studying a dual formulation of the problem (1), which is variational. This follows ideas of \textit{G. Evequoz} and \textit{T. Weth} [Adv. Math. 280, 690--728 (2015; Zbl 1317.35030)] and \textit{G. Evéquoz} [Analysis, München 37, No. 2, 55--68 (2017; Zbl 1376.35004)] for (single) Helmholtz equations. Here, the dual problem is shown to have a nontrivial mountain pass solution under the assumptions stated above, corresponding to a nontrivial strong solution of (1) (Theorem 1). Such solutions are fully nontrivial (i.e., neither \(u\) nor \(v\) is identically zero) under either of the following sets of assumptions (Theorem 2): {\parindent=0.7cm\begin{itemize}\item[(a)] \(2 < p < 4\) and \(\text{ess inf} \, b(x) > 0\), or \item[(b)] \(p \geq 4\), \(\text{ess inf} \, b(x) > \frac{\text{ess sup}\,a(x)}{\text{ess inf}\,a(x)} \cdot 2^{\frac{p-2}{2}}-1\) and \(\mu/\nu\) is sufficiently close to \(1\). \end{itemize}} However, if \(p \geq 4\) and \(0 \leq \text{ess sup} \, b(x) < 2^{\frac{p-2}{2}}\), then every such solution is semitrivial, i.e., either \(u\) or \(v\) is identically zero (Theorem 3).

Recommendations

Cites work

Cited in

(11)

This page was built for publication: Dual variational methods for a nonlinear Helmholtz system

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1635645)