scientific article; zbMATH DE number 4017380

From MaRDI portal
Publication:3029359





[For the entire collection see Zbl 0624.00010.] Let us consider the periodic boundary value problem for the second order differential equation \[ (1)\quad -u=g(u)+h(t),\quad t\in [0,2\pi], \] \[ (2)\quad u(0)=u(2\pi),\quad u'(0)=u'(2\pi), \] where \(g: {\mathbb{R}}\to {\mathbb{R}}\) is continuous and h: [0,2\(\pi\) ]\(\to {\mathbb{R}}\) is Lebesgue integrable. Solutions to (1)-(2) are intended in the Carathéodory sense. It is well known that, when g has at most linear growth at infinity, the solvability of (1)-(2) is determined by the interaction of the ratio g(v)/v with the spectrum \(\sum =\{m^ 2:m=0,1,2,...\}\) of the linear operator \(-d^ 2/dt^ 2\), with periodic boundary conditions on [0,2\(\pi\) ]. Therefore, let us set \(\mu_{\pm}=\liminf_{v\to \pm \infty}g(v)/v\) and \(\nu_{\pm}=\limsup_{v\to \pm \infty}g(v)/v\) and assume that, for some \(m=1,2,...\), (3) \(m^ 2\leq \mu_-\leq \nu_-\leq (m+1)^ 2\) and \(m^ 2\leq \mu_+\leq \nu_+\leq (m+1)^ 2\). Looking at the linear situation (i.e. \(g(v)=\lambda v)\), it follows that assumption (3) is not sufficient, in general, for the solvability of (1)-(2). Hence, some further restriction must be imposed either on g or on h. Following the usual terminology, we classify the conditions on g as nonresonant or resonant, according to whether they assure the solvability of (1)-(2) for every h, or not. In the linear case, they respectively correspond to \(\lambda\not\in \Sigma\), or \(\lambda\in \Sigma\). Here, we propose a nonresonant condition, where it is permitted that both \(\mu_+=\mu_- =m^ 2\) and \(\nu_+=\nu_-=(m+1)^ 2\).











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3029359)