Forced vibrations of wave equations with non-monotone nonlinearities (Q2507954)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Forced vibrations of wave equations with non-monotone nonlinearities
scientific article

    Statements

    Forced vibrations of wave equations with non-monotone nonlinearities (English)
    0 references
    0 references
    0 references
    5 October 2006
    0 references
    This paper deals with the periodically forced nonlinear wave equation \[ u_{tt}-u_{xx}=\varepsilon f(t,x,u;\varepsilon) \] with the Dirichlet boundary condition \(u(t,0)=u(t,\pi)=0\), where \(f\) is \(T\)-periodic with respect to \(t\), in the completely resonant case in which \(T=2\pi\). The aim is to prove existence and regularity of a periodic solution, when \(\varepsilon\) is small. While previous work proving such existence, which is surveyed in the introduction to this paper, has relied heavily on a monotonicity assumption \(\partial_u f>\beta_0\), the authors of this paper develop new ideas using which they are able to prove existence and regularity for a class of perturbations \(f\) which are not monotone with respect to \(u\). The difficulty of the problem arises from the fact that the kernel \(N\) of the linear problem obtained for \(\varepsilon=0\), hence the bifurcation equation obtained by the Lyapunov-Schmidt method, are infinite dimensional. A minimization argument is used to solve the bifurcation equation, which has a variational structure. The main effort is to prove that a minimum of the relevant functional on a sufficiently large closed ball in fact lies in the interior of the ball, and is hence a critical point, and for this the authors develop their method of obtaining a-priori estimates using the variational equation obtained at the minimum point. As an example of the paper's results, Theorem 1 says that if \(f\) is of the form \(f(t,x,u)=\beta u^{2k}+h(t,x)\), where \(h\in N^\bot\) (the orthogonal complement of the kernel of the non-perturbed linear problem) and also either \(h(t,x)>0\) for {\textit{a.e.}} \(x,t\) or \(h(t,x)<0\) for {\textit{a.e.}} \(x,t\), then a \(T\)-periodic weak solution exists for \(\varepsilon\) sufficiently small, which is a classical solution assuming some regularity of \(h\). The proofs are technically involved, but the paper is clearly written, with the needed preliminaries, as well as the main ideas of the proofs, explained in detail.
    0 references
    0 references
    wave equation
    0 references
    periodic solutions
    0 references
    variational methods
    0 references
    a priori estimates
    0 references
    Lyapunov-Schmidt reduction
    0 references
    Dirichlet boundary condition
    0 references
    completely resonant case
    0 references
    0 references
    0 references
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references