New revival phenomena for bidirectional dispersive hyperbolic equations (Q7007045)

The authors study the dispersive revival and fractalization phenomena (also called \textit{Talbot effect}) for bidirectional dispersive equations on a bounded interval, subject to periodic boundary conditions and discontinuous initial profiles.\N\NIn Section 2, they study the periodic initial-boundary value problem of the linear beam equation, in the interval \([0,2\pi]\), with step functions\N\[\Nf(x)=g(x)=\left\{\begin{array}{cl} -1, & 0\leqslant x < \pi,\\\N1, & \pi \leqslant x < 2\pi, \end{array}\right.\N\]\Nas initial data, analyzing the manifestation of the revival phenomenon for the corresponding solution at \textit{rational times}, that is, times \(t>0\) for which \(\frac{t}{\pi}\in \mathbb{Q}\). The main result of this section is Theorem 2.6, which leads to the conclusion that the evolution of the periodic step function \(f\) imparts three different qualitative behaviors. At \textit{irrational times} (which are times \(t>0\) for which \(\frac{t}{\pi}\notin \mathbb{Q}\)), it evolves into continuous but non-differentiable fractal-like profile. At rational times \(t^\ast = \pi \frac pq\), \(q\neq 2\), the solution takes on a discontinuous and piecewise parabola behavior. On the other hand, at each specific rational time \(t_k^0 = \pi\frac{2k-1}{2}\), \(k \in \mathbb{Z}^+\), the quantization effect disappears, and the solution instantly becomes a continuously differentiable function, with an analytical expression.\N\NSection 3 is devoted to periodic initial-boundary value problems of more general bidirectional dispersive equations. They prove that, if the initial functions are of bounded variation, the dynamical evolution of such periodic problems depend essentially upon the large wave number asymptotics of the associated dispersion relations. They study in extent the case of monomial dispersion relation (Theorems 3.1 and 3.8), and are able to obtain that integral polynomial dispersive relations, or asymptotically integral polynomial dispersion relations, produce dispersive revival/fractalization at rational/irrational times, whereas with non-polynomial dispersive relations result in fractal profiles at all times.\N\NFinally, in Section 4, they present numerical simulations (using the Fourier spectral method and the classic fourth-order Runge-Kutta method) of dispersive revival for nonlinear equations, in the concrete case of a nonlinear beam equation, to demonstrate how such effects persist into the nonlinear regime.