Convergence of a Semi-Discrete Numerical Method for a Class of Nonlocal Nonlinear Wave Equations

TL;DR

This paper proves the convergence of a semi-discrete numerical method for nonlocal nonlinear wave equations, demonstrating second-order accuracy and effectiveness in capturing solution blow-up through numerical experiments.

Contribution

It introduces and analyzes a semi-discrete scheme directly applied to nonlocal wave equations, establishing convergence, accuracy, and practical effectiveness.

Findings

Solutions converge uniformly as mesh size decreases

The method achieves second-order spatial convergence

Numerical experiments confirm convergence rate and blow-up detection

Abstract

In this article, we prove the convergence of a semi-discrete numerical method applied to a general class of nonlocal nonlinear wave equations where the nonlocality is introduced through the convolution operator in space. The most important characteristic of the numerical method is that it is directly applied to the nonlocal equation by introducing the discrete convolution operator. Starting from the continuous Cauchy problem defined on the real line, we first construct the discrete Cauchy problem on a uniform grid of the real line. Thus the semi-discretization in space of the continuous problem gives rise to an infinite system of ordinary differential equations in time. We show that the initial-value problem for this system is well-posed. We prove that solutions of the discrete problem converge uniformly to those of the continuous one as the mesh size goes to zero and that they are second-order convergent in space. We then consider a truncation of the infinite domain to a finite one. We prove that the solution of the truncated problem approximates the solution of the continuous problem when the truncated domain is sufficiently large. Finally, we present some numerical experiments that confirm numerically both the expected convergence rate of the semi-discrete scheme and the ability of the method to capture finite-time blow-up of solutions for various convolution kernels.