Periodic traveling waves in a two-dimensional cylinder with saw-toothed boundary and their homogenization limit (Q875638)

The authors study the existence and the properties of traveling waves for a 2D curvature-driven motion. The problem is defined as \(V=\kappa +A\), where \( V \) is the normal velocity of the curve, \(\kappa \) is its curvature and \(A\) is a constant. The domain is defined as \[ \Omega _{\varepsilon }=\{(x,y)\in \mathbb{R}^{2}\mid -H-g_{\varepsilon }(y)<x<H+g_{\varepsilon }(y)\}, \] where \( g_{\varepsilon }\) is built by periodicity from a 1-periodic function \(g\) which satisfies \(\max_{y}g_{\varepsilon }(y)=\tan \alpha \) and \(\min g_{\varepsilon }(y)=-\tan \beta \), for angles \(\alpha \) and \(\beta \) lying in \((0,\pi /4)\). Assuming that the curve is the graph of a function \( y=u(x,t) \), the authors end with the evolution nonlinear equation \( u_{t}=u_{xx}/(1+u_{x}^{2})+A\sqrt{1+u_{x}^{2}}\), with nonhomogeneous Dirichlet boundary conditions on the saw-toothed periodic boundary of \( \Omega _{\varepsilon }\). The authors first prove an existence result for traveling waves \(U^{\varepsilon }\) of this problem. They also prove a stability result for such traveling waves. The last main result of the paper describes the asymptotic behaviour of the solution when \(\varepsilon \) goes to 0. Assuming some further conditions on the data of the problem, the authors prove that the average speed \(c_{\varepsilon }\) of the traveling wave stays in some interval \((c_{0},c_{0}+M\sqrt{\varepsilon })\) for some constant \(c_{0}\), and the convergence of \(U^{\varepsilon }\) to \(\varphi (x;c_{0})+c_{0}t\), in \(C_{loc}^{2,1}((-H,H)\times \mathbb{R})\), where \( \varphi \) is the solution of a stationary nonlinear elliptic equation. For the proof, the authors introduce some change of variables and comparison principles for this kind of problem. The authors then prove a global in time existence result. Thanks to a uniqueness result, they obtain the existence of a periodic traveling wave, which satisfies uniform estimates. The asymptotic behaviour is obtained using the comparison principle with known functions.